Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download the global epidemiology of infectious diseases
Document related concepts
Trichinosis wikipedia , lookup
Sexually transmitted infection wikipedia , lookup
Leptospirosis wikipedia , lookup
Neonatal infection wikipedia , lookup
Onchocerciasis wikipedia , lookup
Marburg virus disease wikipedia , lookup
Chagas disease wikipedia , lookup
African trypanosomiasis wikipedia , lookup
Whooping cough wikipedia , lookup
Middle East respiratory syndrome wikipedia , lookup
Coccidioidomycosis wikipedia , lookup
Oesophagostomum wikipedia , lookup
Hospital-acquired infection wikipedia , lookup
Neglected tropical diseases wikipedia , lookup
Transcript
GLOBAL BURDEN OF DISEASE AND INJURY SERIES VOLUME IV THE GLOBAL EPIDEMIOLOGY OF INFECTIOUS DISEASES Edited By: Christopher J. L. Murray Alan D. Lopez Colin D. Mathers World Health Organization Geneva G L O B A L B U R D E N O F D I S E A S E A N D I N J U RY S E R I E S VOLUME IV The Global Epidemiology of Infectious Diseases Edited by Christopher J. L. Murray Alan D. Lopez Colin D. Mathers World Health Organization Geneva WHO Library Cataloguing-in-Publication Data The global epidemiology of infectious diseases [electronic resource] / edited by Christopher J. L. Murray, Alan D. Lopez, Colin D. Mathers. 1 PDF. -- (Global burden of disease and injury series ; volume 4) 1.Communicable diseases - epidemiology I. Murray, Christopher J. L. II. Lopez, Alan D. III. Mathers, Colin D. ISBN 92 4 159230 3 (NLM classification: WC 100) © Copyright World Health Organization 2004. All rights reserved. Publications of the World Health Organization can be obtained from Marketing and Dissemination, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel: +41 22 791 2476; fax: +41 22 791 4857; email: [email protected]). Requests for permission to reproduce or translate WHO publications – whether for sale or for noncommercial distribution — should be addressed to Marketing and Dissemination, at the above address (fax: +41 22 791 4806; email: [email protected]). The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. The World Health Organization does not warrant that the information contained in this publication is complete and correct and shall not be liable for any damages incurred as a result of its use. The named editors alone are responsible for the views expressed in this publication. Design and production by Digital Design Group, Newton, MA USA Table of Contents List of Tables ........................................................................ v List of Figures ..................................................................... xi Global Burden of Disease and Injury Series Foreword Ralph H. Henderson ..........................................................xiv Global Burden of Disease and Injury Series Foreword Dean T. Jamison.................................................................xvi Preface ..............................................................................xxiv Acknowledgments..............................................................xxv Chapter 1: Diarrhoeal Diseases Caryn Bern .......................................................................... 1 Chapter 2: Pertussis Artur M Galazka, Susan E Robertson.................................. 29 Chapter 3: Diptheria Artur M Galazka, Susan E Robertson.................................. 55 Chapter 4: Measles CJ Clements, GD Hussey ................................................... 75 Chapter 5: Poliomyelitis Kenji Shibuya, Christopher JL Murray............................... 111 Chapter 6: Tetanus A Galazka, M Birmingham, M Kurian, FL Gasse ............... 151 Chapter 7: Leprosy D Daumerie..................................................................... 201 Chapter 8: Dengue and dengue haemorrhagic fever James W LeDuc, Karin Esteves, Norman G. Gratz ............. 219 iv Global Epidemiology of Infectious Diseases Chapter 9: Intestinal nematode infections DAP Bundy, MS Chan, GF Medley, D Jamison, L Savioli .... 243 Chapter 10: Trachoma-Related Visual Loss B Thylefors, K Ranson, A-D Négrel, R Pararajasegaram...... 301 Chapter 11: Chagas Disease A Moncayo ...................................................................... 325 Chapter 12: Schistosomiasis KE Mott .......................................................................... 349 Table of Contents v List of Tables Chapter 1 1.1 International Classification of Diseases (ICD) codes corresponding to watery diarrhoea and dysentery ............ 3 1.2 Previous estimates of diarrhoeal disease burden for children younger than 5 years in developing countries ................... 4 1.3 Diarrhoeal morbidity and mortality in 276 surveys in children younger than 5 years, using WHO methodology, 1981–1986 .................................................................... 5 1.4 Review of empirical databases on morbidity from diarrhoea: longitudinal studies measuring incidence ........ 7 1.5 Review of empirical databases on mortality from diarrhoea: studies measuring annual deaths per 1000 children ........ 10 1.6 Review of databases on mortality from diarrhoea: estimated from vital registration data in Latin America and the Caribbean .............................................................. 13 1.7 Review of empirical databases on morbidity from diarrhoea: morbidity in various age groups .................... 14 1.8 Review of empirical databases on mortality from diarrhoea: various age groups ....................................................... 15 1.9 Studies of persistent diarrhoea and distribution of mortality by type of diarrhoea ..................................................... 16 1.10 Regional summary estimates for morbidity from diarrhoeal disease ........................................................................ 18 1.11 Regional summary estimates for mortality from diarrhoeal disease ........................................................................ 18 1.12 Estimates for distribution of episodes by type of diarrhoea according to age .......................................................... 19 1.13 Assignment of disability weights according to type of diarrhoea; classification used for calculation of disabilityadjusted life years ........................................................ 19 Chapter 2 2.1 Pertussis attack rates, by age, during pertussis outbreaks in Côte d’Ivoire, Guatemala, and Indonesia ................... 33 2.2 Pertussis case fatality rates, by age, United States of America 1927–1989, England and Wales 1980–1991, and Kenya 1974–1976 ........................................................ 36 2.3 Complication rates attributable to pneumonia, seizures, and encephalopathy among pertussis patients in seven studies ......................................................................... 38 2.4 Ninth Revision of the International Classification of Diseases (ICD-9) codes for pertussis ............................. 41 2.5 Number of pertussis cases reported to WHO, by demographic region, 1980, 1985 and 1990 .................... 43 vi Global Epidemiology of Infectious Diseases 2.6 World Health Organization EPI Information System method for estimating number of cases and deaths attributable to pertussis ................................................ 45 2.7 Assumptions concerning completeness of reporting and case fatality rates used in estimating the number of pertussis cases and deaths by demographic region, 1990 .............. 46 2.8 Estimated global number of pertussis cases and deaths in 1990, based on two methods of estimation .................... 47 2.9 Estimated number of pertussis cases, by age, in developing countries, developed countries, and worldwide, 1990 ..... 47 2.10 Estimated number of pertussis deaths, by age, developing countries, developed countries, and worldwide, 1990 ..... 47 2.11 Regional and global levels of coverage with three doses of diphtheria-pertussis-tetanus (DPT) vaccine among children by age 12 months for 1985, 1990 and 1995 ................... 48 Chapter 3 3.1 Ninth revision of the International Classification of Diseases (ICD-9) codes for diphtheria ........................... 58 3.2 Number of diphtheria cases reported to the World Health Organization from developing and developed countries, and percentage of total cases occurring in developing countries, 1974–1994, based on reports received as of March 1996 ................................................................. 60 3.3 Cyclic occurrence of diphtheria in the world, 16th–20th centuries ...................................................................... 62 3.4 Mortality attributable to diphtheria, by age groups, Hampton Falls, New Hampshire, 1735 .......................... 63 3.5 Diphtheria case fatality rates in urban and rural areas, Ukraine, 1992 and 1993 ............................................... 64 3.6 Estimated number cases of diphtheria, and proportion, by age group, in developing countries, developed countries, and worldwide, 1990.................................................... 66 3.7 Estimated number of diphtheria cases and deaths, by age group, worldwide, 1990 ............................................... 66 3.8 Estimated number of diphtheria cases and deaths, by World development report region, 1990................................... 67 Chapter 4 4.1 Complications of measles.............................................. 78 4.2 Studies from five countries on the frequency of complications in measles (as a percentage) ..................... 78 4.3 Distribution of measles deaths by cause ......................... 80 4.4 Measles case fatality rates in prospective community studies, Africa .............................................................. 81 Table of Contents vii 4.5 Measles case fatality rates in prospective community studies, Americas ........................................................ 83 4.6 Measles case fatality rates in prospective community studies, Asia................................................................ 84 4.7 The estimated burden of measles by World Bank Region 88 4.8 Annual estimated burden of measles by World Bank Region, 1990, unadjusted for background mortality....... 90 4.9 Incidence rates of pneumonia, croup and diarrhoea and respective case fatality rates in hospitalized patients with measles ....................................................................... 94 4.10 Frequency of bacterial isolates and complications following lung and tracheal aspirates in measles-associated pneumonia ................................................................... 95 4.11 Association between nutritional status and complications97 4.12 Frequency of central nervous system complications in measles ........................................................................ 99 Chapter 5 5.1 Percentage distribution of notified cases of poliomyelitis by age-group and median age of notified cases: England and Wales, 1912–19 and 1944–50 ..................................... 113 5.2 Annual incidence of poliomyelitis per 100 000 total population during the period 1976–1980..................... 114 5.3 Poliomyelitis cases in Greater Bombay: clinical outcome118 5.4 Change in OPV coverage (percentage) by region .......... 123 5.5a Epidemiological estimates for poliomyelitis, 1990, males ......................................................................... 127 5.5b Epidemiological estimates for poliomyelitis, 1990, females ...................................................................... 130 Chapter 6 6.1 Case fatality ratios in tetanus patients by age, India 1954– 1968 and 1975, Nigeria 1974–1984 ............................ 155 6.2 Case fatality ratios associated with tetanus, 1920–1995 157 6.3 Age distribution of tetanus cases (percentage of cases < 20 and > 50 years of age), 1954–1992 ...................... 163 6.4 Gender ratios of tetanus cases and deaths, 1944–1990 . 165 6.5 Notification efficiencies of tetanus cases, 1969–1992 ... 167 6.6 Estimated neonatal mortality rates and neonatal tetanus mortality rates, and proportion of all neonatal deaths due to tetanus, 1953–1995 ................................................ 172 6.7 Estimated number of non-neonatal tetanus cases and deaths by age in developing and developed countries, 1990..... 182 6.8 Percentage of children immunized in the first year of life with 3 doses of DTP vaccine and percentage of pregnant viii Global Epidemiology of Infectious Diseases women immunized with at least 2 doses of tetanus toxoid, by WHO Region, 1993 ............................................... 186 Chapter 7 7.1 Detection and prevalence of leprosy in five African countries, 1986–1989 ................................................. 208 7.2 Detection rates per 1000 by age and sex in five African countries (people 0–14 years of age not included in the sample) ..................................................................... 209 7.3 Detection by age group and type of leprosy in four districts, India, 1993 .................................................. 209 7.4 Incidence rates by age, sex and type of leprosy, Polambakkam, India, 1967–1971 ............................... 210 7.5 Detection rates of leprosy, Brazil: actual rates, 1950 to 1983; predicted rate, 2000 .......................................... 210 7.6 Age-specific and type-specific prevalence of leprosy in four Governorates, Egypt, 1986 ......................................... 211 7.7 Prevalence of disability among leprosy patients before the implementation of multidrug therapy: results of various studies ....................................................................... 212 7.8 Impact of multidrug therapy on disability related to leprosy—prevalence of disability: results of various studies ....................................................................... 214 Chapter 8 8.1 Estimates of the burden of dengue haemorrhagic fever, Other Asia and Islands ............................................... 225 8.2 Estimates of the burden of dengue haemorrhagic fever, India.......................................................................... 227 8.3 Estimates of the burden of dengue haemorrhagic fever, China ........................................................................ 228 8.4 Estimates of the burden of dengue haemorrhagic fever, SubSaharan Africa ........................................................... 229 8.5 Reported cases of dengue fever, dengue haemorrhagic fever, and deaths attributable to dengue haemorrhagic fever in Latin America and the Caribbean, 1988 to 1992 .......... 230 8.6 Estimates of the burden of dengue haemorrhagic fever, Latin America and the Caribbean ................................ 231 8.7 Dengue haemorrhagic fever reported, selected countries, Other Asia and Islands, 1982–1992............................. 233 8.8 Crude estimate for the global burden of disease attributable to dengue haemorrhagic fever, 1994 ............................ 234 Chapter 9 9.1 Worm burden thresholds for morbidity used in the model ........................................................................ 251 Table of Contents 9.2 9.3 9.4a 9.4b 9.4c 9.5a 9.5b 9.6a 9.6b 9.7a 9.7b 9.8 9.9 9.10 9.11a 9.11b 9.12a 9.12b ix Age weights for prevalences used in the model ............. 252 Prevalence classes and reference (set) prevalences used in the model................................................................... 256 Transition matrix for Ascaris lumbricoides .................. 257 Transition matrix for Trichuris trichiura...................... 257 Transition matrix for hookworms ............................... 257 Estimates of numbers infected and at risk of disability for Ascaris lumbricoides, by World Bank region ............... 259 Estimates of numbers infected and at risk of disability for Ascaris lumbricoides, by age group ............................. 259 Estimates of numbers infected and at risk of disability for Trichuris trichiura, by World Bank region ................... 260 Estimates of numbers infected and at risk of disability for Trichuris trichiura, by age group ................................. 260 Estimates of numbers infected and at risk of disabililty for hookworm, by World Bank region ............................... 261 Estimates of numbers infected and at risk of disability for hookworm, by age...................................................... 261 Summary of global burden of helminth infections, by World Bank region and country or area ................................. 263 Data for DALY calculation, ascariasis, both sexes, 1990.......................................................................... 268 Data for DALY calculation, trichuriasis both sexes, 1990.......................................................................... 270 Data for DALY calculation, hookworm, males, 1990.... 271 Data for DALY calculation, hookworm, females, 1990 . 272 Selected studies investigating the effects of parasitic helminth infections on mental processing..................... 274 Selected studies investigating the effects of parasitic helminth infections on physical fitness......................... 279 Chapter 10 10.1 Review of available data on trachomatous blindness: population-based surveys after 1974 ........................... 305 10.2 Ratio of people with low vision (< 6/18 – 3/60) to people with blindness (< 3/60) ............................................... 312 10.3 Ratio of female prevalence of trachomatous blindness to male prevalence.......................................................... 313 10.4 Age distribution of the prevalence of trachomatous blindness ................................................................... 313 10.5 Prevalence of trachomatous blindness and low vision by region ........................................................................ 313 10.6 Age and sex distribution of the people with blindness or low vision attributable to trachoma, China, 1990 ........ 314 10.7 Age and sex distribution of the people with blindness or low vision attributable to trachoma, India, 1990.......... 314 x Global Epidemiology of Infectious Diseases 10.8 Age and sex distribution of the people with blindness or low vision attributable to trachoma, Middle Eastern Crescent, 1990 ........................................................... 315 10.9 Age and sex distribution of the people with blindness or low vision attributable to trachoma, Other Asia and Islands, 1990 ............................................................. 315 10.10 Age and sex distribution of the people with blindness or low vision attributable to trachoma, Sub-Saharan Africa, 1990.......................................................................... 316 10.11 Age and sex distribution of the people with blindness or low vision attributable due to trachoma throughout the world, 1990 ............................................................... 316 Chapter 11 11.1 Prevalence studies of human T. cruzi infection in Latin America .................................................................... 327 11.2 Prevalence of human T. cruzi infection in Latin America, 1975–1985 ................................................................ 328 11.3 Prevalence of Typarrosoma cruzi infected blood in blood banks of selected countries ......................................... 330 11.4 Incidence of cardiac lesions and T. cruzi infection, rates per 1000 person-years ...................................................... 331 11.5 Excess mortality rates with cardiac lesions, rates per 1000 person-years............................................................... 332 11.6 Efficacy of triatomine control tools: percentage of reinfested houses after different interventions, Argentina, Chile, Honduras and Paraguay, April 1994 .................. 335 11.7 Human infection by Trypanosoma cruzi and reduction of incidence, Southern Cone initiative for the elimination of Chagas disease: 1983–2000 ........................................ 336 11.8 Summary of control activities in Central America, 1999.......................................................................... 341 11.9 The time table towards the interruption of transmission of Chagas disease ........................................................... 343 Chapter 12 12.1 Diagnostic methods for schistosomiasis ....................... 350 12.2 S. Japonicum empirical databases, China .................... 357 12.3 S. japonicum empirical databases by region, Other Asia and Islands ....................................................................... 360 12.4 S. mansoni empirical databases, sub-Saharan Africa .... 362 12.5 S. intercalatum empirical databases by region, sub-Saharan Africa ....................................................................... 365 12.6 S. mansoni empirical databases by region, Latin America and the Caribbean ...................................................... 366 Table of Contents xi 12.7 S. haematobium empirical databases by region, Middle Eastern Crescent ....................................................... 368 12.8 S.mansoni empirical databases by region, Middle Eastern Crescent .................................................................... 369 12.9 Causes of death related to schistosomiasis ................... 371 12.10 Studies of mortality rates associated with schistosomiasis, World Bank regions .................................................... 372 12.11 Epidemiological estimates for schistosomiasis,1990, males ......................................................................... 376 12.12 Epidemiological estimates for schistosomiasis,1990, females ...................................................................... 380 xii Global Epidemiology of Infectious Diseases List of Figures Chapter 2 2.1 Percentage distribution of pertussis cases before (grey bars) and after (black bars) widespread use of pertussis vaccine, Poland (1971 versus 1991), United States (1918–21 versus 1980–89) .................................................................... 34 2.2 Reported annual incidence of pertussis, by WHO region, 1974–1994 .................................................................. 44 Chapter 3 3.1 Annual diphtheria incidence reported to the World Health Organization, by region, 1974–1994, based on reports received as of March 1996 ............................................ 61 3.2 Age-specific immunity of a hypothetical population before and after widespread use of diphtheria vaccine .............. 62 Chapter 4 4.1 Number of cases of measles and measles vaccine coverage reported globally, 1974–1996 ....................................... 76 4.2 Age of onset of measles in the pre-vaccine era ................ 87 Chapter 5 5.1 Global immunization coverage (per cent) of OPV3 for children in the first year of life.................................... 113 5.2 Reported incidence of indigenous poliomyelitis ............ 114 5.3 Annual incidence of paralytic poliomyelitis, 1990 ........ 126 5.4 Annual number of deaths due to poliomyelitis, 1990 .... 133 5.5 DALYs lost due to paralytic poliomyelitis, 1990........... 133 5.6 Immunization status of poliomyelitis cases in Shandong Province, 1990 ........................................................... 136 5.7 Reduction of polio incidence by region, 1990 .............. 137 5.8 Anticipated reduction of polio incidence with full vaccine coverage .................................................................... 138 Chapter 6 6.1 Incidence and mortality rates and case fatality ratios for neonatal tetanus, Denmark, 1920–1960....................... 156 6.2 Reported age-specific tetanus incidence rates, by 5-year intervals, United States, 1965–1994 ............................ 162 6.3 Number of tetanus cases and immunization coverage reported to WHO by national authorities, 1975–1995.. 169 6.4 Reported number of neonatal and non-neonatal tetanus cases by WHO region, 1974–1992 .............................. 170 Table of Contents 6.4 6.5 6.6 6.7 xiii Reported number of neonatal and non-neonatal tetanus cases by WHO region, 1974–1992 (continued)............. 171 Estimated number of tetanus deaths occurring in the world by WHO region, 1990 ................................................ 183 Estimated neonatal tetanus mortality rate per 1000 live births, 1990 ............................................................... 184 Reported coverage with at least two doses of tetanus toxoid among pregnant women, 1995.......................... 187 Chapter 11 11.1 Southern Cone Initiative for the elimination of transmission of Chagas disease: incidence of infection 1980–2000 ... 336 Foreword to The G LOBAL B URDEN OF D ISEASE AND I NJURY S ERIES Ralph H. Henderson The collection and use of timely and reliable health information in support of health policies and programmes have been actively promoted by the World Health Organization since its foundation. Valid health statistics are required at all levels of the health system, ranging from data for health services support at the local community level, through to national statistics and information used to monitor the effectiveness of national health strategies. Equally, regional and global data are required to monitor global epidemics and to continuously assess the effectiveness of global public health approaches to disease and injury prevention and control, as coordinated by WHO technical programmes. Despite the clear need for epidemiological data, reliable and comprehensive health statistics are not available in many Member states of WHO, and, indeed, in many countries the ascertainment of disease levels, patterns and trends is still very uncertain. In recent years, monitoring systems, community-level research and disease registers have improved in both scope and coverage. Simultaneously, research on the epidemiological transition has increased our understanding of how the cause structure of mortality changes as overall mortality rates decline. As a consequence, estimates and projections of various epidemiological parameters, such as incidence, prevalence and mortality, can now be made at the global and regional level for many diseases and injuries. The Global Burden of Disease Study has now provided the public health community with a set of consistent estimates of disease and injury rates in 1990. The Study has also attempted to provide a comparative index of the burden of each disease or injury, namely the number of DisabilityAdjusted Life Years (DALYs) lost as a result of either premature death or years lived with disability. The findings published in the Global Burden of Disease and Injury Series provide a unique and comprehensive assessment of the health of populations as the world enters the third millennium. We also expect that the methods described in the various volumes in the series will stimulate Foreword xv Member States to improve the functioning and usefulness of their own health information systems. Nonetheless, it must be borne in mind that the results from an undertaking as ambitious as the Global Burden of Disease Study can only be approximate. The reliability of the data for certain diseases, and for some regions, is extremely poor, with only scattered information available in some cases. To extrapolate from these sources to global estimates is clearly very hazardous, and could well result in errors of estimation. The methods that were used for some diseases (e.g. cancer) are not necessarily those applied by other scientists or institutions (e.g. the International Agency for Research on Cancer) and hence the results obtained may differ, sometimes considerably, from theirs. Moreover, the concept of the DALY as used in this Study is still under development, and further work is needed to assess the relevance of the social values that have been incorporated in the calculation of DALYs, as well as their applicability in different socio-cultural settings. In this regard, WHO and its various partners are continuing their efforts to investigate burden-of-disease measurements and their use in health policy decision-making. Dr Ralph H. Henderson is former Assistant Director-General of the World Health Organization. Foreword to The G LOBAL B URDEN OF D ISEASE AND I NJURY S ERIES Dean T. Jamison Rational evaluation of policies for health improvement requires four basic types of information: a detailed, reliable assessment of epidemiological conditions and the burden of disease; an inventory of the availability and disposition of resources for health (i.e. a system of what has become known as national health accounts); an assessment of the institutional and policy environment; and information on the cost-effectiveness of available technologies and strategies for health improvement. The Global Burden of Disease and Injury Series provides, on a global and regional level, a detailed and internally consistent approach to meeting the first of these information needs, that concerning epidemiological conditions and disease burden. It fully utilizes what information exists while, at the same time, pointing to great variation—across conditions and across countries—in data quality. In the Global Burden of Disease and Injury Series, Christopher Murray, Alan Lopez and literally scores of their collaborators from around the world present us with a tour de force: its volumes summarize epidemiological knowledge about all major conditions and most risk factors; they generate assessments of numbers of deaths by cause that are consistent with the total numbers of deaths by age, sex and region provided by demographers; they provide methodologies for and assessments of aggregate disease burden that combine—into the Disability-Adjusted Life Year or DALY measure—burden from premature mortality with that from living with disability; and they use historical trends in main determinants to project mortality and disease burden forward to 2020. Publication of the Global Burden of Disease and Injury Series marks the transition to a new era of health outcome accounting—an era for which these volumes establish vastly higher standards for rigour, comprehensiveness and internal consistency. I firmly predict that the official reporting of health outcomes in many countries and globally will increasingly embody the approach and standards described in this series. The Global Burden of Disease and Injury Series culminates an evolutionary process that began in the late 1980s. Close and effective collaboration Foreword xvii between the World Bank and the World Health Organization initiated, supported and contributed substantively to that process. Background Work heading to the Global Burden of Disease and Injury Series proceeded in three distinct phases beginning in 1988. Intellectual antecedents go back much further (see Murray 1996, or Morrow and Bryant 1995); perhaps the most relevant are Ghana’s systematic assessment of national health problems (Ghana Health Assessment Project Team 1981) and the introduction of the QALY (quality-adjusted life year)—see, for example Zeckhauser and Shepard (1976). My comments here focus on the three phases leading directly to the Global Burden of Disease and Injury Series. Phase 1 constituted an input to a four-year long “Health Sector Priorities Review” initiated in 1988 by the World Bank; its purpose was to assess “…the significance to public health of individual diseases for related clusters of diseases…and what is now known about the cost and effectiveness of relevant interventions for their control” (Jamison et al. 1993, p. 3; that volume provides the results of the review). Dr Christopher Murray of Harvard University introduced the DALY as a common measure of effectiveness for the review to use across interventions dealing with diverse diseases, and Dr Alan Lopez of the World Health Organization prepared estimates of child deaths by cause that were consistent with death totals provided by demographers at the World Bank (Lopez 1993). At the same time, and in close coordination, a World Bank effort was preparing consistent estimates for adult (15-59) mortality by cause for much of the developing world (Feachem et al. 1992). Ensuring this consistency was a major advance and is a precondition for systematic attempts to measure disease burden. (Estimates of numbers of deaths by cause that are not constrained to sum to a demographically-derived total seem inevitably to result in substantial overestimates of deaths due to each cause.) Phase 1 of this effort, then, introduced the DALY and established important consistency standards to guide estimation of numbers of deaths by cause. Phase 2 constituted the first attempt to provide a comprehensive set of estimates not only of numbers of deaths by cause but also of total disease burden including burden from disability. This effort was commissioned as background for the World Bank’s World Development Report 1993: Investing in Health; it was co-sponsored by the World Bank and the World Health Organization; and it was undertaken under the general guidance of a committee chaired by Dr JP Jardel (then Assistant Director-General of WHO). The actual work was conceptualised, managed and integrated by Drs Murray and Lopez and involved extensive efforts by a large number of individuals, most of whom were on the WHO staff. First publication of the estimates of 1990 disease burden appeared in Appendix B of Investing in Health (World Bank 1993) and, more extensively, in Murray, Lopez and Jamison (1994). The World Health Organization subsequently published xviii Global Epidemiology of Infectious Diseases a volume containing a full account of the methods used and a somewhat revised and far more extensive presentation of the results (Murray and Lopez 1994). Preparation and publication of the Global Burden of Disease and Injury Series constitutes Phase 3 of this sequence of efforts. As in the earlier phases, the Global Burden of Disease and Injury Series was undertaken to inform a policy analysis—in this case an assessment of priorities for health research and development in developing countries being guided by WHO’s Ad Hoc Committee on Health Research Relating to Future Intervention Options. The Committee sought updated estimates of disease burden for 1990, projections to 2020 and an extension of the methods to allow assessment of burden attributable to selected risk factors. The committee’s report (Ad Hoc Committee 1996) and the Global Burden of Disease and Injury Series appear as companion documents. Chapters summarizing results from the Global Burden of Disease and Injury Series appear in volume I, The Global Burden of Disease; underlying epidemiological statistics for over 200 conditions appear in volume II, Global Health Statistics. The next two volumes of the Series provide, for the first time, chapters detailing the data on each condition or cluster of conditions in the areas of reproductive health and infectious diseases. Will there be a fourth phase? I am sure there will. Reporting of the disease-specific and risk factor analysis in the Global Burden of Disease and Injury Series will provoke constructive and perhaps substantial criticism and improvements; country-specific assessments will multiply (over 20 are now under way) and they, too, will modify and enrich current estimates, see for example, Lozano et al. 1994. Estimates of trends in deaths by cause have recently been initiated, which will serve as the precursor for estimates of trends in disease burden. Methodologies will be criticized and, I would predict, constructively revised. Unfinished elements of the agenda discussed in the next section will be completed. Phase 4, centred on the estimation of global and regional disease burden for 2000 and including a look at the past, will take us well beyond where we now are. The Agenda Disease burden (or numbers of deaths by cause) were initially partitioned in three separate ways for different age, sex and regional groupings (Murray et al. 1994). One partition is by risk factor—genetic, behavioural, environmental and physiological. The second is by disease. The third is by consequence— consequence—premature mortality at different ages and different types of disability (e.g. sensory, cognitive functioning, pain, affective state, etc.). A fourth partition, by reason for remaining burden, was later introduced (Ad Hoc Committee, 1996). Disaggregation by risk factor helps guide policy concerning primary and secondary prevention, including development of new preventive measures. Disaggregation by disease helps guide policy concerning cure, secondary Foreword xix prevention and palliation; and disaggregation by consequence helps guide policies for rehabilitation. Work on the disease burden assessment agenda began with assessments of mortality and burden by disease; the Global Burden of Disease and Injury Series advances the agenda in that domain by revising and adding great detail on disease burden estimates. More recently, the World Health Report for 2002 (World Health Organization, 2002) provides an extensive assessment of the burden from a range of important risk factors; this extends usefulness of the work to the domain of prevention policy. There remains, however, an important unfinished agenda. The disease burden associated with different types of disability remains to be assessed; perhaps part of the reason for neglect of rehabilitation in most discussion of health policy is the lack of even approximate information on burden due to disability or on the DALY gains per unit cost of rehabilitative intervention. A related agenda item—relevant to planning for curative and, particularly, rehabilitative intervention—is to present disease burden estimates from a current prevalence perspective. The dominant perspective of work so far undertaken, including in the Global Burden of Disease and Injury Series, is that of adding up over time the burden that will result from all conditions incident in a given year (here 1990); this well serves the development of primary prevention policy and of treatment policy for diseases of short duration. The prevalence perspective complements the incidence one by assessing how much burden is being experienced during a particular year by chronic conditions or by disabilities; those conditions or disabilities will often have been generated sometime in the past. From an incidence perspective, disability in this year from, say, an injury occurring a decade ago would generate DALY loss in the year of incidence; but to guide investment in rehabilitation we need to know how much disability exists today, i.e., we need a prevalence perspective. Murray and Lopez in The Global Burden of Disease and in Global Health Statistics provide the basic estimates of prevalence of different disabilities and first glimpses of the prevalence perspective. A final major agenda item is to establish for each condition and in the aggregate how much of the potential current burden is in fact being averted by existing interventions and how much of the remaining burden persists because of lack of any intervention, lack of cost-effective interventions, or because of inefficiency of the system. Uses of DALY and Disease Burden Measurement DALYs have six major uses to underpin health policy. Five of these relate to measurement of the burden of disease; the final one concerns judging the relative priority of interventions in terms of cost-effectiveness. Assessing performance. A country-specific (or regional) assessment of the xx Global Epidemiology of Infectious Diseases burden of disease provides a performance indicator that can be used over time to judge progress or across countries or regions to judge relative performance. These comparisons can be either quite aggregated (in terms of DALYs lost per thousand population ) or finely disaggregated to allow focused assessment of where relative performance is good and where it is not. The Global Burden of Disease and Injury Series with its burden assessments for eight regions in 1990—and with the increasing number of country-specific assessments that it will report—will provide, I predict, the benchmark for all subsequent work. The most natural comparison is to the development of National Income and Product Accounts (NIPAs) by Simon Kuznets and others in the 1930s, which culminated in 1939 with a complete NIPA for the United Kingdom prepared by James Meade and Richard Stone at the request of the UK Treasury. NIPAs have, in the subsequent decades, transformed the empirical underpinnings of economic policy analysis. One of the leading proponents of major changes in NIPAs has put in this way: The national income and product accounts for the United States (NIPAs), and kindred accounts in other nations, have been among the major contributions to economic knowledge over the past half century…Several generations of economists and practitioners have now been able to tie theoretical constructs of income output, investment, consumption, and savings to the actual numbers of these remarkable accounts with all their fine detail and soundly meshed interrelations. (Eisner 1989). My own expectation is that this series will, over a decade or two, initiate a transformation of health policy analysis analogous to that initiated for economic policy by the introduction of NIPAs in the late 1930s. Today most health policy work concerns only cost, finance, process and access; burden of disease (and risk factor) assessments should soon allow full incorporation of performance measures in policy analysis. Generating a forum for informed debate of values and priorities. The assessment of disease burden in a country-specific context in practice involves participation of a broad range of national disease specialists, epidemiologists and, often, policy makers. Debating the appropriate values for, say, disability weights or for years of life lost at different ages helps clarify values and objectives for national health policy. Discussing the inter-relations among diseases and their risk factors in the light of local conditions sharpens consideration of priorities. And the entire process brings technically informed participants to the table where policy is discussed. The preparation of a well-defined product generates a process with much value of its own. Identifying national control priorities. Many countries now identify a relatively short list of interventions, the full implementation of which Foreword xxi becomes an explicit priority for national political and administrative attention. Examples include interventions to control tuberculosis, poliomyelitis, HIV infection, smoking and specific micronutrient deficiencies. Because political attention and administrative capacity are in relatively fixed and short supply, the benefits from using those resources will be maximized if they are directed to interventions that are both cost-effective and aimed at problems associated with a high burden. Thus, national assessments of disease burden are instrumental for establishing this short list of control priorities. Allocating training time for clinical and public health practitioners. Medical schools offer a fixed number of instructional hours; training programmes for other levels and types of practitioners are likewise limited. A major instrument for implementing policy priorities is to allocate this fixed time resource well—again that means allocation of time to training in interventions where disease burden is high and cost-effective interventions exist. Allocating research and development resources. Whenever a fixed effort will have a benefit proportional not to the size of the effort but rather to the size of the problem being addressed, estimates of disease burden become essential for formulation of policy. This is the case with political attention and with time in the medical school curriculum; and it is likewise true for the allocation of research and development resources. Developing a vaccine for a broad range of viral pneumonias, for example, would have perhaps hundreds of times the impact of a vaccine against disease from Hanta virus infection. Thus information on disease or risk factor burden is one (of several) vital inputs to inform research and development resource allocation. Indeed, as previously noted, this series—with its disease burden assessments for 1990, its projections to 2020 and its initial assessment of burden due to risk factors—was commissioned to inform a WHO Committee charged with assessing health research and development priorities for developing countries (Ad Hoc Committee 1996). The Committee sought not only to know the burden by condition, but also to partition the burden remaining for each condition into several distinct parts reflecting the importance of the reasons for the remaining disease burden. This division into four parts was undertaken for several conditions; a major agenda item for future analysis is to undertake such a partitioning systematically for all conditions so that there could be reasonably approximate answers to such questions as “How much of the remaining disease burden cannot be addressed without major biomedical advances?" Or, “How much of the remaining disease burden could be averted by utilizing existing interventions more efficiently?” Arguably most of the spectacular gains in human health of the past century have resulted from advances in knowledge (although improvements in income and education have also played a role). If so, improving research and development policy in health may be more important than improving xxii Global Epidemiology of Infectious Diseases policy in health systems or finance; improved assessments and projections of disease burden will be critical to that undertaking. Allocating resources across health interventions. Here disease burden assessment often plays a minor role; the task is to shift resources to interventions which, at the margin, will generate the greatest reduction in DALY loss. When there are major fixed costs in mounting an intervention—as is the case with political and managerial attention for national control priorities—burden estimates are indeed required to optimize resources allocation. But, typically, much progress can be made with only an understanding of how the DALYs gained from an intervention vary with the level of expenditure on it; such assessments are the stuff of cost-effectiveness analysis. The DALY as a common measure of effectiveness allows comparison of cost-effectiveness across intervention addressing all conditions; such an initial effort was undertaken for the World Bank’s “Health Sector Priorities Review” (Jamison et al. 1993) in the late 1980s using a forerunner to the DALY utilized in this series. ——— The Global Burden of disease and Injury Series contains the only available internally consistent, comprehensive and comparable assessments of causes of death, incidence and prevalence of disease and injury, measures and projections of disease burden, and measures of risk factor burden. In that sense the authors’ contributions represent a landmark achievement and provide an invaluable resource for policy analyst and scholars. This effort dramatically raises the standard by which future reporting of health conditions will be judged. Yet, the very need for the ad hoc assessments that the volumes in this series report, points to important gaps in the international system for gathering, analysing and distributing policy-relevant data on the health of populations. Without information on how levels and trends in key indicators in their own countries compare with other countries, national decision-makers will lack benchmarks for judging performance. Likewise students of health systems will lack the empirical basis for forming outcome-based judgements on which policies work—and which do not. I hope, then, that one follow-on to the Global Burden of Disease and Injury Series will be the institutionalization of continued efforts to generate and analyse internationally comparable data on health outcomes. Dean T. Jamison is Professor of Public Health and of Education at the University of California Los Angeles, and a Fellow at the Fogarty International Center of the National Institutes for Health. He recently served as Director, Economics Advisory Service, at the World Health Organization in Geneva, and is currently leading the preparation of the second edition of Disease Control Priorities in Developing Countries. Foreword xxiii References Ad Hoc Committee on Health Research Relating to Future Intervention Options (1996) Investing in health research and development. World Health Organization. Geneva (Document TDR/Gen/96/1). Eisner R (1989) The total incomes system of accounts. Chicago, University of Chicago Press. Feachem RGA et al., eds. (1992) The health of adults in the developing world. New York, Oxford University Press for the World Bank. Ghana Health Assessment Project team (1981) A quantitative method of assessing the health impact of different diseases in less developed countries. International Journal of Epidemiology, 10:73-80. Goerdt A et al. (1996) Disability: definition and measurement issues. In: Murray CJL and Lopez AD, eds., The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020. Cambridge, Harvard University Press. Jamison DT et al., eds. (1993) Disease control priorities in developing countries. New York, Oxford University Press for the World Bank. Lopez AD (1993) Causes of death in industrial and developing countries: estimates for 1985-1990. In: Jamison DT et al., eds. Disease control priorities in developing countries. New York, Oxford University Press for the World Bank, 35-50. Lozano R et al. (1994) El peso de la enfermedad en México: un doble reto. [The national burden of disease in México: a double challenge.] Mexico, Mexican Health Foundation, (documentos para el análisis y la convergencia, No. 3). Morrow RH, Bryant JH (1995) Health policy approaches to measuring and valuing human life: conceptual and ethical issues. American journal of public health, 85(10):1356-1360. Murray CJL (1996) Rethinking DALYs. In: Murray CJL and Lopez AD, eds., The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020. Cambridge, Harvard University Press. Murray CJL and Lopez AD, eds. (1994) Global comparative assessments in the health sector: disease burden, expenditures and interventions packages. Geneva, World Health Organization. Murray CJL, Lopez AD, Jamison DT (1994) The global burden of disease in 1990: summary results, sensitivity analysis and future directions. Bulletin of the World Health Organization, 72(3):495-509. World Bank (1993) World development report 1993: investing in health. New York, Oxford University Press for the World Bank. World Health Organization. World health statistics annual, various years. Geneva, WHO. World Health Organization (2002). World health report 2002: reducing risks, promoting healthy life, Geneva, WHO. Zeckhauser R, Shepard D (1976) Where now for saving lives? Law and contemporary problems, 40: 5-45. Preface This fourth volume of the Global Burden of Disease and Injuries Series provides the reader with information on the epidemiology and burden of major infectious and parasitic diseases. As with previous volumes of the Global Burden of Disease study, the chapters in this book detail the situation as experienced in the year 1990. Since then the epidemiology of some of the conditions described has changed, and where this is the case the authors have added a brief paragraph acknowledging this. The chapters therefore do not provide a detailed update on the current burden of disease, which is accommodated in the documentation of the Global Burden of Disease 2000 and published elsewhere. In Chapter 1 Bern focuses on diarrhoeal diseases which were a major contributor to the global burden of infectious disease in the year 1990 and still account for high numbers of infections in childhood. This is followed by five detailed and focussed chapters on the burden of vaccine-preventable diseases. In Chapters 2 and 3 Robertson and Galazka describe the epidemiology and burden of pertussis and diphtheria, respectively. In Chapter 4 Clemens and colleagues concentrate on the global epidemiology of measles which has undergone significant change in recent years. While approximately 4 million children died of measles in 1990, it is estimated that this number has fallen well below 1 million in 2000. In Chapter 5 Shibuya and Murray summarise the global burden of poliomyelitis, which has similarly been transformed by recent efforts in poliomyelitis eradication. Birmingham and Galazka give a detailed account of the epidemiology of tetanus in Chapter 6. In Chapter 7 Daumerie informs the reader about the global situation of leprosy, followed by Chapter 8 in which LeDuc describes the epidemiology of dengue. Chapter 9 by Bundy and colleagues provides an overview of intestinal nematode infestations, which give rise to considerable disease burden world-wide. In Chapter 10 Thylefors and colleagues focus on the global burden of trachoma whose epidemiology has undergone some change in the recent past. Moncayo gives a detailed account of the burden Preface xxv of Chagas’ Disease in the Americas in Chapter 11 and provides a pertinent update on recent efforts to eliminate the disease. The volume concludes with Chapter 12 by the late Mott and finalised by Mathers who describe the epidemiology of schistosomiasis. This volume summarises the burden of twelve major infectious and parasitic diseases. Volume 3 also contained chapters detailing the global epidemiology of HIV/AIDS and of other sexually transmitted diseases in 1990. These chapters provide information on the data sources and methodological approaches to measuring infectious disease burden that are still relevant today for analysis of current burden, as well as for interpreting time trends between 1990 and 2000 estimates of infectious disease burden. It is our hope that these chapters will also stimulate further debate about the availability and validity of the data, methods and assumptions used to arrive at disease estimates and spur efforts to improve the availability and quality of epidemiological data on these diseases. This will be the final volume of the original Global Burden of Disease and Injuries Series. The Global Burden of Disease study has stimulated further constructive debate, methodological developments and new analyses in summary measures of population health (Murray et al. 2002), in comparative risk assessment (Ezzati et al. 2004) and in the assessment of priorities for interventions (Tan Torres Edejer et al. 2003). Methods and data sources for the assessment of the burden of nutritional deficiencies, non-communicable diseases and injuries have also evolved and are being progressively documented as part of the Global Burden of Disease 2000 project (draft documentation is available at www.who.int/evidence/bod). Substantial documentation of the Global Burden of Disease 2000 study has also been published in peer reviewed journals and books, and we anticipate that there will be further publications containing revised and more extensive presentation of the methods and results. CJLM ADL CDM References Ezzati M, Lopez A, Rodgers A, Murray CJL (2004) Comparative quantification of health risks: global and regional burden of disease attributable to several major risk factors. Geneva, WHO. Murray CJL., Salomon, JA, Mathers CD, Lopez AD (2002) Summary measures of population health: concepts, ethics, measurement and applications. Geneva, WHO. Tan-Torres Edejer T, Baltussen R, Adam T, Hutubessy R, Acharya A, Evans DB, Murray CJL (2003) Making choices in health: WHO guide to cost-effectiveness analysis. Geneva, WHO. Acknowledgments Numerous people have contributed to bring this fourth volume of the 1990 Global Burden of Disease and Injuries Series to conclusion. First and foremost we thank Claudia Stein and Christina Bernard, our editorial managers who were instrumental in ensuring the completion of this task. They pursued chapter submissions with dogged determination, edited the manuscripts, critically reviewed them for consistency with the numbers published in Volumes 1 and 2, and organised the editorial process. Their efforts have been invaluable in achieving the production of this book and are very gratefully acknowledged. We are grateful to David Bramley for expert legal advice, Kai Lashley for excellent editorial assistance, and Kenji Shibuya and Lara Wolfson for much valued scientific input. We gratefully acknowledge the efforts of Emmanuela Gakidou in pursuing chapter submissions in the earlier stages of the process and Marie-Claude von Rulach and Susan Piccolo for superb secretarial support. We also thank Soma Ganguli, Michaela Herinkova, Petra Schuster for additional secretarial support. This volume has been in preparation for some time, and where authors were unable to finalise their chapters others have completed the task, where possible. Former authors who were unable to conclude their work are acknowledged as co-authors in this volume. We are particularly grateful for the pioneering manuscripts of those who sadly passed away while preparing this publication: A Galazka and K Mott. To acknowledge their invaluable contribution to this volume and burden of disease epidemiology at large, Maureen Birmingham and Susan Robertson have kindly taken on the task to complete the chapters by A Galazka, and the editors that of K Mott. CJLM ADL CDM Chapter 1 Diarrhoeal Diseases Caryn Bern Introduction Diarrhoea remains a leading cause of morbidity and mortality in the world, predominantly affecting children in developing countries. Each child in the world experiences an average of one to three episodes of diarrhoea per year, with incidence rates as high as 10 per year for children in some areas (Bern et al. 1992b). During the 1990s diarrhoea was estimated to cause 20 to 25 per cent of mortality among children younger than 5 years in the developing world (Bern et al. 1992b). Effective interventions, including correct case management (oral rehydration therapy, continued feeding and antibiotics in cases of dysentery), promotion of breastfeeding, and better weaning practices, have the potential to reduce the burden of diarrhoeal disease substantially in the future. This chapter reviews the global burden of morbidity and mortality from infectious diarrhoea, especially dysentery and persistent diarrhoea, as well as examining the issues involved in reducing that burden of disease through existing interventions. Aetiology And Clinical Patterns While non-infectious causes of diarrhoea may play a role, for example among the hospitalized elderly (Lew et al. 1991), the most important aetiological agents of diarrhoea are viral and bacterial, with rotavirus and enterotoxigenic E. coli generally identified most frequently among children in developing countries, and viral aetiologies playing a proportionately greater role in industrialized settings (Bern & Glass 1994, Huilan et al. 1991). Even among the elderly, infectious diarrhoea may be an important contributing cause of mortality (Lew et al. 1991). Diarrhoea causes morbidity and mortality through several mechanisms. Acute watery diarrhoea can lead to dehydration severe enough to require hospitalization and to cause death. Rotavirus is the agent most frequently associated with acute dehydrating diarrhoea, and primarily affects children 2 Global Epidemiology of Infectious Diseases younger than 2 years (Kapikian & Chanock 1990). Vibrio cholerae can cause epidemics of dehydrating diarrhoea affecting all age groups, and may lead to high case-fatality rates in the absence of rapid public health intervention, as in the recent epidemic among Rwandan refugees in Goma, Zaire (Goma Epidemiology Group 1995). Dysentery, most often associated with Shigella species, may cause death through bacteraemia or hypoglycaemia (Bennish 1991), but also is associated with a more marked effect on growth (Black, Brown & Becker 1984a). Dysentery may present a refractory problem in that case management depends upon treatment with antibiotics, and resistance is a widespread problem in areas with high rates of dysenteric morbidity and mortality (Salam & Bennish 1991). Persistent diarrhoea, defined as diarrhoea lasting at least 14 days (World Health Organization 1988), has been recognized recently as an entity which carries a risk of both nutritional compromise and of mortality in excess of that for acute diarrhoea (Bhan et al. 1989, Victora et al. 1993). While enteroaggregative E. coli and cryptosporidium have been associated with persistent diarrhoea (Bhan et al. 1989, Baqui et al. 1992a, Henry et al. 1992), often no specific pathogen can be implicated (Lanata et al. 1992), and serial infection with different organisms may play a role (Baqui et al. 1992a). Definitions and codes Diarrhoea is a symptom complex defined by an increased number of stools of a looser consistency than usual per 24-hour period. In a comparison of the definitions commonly used in epidemiological studies, Baqui et al. (1991) reported that those with the best balance of sensitivity and specificity were 3 or more loose stools, or any number of stools containing blood, in a 24-hour period, while the optimal definition of the end of an episode was 3 diarrhoea-free days. The authors concluded that differences in definitions, especially of the end of the episode, made a substantial difference to the final estimate of diarrhoeal incidence, and that validation of the definition of diarrhoea was important to the overall validity of a study. The ninth and tenth revisions of the International Classification of Diseases (ICD-9 and ICD-10) classify diarrhoea according to aetiological agent, rather than symptom complex. Because a number of agents may cause either watery or bloody diarrhoea, and because persistent diarrhoea is defined by duration of illness rather than aetiology, the ICD codes do not correspond neatly to the symptom complexes described in this chapter. An approximate mapping of symptom complexes and codes is, however, possible (Table 1.1). Persistent diarrhoea, which is discussed below, has been associated with a variety of aetiological agents which also cause acute watery diarrhoea and dysentery. Thus it is not included in the table, but may fall under any of several ICD classifications, depending on what agent, if any, is identified. Diarrhoeal Diseases Table 1.1 3 International Classification of Diseases (ICD) codes corresponding to watery diarrhoea and dysentery Symptom complex Aetiological agent ICD-8 and 9 ICD-10 Watery diarrhoea Cholera Salmonella gastroenteritis Bacterial food poisoning Other bacteria Giardiasis Viral gastroenteritis Gastroenteritis, presumed infectious 001 003 005 008.2–008.5 007.1 008.6 009.1, 009.3 A00 A02 A05 A04.0–A04.9 A07.1 A08 A09 Dysentery Shigella Campylobacter Yersinia Amoebiasis Dysentery, aetiology unspecified 004 008.43 008.44 006 009.0, 009.2 A03 A04.5 A04.6 A06 A09 Methods for the measurement of diarrhoeal morbidity and mortality Community longitudinal studies provide the most reliable data for diarrhoeal disease incidence and, if the study size is large enough, mortality estimates. Even for longitudinal studies, however, morbidity estimates can be shown to vary with the intensity of surveillance, with twice-weekly surveillance resulting in higher estimates than less frequent surveillance (Bern et al. 1992b). For example, in one study that made regional estimates for diarrhoeal morbidity, the reported number of episodes per child per year was consistently higher in studies in Latin America than in studies in India or Bangladesh, but the frequency of surveillance in the Latin American studies was also consistently higher (Bern et al. 1992b). However, in a more recent study conducted in India with home visits every 3 days, the incidence of diarrhoea was found to be similar to that in the most frequently surveyed Latin American communities (Bhandari, Bhan & Sazawal 1994). This suggests that among children in the poorest communities, diarrhoeal disease incidence is remarkably similar around the world, and that in studies with less frequent surveillance, the shorter or less severe episodes are likely to be missed. At the same time, several studies which monitored diarrhoeal morbidity among children from different socioeconomic strata in the same locality have demonstrated substantial variation in incidence rates, with poorer children experiencing from 2 to 6 times the number of annual episodes as more affluent children (Araya et al. 1986, Guerrant et al. 1983). While longitudinal studies are best for measurement of incidence of disease, the number of individuals followed is generally too small to allow measurement of mortality. Hence active surveillance, survey data and vital registration are the principal methods used for mortality assessment. For developing countries, where nearly all mortality from diarrhoea occurs, vital registration is unreliable or incomplete, and a so-called “gold 4 Global Epidemiology of Infectious Diseases standard” does not exist for its assessment. In addition, where vital registration is incomplete or absent, cause of death is ascertained on the basis of verbal autopsy methods, and the sensitivity and specificity of those methods for the diagnosis of diarrhoea and of dehydration have varied substantially in different evaluations (Snow et al. 1992, Kalter et al. 1990). In one study, verbal autopsy identified diarrhoeal deaths with a specificity greater than 80 per cent, but a sensitivity less than 50 per cent (Snow et al. 1992), while in another, the performance of verbal autopsy for the diagnosis of diarrhoea was adequate, but the identification of deaths from severe dehydration was problematic (Kalter et al. 1990). Nevertheless, estimates based on these methods reflect the best available data on the magnitude of diarrhoeal mortality, and are useful for comparison to the estimates published by the Global Burden of Disease Study (World Bank 1993, Murray & Lopez 1996). Review of past attempts to quantify disease burden Morbidity from diarrhoea disproportionately affects children younger than 5 years, and mortality from diarrhoea is overwhelmingly a problem of infants and young children in developing countries. Thus, most attempts to quantify disease burden from diarrhoea have focused on these age groups, and on developing countries. Three methods have been used to estimate global disease burden (Table 1.2). Snyder & Merson (1982) developed estimates based primarily on longitudinal studies of children for morbidity, and included vital registration data for mortality estimates. These estimates were recently updated, using more recent longitudinal studies to estimate morbidity and including studies employing a wider range of methodologies, including vital registration and cross-sectional surveys, to estimate mortality (Bern et al. 1992b). For each of these studies, the empirical database was reviewed, and a median diarrhoeal incidence and mortality rate calculated for each region for which data were available, and for all developing countries excluding China. This Table 1.2 Previous estimates of diarrhoeal disease burden for children younger than 5 years in developing countries Deaths per year Deaths per Episodes per (millions) 1000 per year child per year Authors (reference) Year Snyder & Merson (1982) Bern et al. (1992b) 1982 4.6 13.6 2.2 1992 3.3 7.6 2.6 Institute of Medicine 1986 (1986) Martines et al. 1990 (1993) 3.5 7.0 3.5 3.2 6.5 3.5 Methods Longitudinal cohorts and active surveillance Longitudinal cohorts and active surveillance Estimated incidence plus case fatality rate Data from surveys using WHO methods Diarrhoeal Diseases 5 method requires extrapolation from available longitudinal studies, which may not be entirely representative of the country or region. The Institute of Medicine (1986), for the purpose of establishing priorities for vaccine development, published estimates of morbidity and mortality derived by a committee of expert researchers in the field. Based on a review of the literature and personal field experience, the committee made estimates of incidence by region and age group, which were combined with estimates of distribution of diarrhoeal episodes by severity (mild, moderate, severe, death), to yield figures for each of four regions of the developing world. This approach depends on the accuracy of estimates by a small group of experts, and is not derived directly from individual studies. The third approach to estimating disease burden (Martines et al. 1993) is based on data from 276 surveys of diarrhoeal morbidity and mortality in 60 countries conducted between 1981 and 1986. From these data, a median and range were calculated for each region. The methodology for these surveys involved two-stage cluster sampling, and a standardized questionnaire to ascertain two-week prevalence and deaths from diarrhoea among children younger than 5 years (World Health Organization 1989) (Table 1.3). These studies represent the source of data with the broadest available coverage. The reliability of some national survey results have, however, been questioned, in part because of the inadequate training or supervision of field staff. In addition, the method used to estimate mortality, a one-year recall by the household head, is thought to result in underestimates (Martines et al. 1993). Thus, estimates made by these methods include an inherent uncertainty. Table 1.3 Diarrhoeal morbidity and mortality in 276 surveys in children younger than 5 years, using WHO methodology, 1981–1986 Region Latin America and Caribbean Sub-Saharan Africa Middle East and North Africa Asia and the Pacific India China Other All regions Source: Martines 1993 Median Range Diarrhoeal deaths as a percentage of total (median) 12 8 4.9 0.8–10.4 4.2 1.2–9.20 35 67 47 22 10 4.4 2.7 1.6–9.9 2.1–10.8 10.60 5.8 3.1–54.9 1.0–25.3 38 39 150 20 2.6 1.1–5.7 3.2 0.0–17.2 29 — — — 1 1 18 2.7 1.2 2.6 — — — 3.2 0.0 3.3 — — — — — — 276 60 3.5 0.8–10.8 6.5 0.0–54.9 36 Number of surveys Number Episodes/child/year of countries Median Range Deaths per 1000 children per year 6 Global Epidemiology of Infectious Diseases Review of empirical databases Children in developing countries Because community-based studies of morbidity require intensive surveillance, the number of persons studied is generally insufficient to allow mortality measurements. Thus, the groups of empirical studies which examine morbidity (Table 1.4) and mortality (Table 1.5) are different. With two exceptions for which surveillance was monthly (Yang et al. 1990, Chen et al. 1991), the diarrhoeal morbidity studies presented in these tables are all longitudinal, community-based studies of children in developing countries, in which follow-up was at least one year, and surveillance occurred at least every two weeks. Mortality data from studies employing active surveillance or crosssectional ascertainment are sparse, especially for Latin America, but vital registration data support the observation that diarrhoeal mortality rates have declined substantially in Latin America (World Health Organization 1982, Pan American Health Organization 1991) (Table 1.6). Many of the more recent studies presented here for Africa and Asia were studies of interventions for acute respiratory disease (Bang et al. 1990, Pandey et al. 1989, Roesin et al. 1990, de Francisco et al. 1993, Khan et al. 1990) or vitamin A deficiency (Daulaire et al. 1992, West et al. 1991, Herrera et al. 1992), which included longitudinal surveillance for cause-specific mortality. For these studies, as well as for diarrhoeal case management intervention studies, the rates shown are for the nonintervention group or area. Older age groups The Institute of Medicine (1986) has estimated that 60 per cent of overall morbidity and 90 per cent of mortality occur among children younger than 5 years. The few studies that have directly measured incidence of diarrhoea demonstrate that older children and adults experience about 0.1 to 0.5 episodes per year, and that the incidence rises to approximately 1 episode per year among the elderly. The results of some of these studies are seen in Table 1.7. Mortality rates for older children and adults are substantially lower than for children younger than 5 years (Table 1.8). Morbidity and mortality by type of diarrhoea Recently, persistent diarrhoea has been identified as the cause of a substantial proportion of diarrhoeal mortality (Table 1.9), although the percentage differs by location, with a range from 23 to 70 per cent. Differences in prevalent pathogens are likely to contribute to the variation in morbidity and mortality by type of diarrhoea. In addition, Victora et al. (1993) suggest that the differences in mortality may in part be attributable to differential access to oral rehydration therapy. When properly applied, oral rehydration therapy is highly effective in treating acute watery diarrhoea (Duggan, 1985–86 early 1980s early 1980s 1993 Urban Calcutta (Sircar et al. 1984) Rural Andhra Pradesh (Mathur et al. 1985) Rural Northern India (Kumar, Kumar & Datta 1987) Urban Uttar Pradesh (Bhandari, Bhan & Sazawal 1994) 1986-87 Fujian (Chen et al. 1991) 1984–87 1978–79 1978–79 Bangladesh (Huttly et al. 1989) Bangladesh (Black et al. 1982a, 1982b, Black, Brown & Becker 1984a) Bangladesh (Chen et al. 1981) Other Asia and Islands 1986–87 Hebei (Yang et al. 1990) China 1985–86 Period 1.3 6–11 4.4 0.8 0.6 2 0.5 0.6 3 0.4 0.4 4 2.4 2.0 1.7 6.0 5.0 3.8 5.5 4.3 2.3 4.5 1.6 4.5 1.7 —9.9— —9.9 9.9— 2.7 —1.6— 1.6 1.6— 2.0 0.8 1 Age group (years) Episodes of diarrhoea per person per year ——4.1 —— —4.1—— — —7.0— 7.0 7.0— 2.4 — — —4.2 —1.8— 3.1 —1.7— 1.7 1.7— 1.6 0–5 Age group (months) Review of empirical databases on morbidity from diarrhoea: longitudinal studies measuring incidence Rural Uttar Pradesh (Bhan et al. 1989) India Location Table 1.4 4.1 5.6 2.3 2.5 2.2 1.6 1.1 0.7 0–4 continued 0.73 All ages Diarrhoeal Diseases 7 1982–85 Ghana (Biritwum et al. 1986) 1984–86 1982–84 1982–85 1985–87 1978–79 1978–80 Peru (Lanata et al. 1989) Peru (Lopez de Romana et al. 1989, Black et al. 1989) Mexico (Cravioto et al. 1988) Mexico (Cravioto et al. 1990) Brazil (Giugliano et al. 1986) Brazil (Guerrant et al. 1983) Latin America and the Caribbean Egypt (El Alamy et al. 1986) 1982–84 1987–88 Zaire (Manun’ebo et al. 1994) 5.9 3.3 5.0 4.0 10.3 7.6 4.5 0.5 3.2 9.6 7.3 1.5 1.8 —4.0— —4.0 4.0— —3.0— 3.0 3.0— 9.3 1.6 3.8 2 0.7 3.0 3 7.7 6.4 1.3 1.9 3.0 0.6 2.8 4 — a— —5.3 — b— —4.0 —1.2c— 1.1 — — —2.1 2.4 5.0 1.7 1 Age group (years) Episodes of diarrhoea per person per year —10.6— 10.6 10.6— 5.3 2.7 2.7 1982–86 Nigeria (Huttly et al. 1990, Blum et al. 1990) Middle Eastern Crescent 6–11 —7.3— 2.2 1981–84 0–5 Nigeria (Oyejide & Fagbami 1988a, 1988b) Gambia (Rowland et al. 1985) Period Age group (months) 6.4 4.9 1.2 3.1 1.3 6.3 3.7 0–4 Review of empirical databases on morbidity from diarrhoea: longitudinal studies measuring incidence (continued) Sub-Saharan Africa Location Table 1.4 1.0 All ages 8 Global Epidemiology of Infectious Diseases Diarrhoeal Diseases 9 Santosham & Glass 1992), but has limited impact on persistent or dysenteric diarrhoea. 3.7 3.9 4.0 0.8b 0.4c 4.0 3.0 1.5 2.1 5.6 2.0 c. More affluent urban children. b. Poor urban children. a. Poor rural children. 1983–86 Argentina (Grinstein et al. 1989) 1983 Chile (Araya et al. 1986) 0.6 0.8 1986–89 Chile (Ferreccio et al. 1991) 0.7 1981–84 Costa Rica (Simhon et al. 1985) — — —2.3 14.1 1984–86 Brazil (Schorling et al. 1990) 9.4 1982–86 Brazil (Linhares et al. 1989) — — —2.6 15.1 12.2 8.7 1.3 7.2 0.9 1.5 11.3 Developed countries Diarrhoeal disease continues to cause substantial morbidity among children in developed countries, with an estimated incidence of 1.3 to 2.3 episodes per year for children younger than 5 years (Glass et al. 1991), an estimate based on a review of the only four longitudinal studies which provide relevant data (Gurwirth & Williams 1977, Dingle, Badger & Jordan 1964, Hughes et al. 1978, Rodriguez et al. 1987). It is estimated that perhaps 10 per cent of diarrhoeal episodes in the United States entail a physician visit, and 1 per cent a hospital admission (Glass et al. 1991). Lew et al. (1991) examined diarrhoeal mortality using vital registration data from the United States. Mortality rates ranged from < 0.1 per 100 000 for persons 5–24 years, to 2 per 100 000 for children younger than 5 years, and 14 per 100 000 for persons 75 years or older. The only age group in the United States for which the diarrhoeal mortality rate increased from 1978 to 1987 were males 25–54 years old, and this increase appeared to be associated with acquired immunodeficiency syndrome (AIDS). These estimates were based on finding one of the ICD-9 codes for diarrhoea in the underlying cause (45 per cent) or in one of the top three positions (55 per cent) in the death certificate. However, 51 per cent of childhood deaths and 86 per cent of those among the elderly were coded using the ICD-9 code 558 (diarrhoea, presumed non-infectious), despite indirect evidence based on the seasonality of deaths that a substantial proportion were attributable to an infectious agent. This suggests that even good vital registration data may underestimate the contribution of infectious diarrhoea to mortality, because in most cases no aetiological agent is identified before death. 1989 Tamil Nadu (Rahmathullah et al. 1990) 1983 1986–87 Kediri, Indonesia (Roesin et al. 1990) Katmandu Valley, Nepal (Pandey et al. 1989) Sumatra, Indonesia (Nazir, Pardede & Ismail 1985) 1989 1984–85 Jumla, Nepal (Daulaire et al. 1992) 1990 1978–87 Matlab, Bangladesh (Shaikh et al. 1990) Terai, Nepal (West et al. 1991) 1975–77 Matlab, Bangladesh (Chen, Rahman & Sarder 1980) Other Asia and Islands 1987 1982–84 Maharashtra (Bang et al. 1990) 1984–85 Haryana (Bhandari, Bhan & Sazawal 1992) Period 8 624 974 12 264 1 019 3 411 28 000 42 000 7 655 6 176 1 467 5 350 Type of study Active surveillance National survey Active surveillance Active surveillance Active surveillance Active surveillance Active surveillance Active surveillance Active surveillance Active surveillance Active surveillance Study characteristics Population 18.8 70.0 5.7 19.6 9.2 <1 9.2 12.0 10.4 15.1 1–4 9.9 11.3 4.6 23.0 97.5 9.5 16.0 4.3 8.4 16.0 0–4 Diarrhoeal deaths per 1000 by age group (years) 23 22 28 39 77 27 41 20 Diarrhoeal deaths as percentage of total 0–4 year mortality Review of empirical databases on mortality from diarrhoea: studies measuring annual deaths per 1000 children Eight states (Tandon et al. 1987) India Location Table 1.5 10 Global Epidemiology of Infectious Diseases 1988–90 1987–90 1988–89 Sudan (Herrera et al. 1992) Guinea-Bissau (Molbak et al. 1992) Gambia (de Francisco et al. 1993) 10 739 1986 1984 12 156 1980 Egypt (National Control of Diarrheal Diseases Project 1988) Egypt (El-Rafie et al. 1990) 10 418 1982–84 2 071 25 000 1 426 14 149 1 928 5 067 29 000 (total population) 1 064 3 146 9 915 6 584 Yemen (Bagenholm and Nasher 1989) Middle Eastern Crescent 1987 1986–88 Ethiopia (Shamebo et al. 1991) Nigeria (Jinadu et al. 1991) 1975–78 Kenya (Omondi-Odhiambo, van Ginaeken & Voorhoeve 1990) 1984 1982–83 1984–86 Tanzania (Mtango & Neuvians 1986) Gambia (Greenwood et al. 1990) 1983 Central African Republic (Georges et al. 1987) Sub-Saharan Africa Nationwide cluster surveys Local registration 6-monthly ascertainment Active surveillance Active surveillance 6-monthly surveys Retrospective survey (1 year recall) Active surveillance Active surveillance Active surveillance Active surveillance One-time cluster survey 0.8 (1–2 years) 4.5 3.1 0.9 —10.9— (0–2 years) 27.7 11.0 7.6 11.2 4.6 15.3 5.8 62 71 85 21 17 34 44 3.5a 18.9 33 9 20 19 31 continued 14.4 19 20.0 4.0 3.0 11.4 15.9 6.8 5.6 Diarrhoeal Diseases 11 1986 Pakistan (Khan et al. 1990) a. 9–72 months. Overall mortality unexpectedly low. 1985 Ceará, Brazil (Bailey et al. 1990) 1979–89 Ecuador (Ministry of Health, Ecuador & Centers for Disease Control and Prevention 1992) 1983 1988–93 El Salvador (Salvadoran Demographic Association & Centers for Disease Control and Prevention 1994) Pelotas, Brazil (Barros et al. 1987, Victora et al. 1987) 1982–93 Nicaragua (Profamilia & Centers for Disease Control 1993) Latin America and the Caribbean 1980–81 1 677 5 914 Children of 7 961 women Children of 5 752 women Children of 7 150 women 1 194 2 556 8 704 1986 Type of study Birth cohort with 6-monthly ascertainment Birth cohort with one-time ascertainment National demographic survey (retrospective ascertainment) National demographic survey (retrospective ascertainment) National demographic survey (retrospective ascertainment) Active surveillance Family cohort study with active surveillance Study characteristics Population Period 1–4 21.8 4.5 7.0 7.0 12.0 27.5 5.0 3.0 15.0 4.4 —5.5— (0–2 years) <1 5.4 3.8 14.4 9.0 11.2 0–4 Diarrhoeal deaths per 1000 by age group (years) 23 20 31 22 50 Diarrhoeal deaths as percentage of total 0–4 year mortality Review of empirical databases on mortality from diarrhoea: studies measuring annual deaths per 1000 children (continued) Egypt (El Alamy et al. 1986) Location Table 1.5 12 Global Epidemiology of Infectious Diseases Source: Pan American Health Organization 1991 0.98 2.86 4.82 3.98 1.16 0.37 1.19 4.73 7.54 — 9.35 8.57 1.99 4.25 9.83 1.45 4.50 7.80 1.03 2.27 Deaths per 1000 children 0–4 years 9.8 26.8 26.6 23.6 13.7 7.1 10.4 22.8 29.4 — 28.1 38.4 26.1 28.3 44.6 14.3 32.3 24.6 10.0 20.1 Diarrhoeal deaths as % of total for children 0–4 years 1975–1979 0.50 1.11 3.02 1.56 0.37 0.24 0.31 3.01 4.85 4.09 7.46 6.56 1.32 2.69 — 0.70 2.95 5.50 0.61 1.55 Deaths per 1000 children 0–4 years 6.7 13.5 20.7 16.7 7.3 5.4 5.2 17.4 25.7 21.6 27.3 36.1 28.5 25.5 — 9.5 24.4 19.9 8.4 17.2 Diarrhoeal deaths as % of total for children 0–4 years 1980–1984 0.31 0.86 1.94 — 0.27 0.18 0.17 — 3.53 — — — — 2.17 — 0.66 2.35 — 0.31 1.08 Deaths per 1000 children 0–4 years 4.5 12.9 17.2 — 6.1 4.9 3.4 — 23.6 — — — — 24.7 — 9.4 23.0 — 5.2 13.6 Diarrhoeal deaths as % of total for children 0–4 years 1985–90 0.5 3.8 3.8 — 1.5 0.6 0.8 — 8.7 — — — — 7.7 — 2.8 7.0 — 0.6 3.5 Diarrhoeal deaths as % of total for all ages Review of databases on mortality from diarrhoea: estimated from vital registration data in Latin America and the Caribbean Argentina Belize Brazil Colombia Costa Rica Cuba Chile Dominican Republic Ecuador El Salvador Guatemala Honduras Jamaica Mexico Nicaragua Panama Paraguay Peru Uruguay Venezuela Country Table 1.6 Diarrhoeal Diseases 13 14 Table 1.7 Global Epidemiology of Infectious Diseases Review of empirical databases on morbidity from diarrhoea: morbidity in various age groups Age group (years) and morbidity China (Chen et al. 1991) Episodes per person per year 0–4 2.3 Indonesia (El Alamy et al. 1986) 2 week prevalence (%) 0–4 35% 5–12 2% >12 2% Egypt (Nazir, Pardede & Ismail 1985) Episodes per person per year 0–4 3.1 5–14 0.3 >14 0.1 United States (Rodriguez et al. 1985) Episodes per person per year 0–4 1.0 5–14 0.4 >14 0.4 all ages 0.73 Disease Burden Morbidity Table 1.10 highlights the burden of diarrhoeal disease globally: 4 billion episodes of diarrhoea were estimated to occur each year (Murray & Lopez 1996). More than 90 per cent of the morbidity occurs in developing countries. Based on our review of empirical data, it seems likely that some of the regional figures in Table 1.10 are underestimates of the true scale of morbidity. For children in the poorest socioeconomic strata, it is likely that incidence rates are similar in India, Latin America and Africa, and that complete ascertainment of all episodes of diarrhoea would yield rates in the order of 8–14 episodes per year for children between 6 and 24 months, and 5–8 episodes per year for all children younger than 5 years. The overall incidence for a region would be dependent on its socioeconomic mix. It is therefore probable that morbidity is actually somewhat higher than presented in Table 1.10 for Africa, India, the Middle Eastern Crescent and Other Asia and Islands. In Africa, especially, the burden of diarrhoeal disease for adults is undoubtedly increasing because of the magnitude of the AIDS epidemic there. Mortality Table 1.11 presents summary estimates for mortality attributable to diarrhoea (Murray & Lopez 1996). Nearly 3 million people were estimated to die each year from diarrhoeal diseases. More than 99 per cent of the deaths occurred in developing countries and 84 per cent of all diarrhoeal deaths were among children under 5 years in developing countries. The highest diarrhoeal mortality rates on a regional basis were found in sub-Saharan Africa, followed by India, the Middle Eastern Crescent, and Other Asia and Islands. Latin America achieved an impressive decrease in diarrhoeal mortality rates over the past 20 years, a finding corroborated by both empirical and vital registration data. 0–4 25% 0–4 11.2 0–4 0.02 Kenya (Bradley & Gilles 1984) Percentage of total mortality Egypt (El Alamy et al. 1986) Deaths per 1000 per year United States (Lew et al. 1991) Deaths per 1000 per year Bangladesh (Fauveau et al. 1989) (Women only) Deaths per 1000 per year (Percentage of total mortality) 0–4 5–14 1.3 5–24 < 0.001 >9 1.7 (20%) 5–14 28% 5–9 1.2 (27%) >14 0 >14 21% 15–24 0.12 (5%) 35–44 0.5 (14%) 25–54 0.003 25–34 0.4 (13%) Age group (years) and mortality Review of empirical databases on mortality from diarrhoea: various age groups Bangladesh (Shaikh et al. 1990) Deaths per 1000 per year (Percentage of total mortality) Table 1.8 55–74 0.02 > 74 0.14 all ages 0.014 Diarrhoeal Diseases 15 Type of study Case-control study of deaths Longitudinal N = 705 Same cohort as Baqui et al. 1992b Retrospective N = 363 Population-based surveillance: women 15–44 years Population-based 236 deaths Population-based study of 227 infant deaths Longitudinal N = 175 Retrospective study of deaths Population-based 227 infant deaths Longitudinal N = 1426 Bangladesh (Fauveau et al. 1992) Bangladesh (Baqui et al. 1992b) Bangladesh (Baqui et al. 1993) Bangladesh (Henry et al. 1992) Bangladesh (Fauveau et al. 1989) Bangladesh (Victora et al. 1993) Brazil (Victora et al. 1992) Brazil (Schorling et al. 1990) Brazil (Lima et al. 1992) Brazil (Victora et al. 1993) Guinea Bissau (Molbak et al. 1992) 1.5 (age 0–4 years) 0.8 (age 6–11 months) 0.34 (age 0–4 years) 62 58 42b 70 62 23 59 49 Persistent diarrhoea 28 34 24 Persistent diarrhoea Acute watery (episodes per year) diarrhoea 10 42 17 Dysentery Percentage of diarrhoeal deaths attributable to: 135 diarrhoea days per year for those with and 22 for those without persistent diarrhoea 99 per cent of children seen in health care facility during fatal illness 10 per cent of mortality atrributable to diarrhoea; persistent diarrhoea may be important in adults Risk factors for persistent diarrhoea: blood or mucus in stool Malnutrition and immune deficiency independent risk factors for persistent diarrhoea Risk factors for persistent diarrhoea: blood in stool, dehydration Peak mortality for persistent diarrhoea with malnutrition 24–35 months Other findings Studies of persistent diarrhoea and distribution of mortality by type of diarrhoea Location (references) Table 1.9 16 Global Epidemiology of Infectious Diseases Diarrhoeal Diseases 17 8 47 46 Population-based 146 deaths Longitudinal N = 677 Population-based India (Victora et al. 1993) Peru (Lanata et al. 1991) Senegal (Victora et al. 1993) a. Case-fatality rate. b. Includes dysentery lasting < 14 days. Longitudinal N = 963 India (Bhan et al. 1989) 531 deaths Longitudinal N = 1467 India (Bhan et al. 1986) 0.31 (age < 1 year) 0.25 (age 0–3 years) 35 CFRa 0.7 51 CFRa 14 14 Risk factors for persistent diarrhoea: age < 6 months, ≥ 6 stools/day Risk factors for persistent diarrhoea: blood in stool, age 3–5 months Risk factors for mortality: severe malnutrition, persistent diarrhoea Disability calculations In longitudinal studies of diarrhoeal morbidity, the median duration of watery diarrhoea has been observed to be 4 to 5 days, while dysentery lasts longer, 6 to 11 days (Baqui et al. 1991, Black et al. 1982a, Lopez de Romana et al. 1989, Black et al. 1989, Baqui et al. 1992b). Persistent diarrhoea, by definition, lasts at least 14 days (World Health Organization 1988). For the purpose of calculating disability, the distribution of episodes by category (watery, persistent or dysenteric) was assumed to be constant for developing countries and was estimated by WHO based on an average derived from published sources (Bhan et al. 1989, Victora et al. 1993, Schorling et al. 1990, Bhan et al. 1986, Baqui et al. 1992b, Lanata et al. 1991, Henry 1991). Disability weights were assigned according to type of diarrhoea (Tables 1.12 and 1.13). As discussed above, the distribution of types of diarrhoea varies widely in the developing world. In addition, persistent diarrhoea may be a more important cause of mortality among adults than is reflected in the World Development Report disability estimates. Active surveillance data from Bangladesh suggest that diarrhoea may cause up to 10 per cent of mortality among women of child-bearing age, more than half of that mortality being attributable to persistent diarrhoea (Fauveau et al. 1989). Diarrhoea, both persistent and acute, is associated with malnutrition. In particular, severe malnutrition is a significant risk factor for mortality from acute diarrhoea (Bern et al. 1992a) and, more strikingly, from persistent diarrhoea (Bhan et al. 1986, Baqui et al. 1993). Pre-existing malnutrition has also been shown to lead to increased duration (Black, Brown & Becker 1984b) and severity (Tomkins 18 Table 1.10 Global Epidemiology of Infectious Diseases Regional summary estimates for morbidity from diarrhoeal disease Diarrhoeal episodes (millions) Region All ages Established Market Economies Former Socialist Economies of Europe India China Other Asia and Islands Sub-Saharan Africa Latin America and the Caribbean Middle Eastern Crescent World < 5 years Episodes of diarrhoea per person per year All ages < 5 years 167.2 92.6 0.21 1.8 93.8 61.9 0.27 2.3 787.9 1010.3 496.7 653.1 434.5 524.1 318.2 301.0 444.4 225.6 0.93 0.89 0.73 1.28 0.98 4.5 2.3 4.0 5.0 4.0 430.3 307.7 0.86 4.0 4073.9 2275.4 0.77 3.6 Source: Murray & Lopez 1996. Table 1.11 Regional summary estimates for mortality from diarrhoeal disease Diarrhoeal deaths (thousands) Region Established Market Economies Former Socialist Economies of Europe India China Other Asia and Islands Sub-Saharan Africa Latin America and the Caribbean Middle Eastern Crescent World All ages < 5 years Mortality rate from diarrhoea (deaths per 1000) All ages < 5 years Diarrhoeal deaths as percentage of total mortality All ages < 5 years <0.1 0.4 3 <1 0.003 0.007 4 4 0.01 0.1 0.1 3.5 922 93 397 950 153 733 50 352 809 130 1.1 0.1 0.6 1.9 0.4 6.3 0.4 4.1 8.6 2.3 9.8 1.0 7.2 11.6 5.1 22.6 4.7 21.8 20.1 18.4 424 402 0.8 5.0 9.3 21.6 2946 2480 0.6 3.9 5.8 19.4 Source: Murray & Lopez 1996. 1981) of diarrhoea. Diarrhoeal disease can affect growth in the short term (Henry et al. 1987) and long term (Black, Brown & Becker 1984a, Guerrant et al. 1983). This may depend on the type of diarrhoea, with dysentery having a more marked effect on linear growth (Black, Brown & Becker 1984a). It has been shown that sustained nutritional supplementation can obviate the nutritional consequences of diarrhoea in young children (Lutter et al. 1989), but that without supplements, there may be lifelong consequences for growth (Rivera & Martorell 1988). Diarrhoeal Diseases Table 1.12 19 Estimates for distribution of episodes by type of diarrhoea according to age Percentage of episodes attributable to Age (years) Watery diarrhoea Persistent diarrhoea 80 89 90 90 85 10 1 0 0 0 0–4 5–14 15–44 45–59 ≥ 60 Percentage of mortality attributable to Dysentery Watery diarrhoea Persistent diarrhoea Dysentery 10 10 10 10 15 50 75 80 80 85 35 5 0 0 0 15 20 20 20 15 Source: WHO. Table 1.13 Assignment of disability weights according to type of diarrhoea; classification used for calculation of disabilityadjusted life years Percentage assigned to severity classa Type of diarrhoea Acute watery Persistent Dysentery I II III IV 70 50 20 40 40 10 10 40 20 a. Severity classes based on limited ability to perform activities in the following areas: recreation, education, procreation or occupation. Class I denotes limited ability to perform at least one activity in one of these areas. Class II denotes limited ability to perform most activities in one area. Class III denotes limited ability to perform most activities in two or more areas. Class IV denotes limited ability to perform most activities in all areas. Conclusion Oral rehydration therapy (ORT) is the major intervention that has been promulgated globally for control of morbidity and mortality from diarrhoeal disease. ORT is not a primary intervention, in that it does not affect incidence of diarrhoeal disease, but it has been shown to be effective in reducing hospitalizations and mortality from diarrhoea in intensive studies (Oberle et al. 1980, Rahaman et al. 1979, Heymann et al. 1990). It has been more difficult to demonstrate an effect of ORT diarrhoeal control programmes on a larger scale (Santosham & Greenough 1991). Data from the Egyptian national programme have demonstrated a decrease in mortality over the 1970s and 1980s (National Control of Diarrheal Diseases Project 1988), but the decrease in mortality had begun before the programme was in place, suggesting that other changes, such as socioeconomic development with attendant improvements in nutrition, sanitation and water, may have been at least partly responsible. WHO estimates suggest that increased global access to ORT coupled with methods to improve usage have the potential to decrease global mortality from diarrhoea (World Health Organization 1992). The extent to which this will be effective will depend upon the proportion of mortality resulting from acute watery diarrhoea in the locality in question (Victora 20 Global Epidemiology of Infectious Diseases et al. 1993). Recent data on diarrhoeal mortality attributable to persistent diarrhoea and dysentery suggest that once ORT use is high, further reduction of diarrhoeal mortality will depend upon additional interventions such as improved case management, hygiene education, encouragement of breast-feeding, improved sanitation, and general socioeconomic development (Feachem, Hogan & Merson 1983, Esrey, Feachem & Hughes 1985, Feachem 1984, Ashworth & Feachem 1985, Feachem & Koblinsky 1984, Ronsmans et al. 1991), all of which are more difficult to implement. Thus, in areas where diarrhoeal mortality is very high and ORT access is low, such as sub-Saharan Africa, a large fraction of deaths may be preventable through improved ORT access and education, while areas such as Brazil may already be at the point where further reduction will require other interventions in addition to the maintenance of ORT promotion. References Araya M et al. (1986) Acute diarrhoea and asymptomatic infection in Chilean preschoolers of low and high socio-economic strata. Acta Paediatrica Scandinavica, 75:645–651. Ashworth A, Feachem RG (1985) Interventions for the control of diarrhoeal diseases among young children: weaning education. Bulletin of the World Health Organization, 63:1115–1127. Bagenholm GC, Nasher AAA (1989) Mortality among children in rural areas of the People’s Democratic Republic of Yemen. Annals of Tropical Paediatrics, 9:75–81. Bailey P et al. (1990) A study of infant mortality and causes of death in a rural north-east Brazilian community. Journal of Biosocial Science, 22:349–363. Bang AT et al. (1990) Reduction in pneumonia mortality and total childhood mortality by means of community-based intervention trial in Gadchiroli, India. Lancet, 336:201–206. Baqui AH et al. (1991) Methodological issues in diarrhoeal diseases epidemiology: definition of diarrhoeal episodes. International Journal of Epidemiology, 20:1057–1063. Baqui AH et al. (1992a) Enteropathogens associated with acute and persistent diarrhea in Bangladeshi children less than 5 years of age. Journal of Infectious Diseases, 166:792–796. Baqui AH et al. (1992b) Epidemiological and clinical characteristics of acute and persistent diarrhoea in rural Bangladeshi children. Acta Paediatrica Supplements, 381:15–21. Baqui AH et al. (1993) Cell-mediated immune deficiency and malnutrition are independent risk factors for persistent diarrhea in Bangladeshi children. American Journal of Clinical Nutrition, 58:543–548. Barros FC et al. (1987) Infant mortality in southern Brazil: a population based study of causes of death. Archives of Disease in Childhood, 62:487–490. Diarrhoeal Diseases 21 Bennish ML (1991) Potentially lethal complications of shigellosis. Reviews of Infectious Diseases, 13(Suppl 4):S319–324. Bern C, Glass RI (1994) Impact of diarrheal diseases worldwide. In: Kapikian AZ, ed. Viral Infections of the Gastrointestinal Tract. New York, Marcel Dekker, pp. 1–26. Bern C et al. (1992a) Rotavirus diarrhea in Bangladeshi children: correlation of disease severity with serotypes. Journal of Clinical Microbiology, 30:3234–3238. Bern C et al. (1992b) The magnitude of the global problem of diarrhoeal disease: a ten-year update. Bulletin of the World Health Organization, 70:705–714. Bhan MK et al. (1986) Major factors in diarrhoea related mortality among rural children. Indian Journal of Medical Research, 83:9–12. Bhan MK et al. (1989) Descriptive epidemiology of persistent diarrhoea among young children in rural northern India. Bulletin of the World Health Organization, 67:281–288. Bhandari N, Bhan MK, Sazawal S (1992) Mortality associated with acute watery diarrhea, dysentery, and persistent diarrhea in rural north India. Acta Paediatrica Supplements, 381:3–6. Bhandari N, Bhan MK, Sazawal S (1994) Impact of massive dose of vitamin A given to preschool children with acute diarrhoea on subsequent respiratory and diarrhoeal morbidity. British Medical Journal, 309:1404–1407. Biritwum RB et al. (1986) Growth and diarrhoeal disease surveillance in a rural Ghanaian pre-school child population. Transactions of the Royal Society of Tropical Medicine and Hygiene, 80:208–213. Black RE et al. (1982a) Longitudinal studies of infectious diseases and physical growth of children in rural Bangladesh. I. Patterns of morbidity. American Journal of Epidemiology, 115:305–314. Black RE et al. (1982b) Longitudinal studies of infectious diseases and physical growth of children in rural Bangladesh. II. Incidence of diarrhea and association with known pathogens. American Journal of Epidemiology, 115:315–324. Black RE et al. (1989) Incidence and etiology of infantile diarrhea and major routes of transmission in Huascar, Peru. American Journal of Epidemiology, 129:785–799. Black RE, Brown KH, Becker S (1984a) Effects of diarrhea associated with specific enteropathogens on the growth of children in rural Bangladesh. Pediatrics, 73: 799–805. Black RE, Brown KH, Becker S (1984b) Malnutrition is a determining factor in diarrheal duration, but not incidence, among young children in a longitudinal study in rural Bangladesh. American Journal of Clinical Nutrition, 39:87–94. Blum D et al. (1990) The Imo State (Nigeria) Drinking Water Supply and Sanitation Project1. Description of the project, evaluation methods, and impact on intervening variables. Transactions of the Royal Society of Tropical Medicine and Hygiene, 84:309–315. Bradley AK, Gilles HM (1984) Malumfashi Endemic Diseases Research Project, XXI. Pointers to causes of death in the Malumfashi area, northern Nigeria. Annals of Tropical Medicine and Parasitology, 78:265–271. 22 Global Epidemiology of Infectious Diseases Chen KC et al. (1991)The epidemiology of diarrhoeal diseases in southeastern China. Journal of Diarrhoeal Diseases Research, 9:94–99. Chen LC, Huq E, Huffman SL (1981) A prospective study of the risk of diarrheal diseases according to the nutritional status of children. American Journal of Epidemiology, 114:284–292. Chen LC, Rahman M, Sarder AM (1980) Epidemiology and causes of death among children in a rural area of Bangladesh. International Journal of Epidemiology, 9:25–33. Cravioto A et al. (1988) Prospective study of diarrhoeal disease in a cohort of rural Mexican children: incidence and isolated pathogens during the first two years of life. Epidemiology and Infection, 101:123–134. Cravioto A et al. (1990) Risk of diarrhea during the first year of life associated with initial and subsequent colonization by specific enteropathogens. American Journal of Epidemiology, 131:886–904. Daulaire NMP et al. (1992) Childhood mortality after a high dose of vitamin A in a high risk population. British Medical Journal, 304:207–210. de Francisco A et al. (1993) The pattern of infant and childhood mortality in Upper River Division, The Gambia. Annals of Tropical Paediatrics, 13:345–352. Dingle JH, Badger DG, Jordan WS (1964) Illness in the home: a study of 25,000 illnesses in a group of Cleveland families. Cleveland, Case Western Reserve Press. Duggan C, Santosham M, Glass RI (1992) The management of acute diarrhea in children: oral rehydration, maintenance, and nutritional therapy. Morbidity and Mortality Weekly Report, 41:1–20. El Alamy MA et al. (1986) The incidence of diarrheal disease in a defined population of rural Egypt. American Journal of Tropical Medicine and Hygiene, 35: 1006–1012. El-Rafie M et al. (1990) Effect of diarrhoeal disease control on infant and childhood mortality in Egypt. Lancet, 335:334–338. Esrey SA, Feachem RG, Hughes JM (1985) Interventions for the control of diarrhoeal diseases among young children: improving water supplies and excreta disposal facilities. Bulletin of the World Health Organization, 63:757–772. Fauveau V et al. (1989) Epidemiology and cause of deaths among women in rural Bangladesh. International Journal of Epidemiology, 18:139–145. Fauveau V et al. (1992) Persistent diarrhoea as a cause of childhood mortality in rural Bangladesh. Acta Paediatrica Supplements, 381:12–14. Feachem RG (1984) Interventions for the control of diarrhoeal diseases among young children: promotion of personal and domestic hygiene. Bulletin of the World Health Organization, 62:467–476. Feachem RG, Koblinsky MA (1984) Interventions for the control of diarrhoeal diseases among young children: promotion of breastfeeding. Bulletin of the World Health Organization, 62:271–291. Feachem RG, Hogan RC, Merson MH (1983) Diarrhoeal disease control: reviews of potential interventions. Bulletin of the World Health Organization, 61:637–640. Diarrhoeal Diseases 23 Ferreccio C et al. (1991) Epidemiologic patterns of acute diarrhea and endemic Shigella infections in children in a poor periurban setting in Santiago, Chile. American Journal of Epidemiology,134:614–627. Georges MC et al. (1987) Diarrheal morbidity and mortality in children in the Central African Republic. American Journal of Tropical Medicine and Hygiene, 36:598–602. Giugliano LG et al. (1986) Longitudinal study of diarrhoeal disease in a periurban commubity in Manaus (Amazon-Brazil). Annals of Tropical Medicine and Parasitology, 80:443–450. Glass RI et al. (1991) Estimates of morbidity and mortality rates for diarrheal diseases in American children. Journal of Pediatrics, 118:S27–33. Goma Epidemiology Group (1995) Public health impact of Rwandan refugee crisis: what happened in Goma, Zaire, in July, 1994? Lancet, 345:339–344. Greenwood BM et al. (1990) Evaluation of a primary health care programme in The Gambia. II. Its impact on mortality and morbidity in young children. Journal of Tropical Medicine and Hygiene, 93:87–97. Grinstein S et al. (1989) Epidemiology of rotavirus infection and gastroenteritis in prospectively monitored Argentine families with young children. American Journal of Epidemiology, 130:300–308. Guerrant RL et al. (1983) Prospective study of diarrheal illnesses in northeastern Brazil: patterns of disease, nutritional impact, etiologies, and risk factors. Journal of Infectious Diseases, 148:986–997. Gurwith MJ, Williams TW (1977) Gastroenteritis in children: a two-year review in Manitoba. I. Etiology. Journal of Infectious Diseases, 136:239–247. Henry FJ (1991) The epidemiologic importance of dysentery in communities. Reviews of Infectious Diseases, 13(Suppl 4):S238–244. Henry FJ et al. (1987) Dysentery, not watery diarrhoea, is associated with stunting in Bangladeshi children. Human Nutrition–Clinical Nutrition, 41:243–249. Henry FJ et al. (1992) Epidemiology of persistent diarrhea and etiologic agents in Mirzapur, Bangladesh. Acta Paediatrica Supplements, 381:27–31. Herrera MG et al. (1992) Vitamin A supplementation and child survival. Lancet, 340:267–271. Heymann DL et al. (1990) Oral rehydration therapy in Malawi: impact on the severity of disease and on hospital admissions, treatment practices, and recurrent costs. Bulletin of the World Health Organization, 68:193–197. Hughes JM et al. (1978) Acute gastrointestinal illness in Charlottesville: a prospective family study [abstract]. Clinical Research, 26:28A. Huilan S et al. (1991) Etiology of acute diarrhoea among children in developing countries: a multicentre study in five countries. Bulletin of the World Health Organization, 69:549–555. Huttly SRA et al. (1989) Persistent diarrhoea in a rural area of Bangladesh: a community based longitudinal study. International Journal of Epidemiology, 18:964–969. 24 Global Epidemiology of Infectious Diseases Huttly SRA et al. (1990) The Imo state (Nigeria) Drinking Water Supply and Sanitation Project, 2. Impact on dracunculiasis, diarrhoea, and nutritional status. Transactions of the Royal Society of Tropical Medicine and Hygiene, 84:316–321. Institute of Medicine (1986) The burden of disease resulting from various diarrheal pathogens. In: New Vaccine Development: Establishing Priorities. Washington, DC, National Academy Press, pp. C1–13. Jinadu MK et al. (1991) Childhood diarrhoea in rural Nigeria. I. Studies on prevalence, mortality and socio-environmental factors. Journal of Diarrhoeal Diseases Research, 9:323–327. Kalter HD et al. (1990) Validation of postmortem interviews to ascertain selected causes of death in children. International Journal of Epidemiology, 19: 380–386. Kapikian AZ, Chanock RM (1990) Rotaviruses. In: Fields BN, Knipe DM, eds. Virology. New York, Raven Press, Ltd., pp. 1353–1403. Khan AJ et al. (1990) Acute respiratory infections in children: a case management intervention in Abbottabad District, Pakistan. Bulletin of the World Health Organization, 68:577–585. Kumar V, Kumar R, Datta N (1987) Oral rehydration therapy in reducing diarrhoea-related mortality in rural India. Journal of Diarrhoeal Diseases Research, 5:159–164. Lanata CF et al. (1989) Protection of Peruvian children against rotavirus diarrhea of specific serotypes by one, two, or three doses of the RIT 4237 attenuated bovine rotavirus vaccine. Journal of Infectious Diseases, 159:452–459. Lanata CF et al. (1991) Epidemiologic, clinical, and laboratory characteristics of acute vs. persistent diarrhea in periurban Lima, Peru. Journal of Pediatric Gastroenterology and Nutrition, 12:82–88. Lanata CF et al. (1992) Etiologic agents in acute vs. persistent diarrhea in children under three years of age in peri-urban Lima, Peru. Acta Paediatrica Supplements, 381:32–38. Lew JF et al. (1991) Diarrheal deaths in the United States, 1979 through 1987: a special problem for the elderly. Journal of the American Medical Association, 265:3280–3284. Lima AAM et al. (1992) Persistent diarrhea in northeast Brazil: etiologies and interactions with malnutrition. Acta Paediatrica Supplements, 381:39–44. Linhares AC et al. (1989) Longitudinal study of rotavirus infections among children from Belém, Brazil. Epidemiology and Infection, 102:129–145. Lopez de Romana G et al. (1989) Longitudinal studies of infectious diseases and physical growth of infants in Huascar, an underprivileged peri-urban community in Lima, Peru. American Journal of Epidemiology, 129:769–784. Lutter CK et al. (1989) Nutritional supplementation: effects on child stunting because of diarrhea. American Journal of Clinical Nutrition, 50:1–8. Manun’ebo MN et al. (1994) Influence of demographic, socioeconomic and environmental variables on childhood diarrhoea in a rural area of Zaire. Journal of Tropical Medicine and Hygiene, 97:31–38. Diarrhoeal Diseases 25 Martines J, Phillips M, Feachem RG (1993) Diarrheal diseases. In: Jamison DT, et al., eds. Disease Control Priorities in Developing Countries. New York, Oxford University Press for the World Bank, pp. 91–116. Mathur R et al. (1985) Nutritional status and diarrhoeal morbidity: a longitudinal study in rural Indian preschool children. Human Nutrition–Clinical Nutrition, 39:447–454. Ministry of Health, Ecuador, and Centers for Disease Control and Prevention (1992) Family Planning and Child Survival Survey, Final report. Atlanta GA. Molbak K et al. (1992) Persistent and acute diarrhoea as the leading causes of child mortality in urban Guinea Bissau. Transactions of the Royal Society of Tropical Medicine and Hygiene, 86:216–220. Mtango FDE, Neuvians D (1986) Acute respiratory infections in children under five years. Control project in Bagamoyo District, Tanzania. Transactions of the Royal Society of Tropical Medicine and Hygiene, 80:851–858. Murray CJL, Lopez AD (1996) Global health statistics: a compendium of incidence, prevalence and mortality estimates for over 200 conditions. Global Burden of Disease and Injury Series, Volume II. Harvard University Press, Boston, 1996. National Control of Diarrheal Diseases Project (1988) Impact of the National Control of Diarrheal Diseases Project on infant and child mortality in Dakahalia, Egypt. Lancet, 2:145–148. Nazir HZ M, Pardede N, Ismail R (1985) The incidence of diarrhoeal diseases and diarrhoeal diseases related mortality in rural swampy low-land area of south Sumatra, Indonesia. Journal of Tropical Pediatrics, 31:268–272. Oberle MW et al. (1980) Diarrhoeal disease in Bangladesh: epidemiology, mortality averted and costs at a rural treatment centre. International Journal of Epidemiology, 9:341–348. Omondi-Odhiambo, van Ginneken JK, Voorhoeve AM (1990) Mortality by cause of death in a rural area of Machakos District, Kenya in 1975–78. Journal of Biosocial Science, 22:63–75. Oyejide CO, Fagbami AH (1988a) An epidemiological study of rotavirus diarrhoea in a cohort of Nigerian infants: I. Methodology and experiences in the recruitment and follow-up of patients. International Journal of Epidemiology, 17:903–907. Oyejide CO, Fagbami AH (1988b) An epidemiological study of rotavirus diarrhoea in a cohort of Nigerian infants: II. Incidence of diarrhoea in the first two years of life. International Journal of Epidemiology, 17:908–912. Pan American Health Organization (1991) Mortality due to intestinal infectious diseases in Latin America and the Caribbean, 1965–1990. Epidemiological Bulletin of the Pan American Health Organization, 12(3):1–10. Pandey MR et al. (1989) Impact of a pilot acute respiratory infection (ARI) control programme in a rural community of the hill region of Nepal. Annals of Tropical Paediatrics, 9:212–220. Profamilia and Centers for Disease Control (1993) Encuesta sobre Salud Familiar, [Family Health Survey] Nicaragua 92–93, Final Report, Managua, Nicaragua. 26 Global Epidemiology of Infectious Diseases Rahaman MM et al. (1979) Diarrhoeal mortality in two Bangladeshi villages with and without community-based oral rehydration therapy. Lancet, II:809–812. Rahmathullah L et al. (1990) Reduced mortality among children in southern India receiving a small weekly dose of Vitamin A. New England Journal of Medicine, 323:929–935. Rivera J, Martorell R (1988) Nutrition, infection and growth. Part I: effects of infection on growth. Clinical Nutrition, 7:156–162. Rodriguez WJ et al. (1985) Fecal adenoviruses from a longitudinal study of families in metropolitan Washington, DC: laboratory, clinical, and epidemiologic observations. Journal of Pediatrics, 107:514–520. Rodriguez WJ et al. (1987) Longitudinal study of rotavirus infection and gastroenteritis in families served by a pediatric medical practice: clinical and epidemiologic observations. Pediatric Infectious Disease Journal, 6:170–176. Roesin R et al. (1990) ARI intervention study in Kediri, Indonesia (a summary of study results). Bulletin of the International Union Against Tuberculosis and Lung Disease, 65:23–30. Ronsmans C et al. (1991) Current practices for treatment of dysentery in rural Bangladesh. Reviews of Infectious Diseases, 13 (Suppl 4):S351–356. Rowland SGJG et al. (1985) The etiology of diarrhea studied in the community in young urban Gambian children. Journal of Diarrhoeal Diseases Research, 3:7–13. Salam MA, Bennish ML (1991) Antimicrobial therapy for shigellosis. Reviews of Infectious Diseases, 13(Suppl 4):S332–341. Salvadoran Demographic Association and Centers for Disease Control and Prevention (1994) National Family Health Survey, Final Report, San Salvador, El Salvador. Santosham M, Greenough WB (1991) Oral rehydration therapy: a global perspective. The Journal of Pediatrics, 118:S44–51. Schorling JB et al. (1990) A prospective study of persistent diarrhea among children in an urban Brazilian slum: patterns of occurrence and etiologic agents. American Journal of Epidemiology, 132:144–156. Shaikh K et al. (1990) Pattern of diarrhoeal deaths during 1966–1987 in a demographic surveillance area in rural Bangladesh. Journal of Diarrhoeal Diseases Research, 8:147–154. Shamebo D et al. (1991) The Butajira rural health project in Ethiopia: mortality pattern of the under fives. Journal of Tropical Pediatrics, 37:254–261. Simhon A et al. (1985) Low endemicity and low pathogenicity of rotaviruses among rural children in Costa Rica. Journal of Infectious Diseases, 152:1134–1142. Sircar BK et al. (1984) A longitudinal study of diarrhoea among children in Calcutta communities. Indian Journal of Medical Research, 80:546–550. Snow RW et al. (1992) Childhood deaths in Africa: uses and limitations of verbal autopsies. Lancet, 340:351–355. Diarrhoeal Diseases 27 Snyder JD, Merson MH (1982) The magnitude of the global problem of acute diarrhoeal disease: a review of active surveillance data. Bulletin of the World Health Organization, 60:605–613. Tandon BN et al. (1987) Morbidity pattern and cause specific mortality during infancy in ICDS projects. Journal of Tropical Pediatrics, 33:190–193. Tomkins A (1981) Nutritional status and severity of diarrhoea among pre-school children in rural Nigeria. Lancet, 1:860–862. Victora CG et al. (1987) Birthweight and infant mortality: a longitudinal study of 5914 Brazilian children. International Journal of Epidemiology, 16:239–245. Victora CG et al. (1992) Deaths due to dysentery, acute and persistent diarrhoea among Brazilian infants. Acta Paediatrica Supplements, 381:7–11. Victora CG et al. (1993) International differences in clinical patterns of diarrhoeal deaths: a comparison of children from Brazil, Senegal, Bangladesh, and India. Journal of Diarrhoeal Diseases Research, 11:25–29. West KP et al. (1991) Efficacy of vitamin A in reducing preschool child mortality in Nepal. Lancet, 338:67–71. World Bank (1993) World development report 1993: investing in health. New York, Oxford University Press. World Health Organization (1982) Diarrhoeal Diseases Control Programme: an overview of the problem in the Americas. Weekly Epidemiological Record, 57:353–355. World Health Organization (1988) Persistent diarrhoea in children in developing countries: memorandum from a WHO meeting. Bulletin of the World Health Organization, 66:709–717. World Health Organization (1989) Household survey manual: diarrhoea case management, morbidity, and mortality. Geneva, World Health Organization (Document CDD/SER/86.2 Rev.1). World Health Organization (1992) Eighth programme report 1990–1991. Geneva, World Health Organization. Yang CR et al. (1990) Diarrhoea surveillance in children aged under 5 years in a rural area of Hebei Province, China. Journal of Diarrhoeal Diseases Research, 8:155–159. Chapter 2 Pertussis Artur M Galazka, Susan E Robertson Introduction “The severity of whooping cough…is much underrated, both in the direct and the indirect damage it causes to the health, growth and resistance of young children…. Deaths are inordinately frequent among infants aged less than one year. As a world problem of preschool children, whooping cough is in the front rank of acute infections…. Despite all this, the disease continues to be lightly regarded by the public, and also by many physicians.” This situation analysis was reported by Gordon & Hood in 1951, but it remains for the most part true today. Over the past three decades, immunization programmes have been implemented successfully in most countries of the world, and morbidity and mortality attributable to pertussis have declined dramatically (Galazka 1992). Since the 1970s, however, six developed countries (Germany, Italy, Japan, Sweden, United Kingdom, and United States of America) have grappled with low levels of pertussis vaccination, leading to substantial problems with this disease in places where it had previously been well controlled. The underlying causes of this situation include apathy and complacency on the part of physicians and parents, negative attitudes toward immunization spread by anti-immunization pressure groups, and litigation over liability for alleged vaccine related injuries (Galazka 1992, Mortimer 1988). In developing countries, coverage with a primary series of three doses of diphtheria-pertussis-tetanus (DPT) vaccine in infants reached 81 per cent in 1995 (Expanded Programme on Immunization 1996). There were, however, considerable differences in coverage rates between WHO regions (58 per cent in Africa compared with 92 per cent in the Western Pacific) and, likewise, between and within countries. Failure to reach and maintain high immunization coverage in developing countries is the result of multiple factors, including weak management of immunization services, 30 Global Epidemiology of Infectious Diseases inadequate resources, missed opportunities to immunize eligible children, and ineffective motivation of mothers to return to complete immunization series for their infants. In developing countries, pertussis has not received the epidemiological attention given to measles or to infections with the other major respiratory viral and bacterial pathogens (Wright 1991). In countries where conditions predispose to early acquisition of disease, where malnutrition and concomitant diseases worsen early infections, and where preventive and curative interventions remain unsatisfactory, the burden from pertussis is still significant. Causative agent The majority of classic pertussis cases result from infection with Bordetella pertussis, a pathogenic organism with marked tropism for ciliated cells of the respiratory epithelium. B. pertussis is a natural pathogen for humans only. Pertussis-like illness can also be seen with infection by B. parapertussis and adenovirus types 1, 2, 3, and 5. Characteristics of pertussis Clinically, pertussis is unrecognizable in the first, catarrhal stage and becomes a real disease only afterwards, when temperature is normal and the patient is not so much sick as afflicted. Pathologically, changes in the respiratory tract are neither deep-seated nor general since the primary infection attacks superficial mucous membranes. Yet deaths occur and resistance to a second attack continues for a long time. In spite of its being a bacterial infection, pertussis resembles a superficial mucosal viral infection such as those with influenza and respiratory syncytial viruses (Wright 1991). Epidemiologically, the age distribution of deaths is different from that of other communicable diseases of childhood; deaths from pertussis in the first year of life are inordinately frequent because of lack of passive protection acquired from the mother. Clinical aspects Phases of disease Pertussis is an exhausting illness that can persist for several months. The incubation period lasts for 7 to 13 days. This is followed by the catarrhal (or prodromal) phase, which lasts for one to two weeks. Patients in the catarrhal phase exhibit only mild symptoms of an uncomplicated upper respiratory infection with rhinorrhoea, conjunctivitis, lacrimation, mild cough, and low-grade fever. This phase is indistinguishable from a viral infection, making it virtually impossible to diagnose pertussis clinically at Pertussis 31 this point, except in the setting of an epidemic or known exposure. The catarrhal stage is the most infectious period and transmission occurs by airborne droplets. The paroxysmal phase, which follows, usually lasts for 2 to 4 weeks, but occasionally persists as long as 20 weeks. It is characterized by a repetitive series of paroxysmal coughs during a single expiration, followed by a sudden massive inspiratory effort that produces a “whoop” as air is inhaled forcefully against the narrowed glottis. Facial redness or cyanosis, bulging eyes, protruding tongue, lacrimation, and salivation are commonly seen during paroxysms. Haemoptysis, subconjuctival haemorrhages, and hernias may result from severe paroxysms. Young infants may experience apnoea. The coughing is so severe that the patient is commonly unable to eat or sleep. The coughing paroxysm may lead to vomiting of sufficient severity to result in weight loss and malnutrition, especially in young infants with borderline nutritional status (Morley, Woodland & Martin 1966). The final phase of pertussis is the convalescent phase, when paroxysmal episodes gradually decrease in frequency and severity. Coughing may persist for months, especially in association with subsequent respiratory infections. With each such infection the cough may return, although it will not be as severe as in the initial stage. The course of the disease may be influenced by immunization against pertussis. Among 100 Finnish children with culture-confirmed pertussis, all unimmunized patients showed typical pertussis with paroxysmal cough and whoops, while among immunized patients, 43 per cent developed typical pertussis, 49 per cent had atypical disease (with paroxysmal cough but no whoops), 6 per cent had an abortive form of pertussis with sore throat, slightly elevated temperature and transient coughing, and 2 per cent were asymptomatic carriers (Huovila 1982). The disease is more severe in unimmunized individuals, as judged by rates of admission to hospital. Among 1265 children aged 1–4 years, 1.8 per cent of immunized children required hospital admission compared with 7.3 per cent of unimmunized children (Ramsay, Farrington & Miller 1993). Diagnosis Typical cases of pertussis with characteristic cough are easily diagnosed on clinical grounds by health workers and parents who have had previous experience with pertussis manifestations. However, in young infants the presentation of pertussis may be atypical, without the distinctive whoop (Cherry et al. 1988, Walker 1981). Symptoms may be milder in partially immunized individuals and in fully immunized older children or adults with waning immunity. Problems in laboratory diagnosis of pertussis contribute to the substantial underreporting of cases (Halperin, Bortolussi & Wort 1989, Onorato & Wassilak 1987). Isolation of B. pertussis from nasopharyngeal swabs remains the gold standard for the diagnosis of pertussis infection. Isolation of the organism is, however, fraught with potential difficulties. To 32 Global Epidemiology of Infectious Diseases ensure a reasonable chance of bacterial identification, special care must be taken with preparation of culture plates, transport of specimens, and identification of colonies (Wright 1991). Under ideal conditions, 80 per cent of suspected cases can be confirmed by culture, but in most clinical situations isolation rates are much lower (Onorato & Wassilak 1987). The isolation rate of the organism from the respiratory tract is greatest during the catarrhal phase. The frequency of positive culture diminishes when specimens are taken more than two weeks after onset of symptoms or when patients are receiving antibiotics or have been previously immunized (Onorato & Wassilak 1987). Fluorescent antibody staining has been used for the identification of B. pertussis in direct smears from nasopharyngeal swabs and for identification of organisms growing on plates. Fluorescent antibody testing offers the advantage of more rapid laboratory identification; however, this method is less sensitive and less specific (Halperin, Bortolussi & Wort 1989). Serological methods involve measurement of IgG and IgA antibodies to pertussis toxin (PT) and filamentous haemagglutinin (FHA) using the ELISA method. There is a significant rise of ELISA IgG titres to PT and FHA during the pertussis infection. In unimmunized children, a combination of these two determinations and the ELISA IgA for FHA seems to be the most sensitive method for serological diagnosis of pertussis. In immunized persons, however, the interpretation of serological test results is complicated since it is not possible to distinguish between antibodies produced in response to natural infection and antibodies produced following immunization. Treatment The treatment of pertussis is mainly supportive. In severe cases it is critical to prevent hypoxia and pulmonary complications. Delivery of oxygen and gentle suction to remove profuse, viscid secretions may be required, particularly in infants with pneumonia and significant respiratory distress. Maintenance of hydration and nutrition is important. The efficacy of antibiotic therapy with erythromycin depends on the time of its administration. When given during the incubation period, it can prevent the disease or modify its course. When given early in the catarrhal stage, erythromycin can reduce symptoms and/or shorten the course of the disease. Initiation of erythromycin treatment within 7 days after cough onset may be associated with a shorter duration of cough and fewer whoops (Bergquist et al. 1987, Farizo et al. 1992, Steketee et al. 1988). Among children under 12 months of age, those who received erythromycin (irrespective of timing of therapy) were significantly less likely to have pneumonia than those who did not receive such therapy (Farizo et al. 1992). Erythromycin treatment shortens the period of communicability and thus may be useful in an epidemic situation (Bergquist et al. 1987, Huovila 1982). Pertussis 33 Epidemiological aspects Communicability Pertussis is a highly communicable disease. It is likely that no one escapes pertussis in the absence of immunization. By the age of 16 years, almost 100 per cent of children have suffered an episode of pertussis but about 25 per cent of episodes are unrecognized (Thomas 1989). This has been demonstrated by data from epidemic investigations, studies of secondary spread within families, and serological surveys. In pertussis epidemics, attack rates in unimmunized children are high, ranging between 11 per cent and 81 per cent depending on age (Table 2.1). The high degree of communicability has been repeatedly demonstrated by secondary attack rates of 70 to 100 per cent among susceptibles within families (Gordon & Hood 1951). In Sweden, one study of 372 unvaccinated children showed how the cumulative proportion with clinically typical pertussis increases with age: 12 per cent by age 2 years, 35 per cent by age 4 years, 55 per cent by age 6 years, and 61 per cent by age 10 years. Furthermore, an additional 25 per cent of children who did not experience clinical disease had a mild or atypical disease, as evidenced by IgG antibody against pertussis toxin (Isacson et al. 1993). Age distribution of pertussis cases The epidemiological pattern of pertussis is influenced by the age-specific proportions of susceptible and resistant individuals in the community. Although no specific antibody against B. pertussis antigens has been Table 2.1 Pertussis attack rates, by age, during pertussis outbreaks in Côte d’Ivoire, Guatemala, and Indonesia Attack rate (percentage), by country and year of outbreak Age in years Côte d’Ivoire 1982 Guatemala 1968 Indonesia 1983 <1 1 2 59 46 77 35 40 45 50 3 4 71 63 47 20 5 6 61 40 7 8 9 10–14 63 — — — 11 1 — Total 62 19 68 76 81 63 65 Sources: Côte d’Ivoire: Sokal, Soga & Imboua-Bogui 1986, Guatemala: Mata 1978, Indonesia: Expanded Programme on Immunization 1984 34 Global Epidemiology of Infectious Diseases convincingly shown to be responsible for immunity against the disease, the antibody prevalence at different ages may be a valuable index of the exposure to pertussis antigens. The proportion of persons with measurable IgG antibodies against pertussis toxin (PT), considered an important antigen in the pathogenesis of the disease, increases with age, reflecting contact with B. pertussis organisms. The rate of this increase seems to be more rapid in developing countries than in developed countries. In Cameroon, 82 per cent of children were seropositive by 10–11 years of age (Stroffoloni et al. 1991), compared with 67 per cent of Italian children 11–12 years of age (Stroffolini et al. 1989). In New Zealand only 39 per cent were seropositive by 15 years of age (Lau 1989). Before the introduction of pertussis immunization, the peak of the reported incidence was in children 1–4 years of age (Figure 1). In Sweden, where pertussis vaccination was discontinued in 1979, 69 per cent of cases in 1980–1985 occurred in children 1–6 years of age (Romanus, Jonsell & Bergquist 1987). In Kenya during 1974–1977 prior to widespread introduction of pertussis vaccine, 87 per cent of pertussis cases occurred in children 6 years of age or younger, with a median age of 3.5 years, and highest attack rates in children 3–4 years of age (167 cases per 1000) and infants (159 per 1000) (Voorhoeve et al. 1978). The introduction of widespread immunization against pertussis has changed the pattern of the disease (Figure 2.1). Apart from a considerable reduction in the number of cases and abolishing the endemic pattern Figure 2.1 Poland Percentage distribution of pertussis cases before (grey bars) and after (black bars) widespread use of pertussis vaccine, Poland (1971 versus 1991), United States (1918–21 versus 1980–89) <1 year 1–4 years 5–14 years USA <1 year 1– 4 years 5–9 years >10 years 0% 10% 20% 30% Sources: Gordon & Hood (1951), Adonajlo (1975,1993), Farizo et al. (1991). 40% 50% 60% Pertussis 35 of the disease, there has been a clear change in the age distribution of pertussis morbidity. The scope of these changes differs depending on the schedule of vaccine delivery and the coverage rates achieved. In Poland, for example, the most noticeable reduction of pertussis morbidity has been among children 1–4 years of age and the peak incidence has shifted to infants. Infants represented only 12 per cent of all pertussis cases in Poland 1973, compared with 49 per cent in 1993 (Adonajlo 1975, 1993). In the United States of America during 1980–1989, children under one year of age accounted for nearly 50 per cent of all cases; the incidence rate among infants was nearly 10 times higher than that among children of 1–4 years of age, and more than one hundred times higher than that among adolescents or adults (Farizo et al. 1992). There is a lack of reliable information from developing countries on the influence of pertussis immunization on age-specific pertussis morbidity. Data from pertussis epidemics during the pre-vaccine era in Côte d’Ivoire (Sokal, Soga & Imboua-Bogui 1986), Guatemala (Mata 1978) and Indonesia (Expanded Programme on Immunization 1984) show that incidence rates were highest in children aged 1–4 years (Table 2.1). These findings are similar to the situation observed in developed countries in the pre-vaccine era. Death rates Before the mid-20th century, pertussis was a major cause of childhood mortality. Uttley (1960) has compared the evolution of pertussis mortality under the age of 15 years in England and Wales with the mortality in Antigua, a small Caribbean island with relatively accurate mortality statistics going back more than 100 years. In the latter half of the 19th century, pertussis death rates were about the same in England and Wales, and in Antigua; in both places nearly 1 in 1000 children under 15 years of age died from pertussis. In England and Wales, death rates began to decline by the end of the 19th century and had fallen to 20 per million by the 1960s. In contrast, the rates in Antigua did not show significant changes until the mid-20th century. In the 1950s, death rates in Antigua were still ten times higher than in England and Wales. Improved economic conditions, better nutrition, decreasing family size and falling birth rates were probable factors contributing to these decreases in pertussis mortality (Gordon & Hood 1951). Age is an important determinant of pertussis severity and prognosis. The case fatality rate (CFR) is highest in infants (Table 2.2). Infants under 6 months and children born prematurely or with congenital disorders are especially vulnerable to pulmonary complications and suffer higher fatality rates than other groups. In the United States of America, the pertussis CFR declined from high levels in the 1920s (Gordon & Hood 1951) to 0.4 per cent in the 36 Table 2.2 Global Epidemiology of Infectious Diseases Pertussis case fatality rates, by age, United States of America 1927–1989, England and Wales 1980–1991, and Kenya 1974–1976 Fatality rate (percentage) Hospital data, Detroit, USA Age group National surveil- National surveil- Biweekly houselance system, lance system, hold visits, rural USA England and Wales community, Kenya 1927–1929 1950 1980–1989 0 months 1 month 2 months 3 months 4 months 5 months 6–11 months < 1 year 1–4 years 5–9 years 10–14 years 15–19 years 20+ years — — — — — — — 42.3 19.2 2.9 — — — — — — — — — — 3.3 2.2 5.3 — — — 1.3 1980–1991 0.3 0.17 0.3 0.6 0.3 < 0.1 0 0 0 0.1 0.07 0.3 0.007 0.005 0.01 — — Total 22.2 3.1 0.4 0.02 1.3 0.5 0.3 1974–1976 — — — — — — — 3.2 1.5 0.05 Sources: United States, Detroit: Gordon & Hood (1951), United States, national: Farizo et al. (1992), England and Wales: Miller,Vurdien & White (1992), Kenya: Mahieu et al. (1978) 1980s (Farizo et al. 1992), although the rate remained above 1 per cent in infants under 2 months of age (Table 2). In England and Wales, the overall CFR was 0.02 per cent in 1980–1991 (Miller, Vurdien & White 1992). In Sweden, where routine immunization against pertussis was stopped in 1979, three deaths were reported among 2282 hospitalized patients in 1981–1982, yielding a CFR of 0.13 per cent (Romanus, Jonsell & Bergquist 1987). Data on hospitalized patients in developing countries show a high CFR, ranging from 3.3 per cent in South Africa (Ramkissoon, Coovadia & Loening 1989) to 9 to 16 per cent in Kenya (Fendall & Grounds 1965), Nigeria (Omololu 1965, Morley Woodland & Martin 1966), and Uganda (Bwibo 1971). In these studies, CFR was dependent on age, with the highest rates among infants. Hospital data may be biased since young and more severely ill patients and those with complications are those more likely to be hospitalized, influencing the high CFR reported. Thus, the CFR is lower in community-based surveys and in outbreak data. CFRs have ranged from 0.3 per cent in an urban outbreak in the immunized population in South Africa (Strebel et al. 1991) to 1–3 per cent in rural areas of Kenya (Mahieu et al. 1978, Voorhoeve et al. 1978), Côte d’Ivoire (Sokal, Soga & ImbouaBogui 1982) and among Mexican Indians (Heimgartner 1977). In 1968 Pertussis 37 there was a pertussis outbreak with high mortality in one community in Guatemala, with a CFR approaching 15 per cent for children less than 15 years of age (Mata 1978). Only a few countries report the number of deaths attributable to pertussis. In countries that report mortality data, there is evidence that pertussis may be responsible for many more deaths than official mortality statistics show (Miller & Farrington 1988). Analysis of trends in infant mortality rates from infectious respiratory illness has shown an excess of 460 to 700 deaths in England and Wales coinciding with pertussis outbreaks over the period 1977–1982; this amounts to 14 to 22 times the number of pertussis deaths reported during this period (Nicoll & Gardner 1988). Complications following pertussis There are three major categories of complications following pertussis: respiratory, neurological, and nutritional. Respiratory complications The most frequent complications are respiratory (Table 2.3). Pulmonary involvement may be sufficiently severe to compromise respiratory function and cause death. The majority of full-blown cases of pertussis exhibit atelectasis or bronchopneumonia or both. Bronchopneumonia was reported in 14 to 15 per cent of hospitalized cases in Sweden (Romanus, Jonsell & Bergquist 1987) and the United States of America (Farizo et al. 1992), 26 to 37 per cent in the United Kingdom (Miller & Fletcher 1976, Walker et al. 1981) and 74 per cent in South Africa (Ramkissoon, Coovadia & Loening 1989). Pulmonary complications were clearly age-dependent and ranged from 22 to 32 per cent in infants, to 4 to 17 per cent in children over 5 years of age (Farizo et al. 1992, Walker et al. 1981). Apnoea was noted in 2 to 5 per cent of hospitalized cases (Swansea Research Unit of the Royal College of General Practitioners Unit 1987, Walker et al. 1981). Controversy surrounds the issue of whether children who have had pertussis may have long term respiratory sequelae. One study found that children who had experienced pertussis under 5 years of age showed long term respiratory sequelae, such as deterioration of lung function, increase in respiratory symptoms, and increased admission to hospital for both upper and lower respiratory conditions (Swansea Research Unit of the Royal College of General Practitioners 1985). In another study, children who had been hospitalized for pertussis when younger than one year, approximately 9 years previously, showed more respiratory symptoms than controls, although the differences were not statistically significant, and there were no significant differences between cases and controls in lung function indices or in bronchial reactivity (Johnston et al. 1986). Neurological complications There may be a wide variety of neurological manifestations, most common of which are convulsions and altered consciousness. Acute neuro- 38 Table 2.3 Global Epidemiology of Infectious Diseases Complication rates attributable to pneumonia, seizures, and encephalopathy among pertussis patients in seven studies Percentage of patients with Country, study period Reference Type of study Age group Pneumonia Seizures Encephalopathy South Africa 1982–1987 Ramkissoon, Hospital Coovadia study & Loening 1989 All ages 74.0 — 6.9 Sweden 1981–1983 Romanus, Jonsell & Bergquist 1987 Hospital study All ages 14.0 3.2 0.5 UK 1977–1979 Swansea Research Unit 1981 Telephone All ages survey of general practitioners 0.8 0.6 0.1 UK 1969–1980 Walker et al. Hospital 1981 study < 3 months 3–5 months 6–11 months 1–4 years 5 years 0–5 years 32.0 26.0 27.0 26.0 17.0 26.0 — — — — — 4.0 — — — — — — USA 1980–1989 Farizo et al. 1992 National surveillance system 0–1 month 2–3 months 4–5 months 6–11 months < 12 months 1–4 years 5–9 years 10–19 years 20+ years All ages 24.7 22.5 18.8 19.3 21.7 12.3 5.0 4.0 2.9 14.6 4.1 2.7 1.8 3.1 3.0 2.1 0.9 0.5 0.8 2.2 1.4 1.0 0.5 0.7 0.9 0.4 0.4 0.2 0.5 0.7 USA 1985–1989 Sutter & Cochi 1992 Hospital study using national surveillance system All ages 31.0 3.7 1.2 USA 1985–1989 Sutter & Cochi 1992 Hospital discharge database All ages 20.0 2.1 0.2 logical manifestations may also include hemiplegia, paraplegia, ataxia, aphasia, blindness, deafness, and decerebrate rigidity. Central nervous system complications are apparently secondary to anoxia and/or cerebral haemorrhages caused by raised intracranial pressure during episodes of paroxysmal coughing or encephalitis. The risk of neurological complications is markedly age dependent Pertussis 39 (Table 2.3). It is higher in young infants and negligible in older children. Occasionally, cases of neurological complications after pertussis have been reported in adults (Halperin & Marrie 1991). Estimates of the risk of encephalopathy associated with pertussis are not precise because they are based on hospital cases and do not include in the denominator the much larger number of pertussis cases that do not require hospitalization (Mortimer 1992). In hospitalized series, convulsions were noted in 3.7 to 4 per cent (Miller & Fletcher 1976, Walker et al. 1981, Sutter and Cochi 1992) and encephalopathy in 1.2 to 6.9 per cent (Ramkissoon, Coovadia & Loening 1989, Sutter & Cochi 1992). Data from routine reporting systems or outbreak investigations indicate a frequency of encephalopathy between 0.1 and 0.7 per cent (Swansea Research Unit 1981, Romanus, Jonsell & Bergquist 1987, Farizo et al. 1992) and a frequency of convulsions between 0.08 and 3.2 per cent (Swansea Research Unit 1981, 1987, Farizo et al. 1992, Romanus, Jonsell & Bergquist 1987). Complications involving the central nervous system may result in permanent sequelae. Zellweger (1959) estimated that among patients with pertussis encephalopathy, one-third will survive and appear normal, one-third will survive with sequelae, and the remaining one-third will die. Early studies, largely in the pre-vaccination era, suggested an association between pertussis and mental retardation or behavioural disturbance. The prognosis of pertussis encephalopathy depends on the severity of the symptoms. Although simple seizures may have no sequelae, mental retardation or residual sensory and motor deficit may occur after more severe disease (Halperin & Marrie 1991). A large longitudinal cohort study in Britain found that children who had been hospitalized for pertussis were three times more likely to be intellectually abnormal (Butler et al. 1982). Another study in the United Kingdom suggested that convulsions or apnoea as a complication of pertussis may be associated with subsequent intellectual impairment; the study group of children had a significantly lower median intelligence quotient and poorer school attainment than control groups of children who had never had pertussis or who had pertussis without complications (Swansea Research Unit 1987). Nutritional complications Nutritional problems related to repeated vomiting are occasionally of major importance. Inability to maintain adequate caloric intake is a severe problem in previously malnourished children who develop pertussis, a frequent experience in the least developed countries (Morley, Woodland & Martin 1966). Hospitalization rates The likelihood of hospitalization in pertussis cases depends on various factors. Younger patients with more severe pertussis illness are hospitalized more frequently and for longer periods than those who are older and have 40 Global Epidemiology of Infectious Diseases less severe disease. Infants less than 6 months of age have the highest rate of hospitalization. In the United States of America during 1980–1989, the reported hospitalization rate for pertussis cases was 43 per cent overall, with 69 per cent among infants, 26 per cent in children 1–4 years of age and 6 to 9 per cent in persons over 5 years of age (Farizo et al. 1992). Outbreak investigation data found lower hospitalization rates of 46 per cent for infants and 4 per cent for patients older than one year (Sutter & Cochi 1992). In Sweden, the hospitalization rate attributable to pertussis was 12 to 18 per cent for all ages (Romanus, Jonsell & Bergquist 1987). In the United Kingdom, almost 10 per cent of patients were admitted to hospital and this proportion ranged from 60 per cent in children less than 6 months of age compared with 3 per cent in those over 3 years of age (Miller & Fletcher 1976). Other reasons for hospitalization included additional medical factors, such as congenital heart disease, prematurity, asthma or social reasons (Miller & Fletcher 1976). In the United States of America, the average length of inpatient stay was 7 days in 1980–1989 (Farizo et al. 1992). In Sweden, the mean duration of hospital stay was 8 days for infants younger than 6 months, 6 days for children 6–11 months of age, and 4 days for patients older than 12 months (Romanus, Jonsell & Bergquist 1987). In the United Kingdom, the mean duration of hospital stay was 15.2 days, ranging from 17.8 days in children under 6 months of age to 13.8 days in children over 5 years of age (Miller & Fletcher 1976). In South Africa, the average duration of hospital stay was 13 days (Ramkissoon, Coovadia & Loening 1989). Impact of pertussis on child and family health Even today, pertussis may be a prolonged, severe and frightening disease occasionally resulting in serious sequelae. The disease causes considerable distress to both child and family (Johnston et al. 1985). Because of the longlasting course of the disease, pertussis patients are exhausted, lose appetite and weight, and have disturbed sleeping habits. Eight weeks following onset about one-third of children still have symptoms affecting their behaviour. Behavioural changes observed in pertussis patients include irritability, anxiety, and setbacks in development (Mark & Granstrom 1992). Pertussis becomes a “family affair” (Mortimer 1990) because of social and economic consequences for the afflicted families. Episodes of choking, apnoea or cyanosis in ill children are distressing events for the entire family. One study reported disturbed sleep for 78 per cent of parents, with 53 per cent having to attend to the child 4 times or more each night (Mark & Granstrom 1992). The economic consequences of the disease include expenses for medical visits and drugs, and the need to stay at home from work for a prolonged period to take care of the ill child. Pertussis 41 Reported pertussis morbidity Case definitions The ninth revision of the International Classification of Diseases (ICD-9) codes related to pertussis are listed in Table 2.4 (World Health Organization 1975). The ICD-9 has a major category for “whooping cough” (ICD-9 033) which includes diseases caused by all three species of the bacterial genus Bordetella. The subcategories are as listed in Table 2.4. B. pertussis (ICD-9 033.0) causes pertussis, which is the disease entity described in this chapter. B. parapertussis (ICD-9 033.1) usually causes milder respiratory disease and is much less common than pertussis. B. bronchiseptica (ICD-9 033.8) causes tracheobronchitis, and is uncommon. The International Nomenclature of Diseases gives the following synonyms for pertussis: chin cough, morbus cucullaris, and whooping cough; however, it is noted that the term “whooping cough” applies to a clinical manifestation that is not always present (Council for International Organizations of Medical Sciences & World Health Organization 1985). The International Nomenclature of Diseases indicates that the term “pertussis” should be reserved for disease caused by B. pertussis, and should not be used for disease caused by B. parapertussis or B. bronchiseptica. For purposes of surveillance, the following case definitions have been recommended by the World Health Organization (Strassburg 1984): A suspect case of pertussis arises when an individual with a history of severe cough also has a history of any one of the following: • cough persisting 2 or more weeks; • fits of coughing; cough followed by vomiting. A probable case of pertussis is a suspect case with any one of the fol- lowing: • typical findings on physical investigation by a qualified health worker: in young infants prolonged coughing (sometimes without Table 2.4 Ninth Revision of the International Classification of Diseases (ICD-9) codes for pertussis Code Description 033 Whooping cough Includes: pertussis Use additional code, if desired, to identify any associated pneumonia Bordetella pertussis [B. pertussis] Bordetella parapertussis [B. parapertussis] Whooping cough attributable to other specified organism Bordetella bronchiseptica [B. bronchiseptica] Whooping cough, unspecified organism Pneumonia in whooping cough Code also underlying disease (033.0–033.9) 033.0 033.1 033.8 033.9 483.4 Source: World Health Organization (1975). 42 Global Epidemiology of Infectious Diseases • • • • whoop) is followed by a period of apnoea and cyanosis; in older children the paroxysm is often followed by choking and sometimes vomiting and the production of sticky and stringy mucus; paroxysmal coughing is usually followed by a typical breath intake and “whoop”; subconjunctival haemorrhages; exposure to a suspect case in the previous 2–4 weeks; epidemic of pertussis in the area; white blood cell count with 15 000 lymphocytes per ml or more. A confirmed case of pertussis is a probable case with positive culture or immunofluorescence of nasopharyngeal secretions for B. pertussis. Weakness of pertussis surveillance Reported pertussis incidence is likely to be grossly underestimated for many reasons. The disease is frequently underdiagnosed, particularly if the classic symptoms of paroxysmal cough and whoop have not occurred. Difficulties in laboratory confirmation also contribute to underdiagnosis. These problems are compounded by reliance on a passive reporting system with its inherent weakness of underreporting. Reported data based on hospitalized cases may suffer from disproportionate representation of severe cases in younger children and infants. During outbreaks, reporting rates may increase because of temporarily enhanced awareness of physicians, anxiety in the community, and media attention (Crombie 1983). Only an estimated 5 to 25 per cent of all pertussis cases are reported in the United Kingdom and the United States of America (Jenkinson 1983, Hinman & Koplan 1984, Clarkson & Fine 1985, Thomas 1989). Reporting is disproportionately higher for hospitalized patients with classic, laboratory-confirmed disease (Centers for Disease Control 1990). In the United States of America, the completeness of reporting pertussis hospitalizations was highest in infants (34 per cent), declining to 28 per cent for children aged 1 to 4 years, 14 per cent for children aged 5 to 9 years, and 9 per cent for persons 10 years of age and above (Sutter & Cochi 1992). In Canada, there was a nine-fold increase in the number of reported pertussis cases when an active enhanced pertussis surveillance system was introduced in Nova Scotia Province (Halperin et al. 1989). Limitations in the available data on pertussis cases and deaths make comparisons between different regions or countries difficult. Such comparisons are also hampered by differences in the reliability of surveillance data, differences in completeness of reporting of pertussis, and differences in the quality of clinical and bacteriological diagnosis in different countries. In some developing countries, pertussis is still not notifiable, and in others, statistics on pertussis incidence are not reliable. In many countries, the reported numbers reflect only hospitalized cases, reporting is considerably delayed, and available data cannot serve as an operational tool for Pertussis 43 controlling the disease. As of March 1996, only 82 per cent, 69 per cent and 60 per cent of countries had provided data on pertussis incidence for 1993, 1994, and 1995, respectively, to the World Health Organization (Expanded Programme on Immunization 1996). Global trends The worldwide total number of cases reported to the World Health Organization ranged from 2.1 to 2.8 million cases per year at the end of the 1970s, and 600 000 to 800 000 cases at the end of the 1980s (Expanded Programme on Immunization 1996). A further decline in reported cases of pertussis was noted by the beginning of the 1990s when 200 000 to 440 000 cases were reported annually. This represents a global decrease of pertussis morbidity by 85 to 90 per cent over a 15-year period. Obviously, these data represent only a small proportion of the real pertussis morbidity. With an estimated number of pertussis cases in the world of about 30 million in 1990 (see estimates, below), the global completeness of reporting is in the range of 1 to 2 per cent. Regional trends Although existing surveillance data cannot be considered qualitatively accurate, World Health Organization data show substantial decreases in the number of reported cases from 1980 to 1990 (Expanded Programme on Immunization 1996). Data from China show a striking 97 per cent decline during this period. Declines of 75 per cent or more were seen in developing countries of Asia, Latin America, the Middle Eastern Crescent, and sub-Saharan Africa (Table 2.5). India experienced a 65 per cent decrease in number of reported cases. No clear trend is apparent in the Established Market Economies and Formerly Socialist Economies of Europe. Table 2.5 Number of pertussis cases reported to WHO, by demographic region, 1980, 1985 and 1990 Number of reported pertussis cases (in thousands) Demographic region Established Market Economies Formerly Socialist Economies of Europe India China Other Asia and Islands Sub-Saharan Africa Latin America and the Caribbean Middle Eastern Crescent World 1980 1985 1990 Change from 1980 to 1990 66.1 32.7 129.1 60.2 70.3 34.1 6% increase 4% increase 320.1 613.6 289.7 395.9 116.7 145.3 184.4 147.3 136.2 245.5 48.6 104.8 112.4 20.0 57.7 90.0 24.5 31.1 65% decrease 97% decrease 80% decrease 77% decrease 79% decrease 79% decrease 1980.1 1 065.1 440.1 78% decrease Source: Expanded Programme on Immunization 1996 44 Global Epidemiology of Infectious Diseases Figure 2.2 Reported annual incidence of pertussis, by WHO region, 1974–1994 Africa Americas Eastern Mediterranean Europe South-East Asia Western Pacific 10 000 000 Number of Reported Cases 1 000 000 100 000 10 000 1 000 100 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 Year Source: Expanded Programme on Immunization (1996). Similar conclusions can be drawn from the analysis of reported pertussis morbidity by World Health Organization regions (Figure 2.2). The most acute downward trends are seen in the Eastern Mediterranean and Western Pacific. More gradual declines are seen in Africa, the Americas, and SouthEast Asia. No clear trend is apparent in the European Region. Regional data summarize cases from surveillance systems of varying quality, and real trends may be clouded by mixing reliable and unreliable data. The analysis of pertussis incidence at the country level indicates an obvious downward trend in many countries. However, some countries, especially in Europe, have had consistently high or increasing pertussis morbidity, and this is cause for concern. Several European countries continue to have high incidence rates, including Italy, which has had low coverage (Binkin et al. 1992), Germany where vaccine coverage has been spotty, and Sweden, where the immunization programme was interrupted in the mid-1980s (Dittmann 1996). Pertussis 45 Estimated pertussis morbidity and mortality Methods for estimating morbidity and mortality Since 1985, the WHO EPI Information System has published the estimated number of pertussis cases and deaths occurring and prevented through immunization. Since 1992, the method used in the estimation process has involved the following four steps (Table 2.6): The number of pertussis cases without immunization is estimated for each country. The underlying assumption is that 80 per cent of surviving newborns will acquire pertussis in the first 5 years of life. The impact of immunization on the number of cases is assessed by the use of the protection factor (the result of vaccination coverage with 3 primary doses of DPT vaccine in infants and 80 per cent assumed vaccine efficacy). The number of deaths without or with immunization is estimated by using estimated CFRs of 0.04 per cent for developed countries and 1 per cent for developing countries (Galazka 1991). The estimated numbers of cases and deaths are summarized, forming regional and global totals. For this chapter we have developed a second method for estimating pertussis cases, and deaths is based on the following two steps (Table 2.7): Table 2.6 World Health Organization EPI Information System method for estimating number of cases and deaths attributable to pertussis Pertussis immunization status Country status Formula for estimation Number of cases attributable to pertussis Without With Any Any 80% of all newborns No. cases without immunization × protection factor a Number of deaths attributable to pertussis Without Developing Without Developed With Developing With Developed No. cases without immunization × 0.01b No. cases without immunization × 0.0004 c No. cases with immunization × 0.01b No. cases with immunization × 0.0004 c Cases or deaths a. Protection factor = (1 – coverage with 3 doses of DPT vaccine in infants × 80 per cent vaccine efficacy). b. Case fatality rate for developing countries, 1 per cent. c. Case fatality rate for developed countries, 0.04 per cent. Source: Galazka (1991). 46 Global Epidemiology of Infectious Diseases Table 2.7 Demographic region Established Market Economies Formerly Socialist Economies of Europe India China Other Asia and Islands Sub-Saharan Africa Latin America and the Caribbean Middle Eastern Crescent World Assumptions concerning completeness of reporting and case fatality rates used in estimating the number of pertussis cases and deaths by demographic region, 1990 No. cases reported in 1990 Assumed reporting completeness (%) 70 300 12.500 562 000 0.04 225 34 100 6.25 546 000 0.06 328 112 400 20 000 57 700 1.89 0.52 1.71 5 947 000 3 860 000 3 382 000 1.35 0.44 0.62 80 000 17 000 21 000 90 000 24 500 1.04 0.85 8 638 000 2 870 000 1.62 0.33 140 000 17 000 31 100 0.69 4 511 000 0.33 15 000 440 100 Estimated number of cases in 1990 Assumed case fatality rate (%) 30 316 000 Estimated number of deaths in 1990 347 000 The number of pertussis cases is estimated by multiplying the number of reported cases by arbitrarily allocated reporting completeness rates for different World Development Report regions. The number of pertussis deaths is estimated by multiplying the number of estimated cases by arbitrarily allocated CFRs for different World Development Report regions. Estimated global number of pertussis cases and deaths Global estimates obtained for 1990 by using the two methods differ somewhat in the number of cases (Table 2.8), amounting to 39 million with the WHO EPI Information System method compared with 30 million based on the reported cases method. With both methods, the estimated number of pertussis deaths was about 350 000. Thus, for 1990, pertussis occupies the third place among vaccine preventable diseases after the two main killers, measles and neonatal tetanus. For practical reasons, we have used the reported cases method as the basis for further calculations in this chapter. Estimated number of pertussis cases and deaths by age To estimate the number of pertussis cases by age, we assumed that in developing countries 30 per cent, 50 per cent and 20 per cent of pertussis cases occur among children in age groups < 1, 1–4 and 5–14 years (Table 2.9). For developed countries, we assumed that 50 per cent, 25 per cent, 20 per cent and 5 per cent of pertussis cases occur among age groups Pertussis 47 Table 2.8 Estimated global number of pertussis cases and deaths in 1990, based on two methods of estimation WHO EPI Information System method Reported cases method 39 200 000 30 316 000 348 800 347 000 Estimated number of pertussis cases in 1990 Estimated number of pertussis deaths in 1990 Table 2.9 Estimated number of pertussis cases, by age, in developing countries, developed countries, and worldwide, 1990 Developing countries Developed countries Estimated number of cases World Age in years Estimated age distribution (%) Estimated number of cases Estimated age distribution (%) Estimated number of cases <1 1–4 5–14 15+ 30 50 20 0 8 762 000 14 604 000 5 842 000 0 50 25 20 5 554 000 277 000 222 000 55 000 9 316 000 14 881 000 6 064 000 55 000 Total 100 29 208 000 100 1 108 000 30 316 000 < 1, 1–4, 5–14, and > 15 years, respectively. Pertussis deaths in different age groups were estimated by applying various CFRs to numbers of cases in particular age groups (Table 2.10). For 1990, worldwide there were estimated to be around 230 000 pertussis deaths in infants and 88 000 in children 1 to 4 years of age. Estimated number of pertussis complications We estimate that the frequency of bronchopneumonia, convulsions and encephalopathy is 15 per cent, 1 per cent and 0.5 per cent, respectively, of all estimated pertussis cases (30 million in 1990). Thus, we estimate that in 1990, pertussis caused 4.5 million cases of bronchopneumonia and 300 000 cases of convulsions. An estimated 150 000 children suffered Table 2.10 Estimated number of pertussis deaths, by age, developing countries, developed countries, and worldwide, 1990 Developing countries Developed countries World Age in years Estimated case fatality rate (%) Estimated number of deaths Estimated case fatality rate (%) Estimated number of deaths Estimated number of deaths <1 1–4 5–14 15+ 2.6 0.6 0.5 0.0 229 000 88 000 29 000 0 0.10 0.05 0.02 0.02 550 140 40 10 229 550 88 140 29 040 10 740 346 740 Total 346 000 48 Global Epidemiology of Infectious Diseases pertussis encephalopathy in 1990, with about 50 000 of these resulting in long-lasting sequelae. Impact of immunization against pertussis Immunization is the key to preventing pertussis. Whole cell pertussis vaccines, widely used in industrialized countries since the late 1950s and 1960s, and introduced in developing countries within the WHO Expanded Programme on Immunization in the 1970s and 1980s, are of proven efficacy. Experience in Japan, Sweden and the United Kingdom has provided additional evidence of vaccine efficacy. Stopping immunization with pertussis vaccine (Sweden) or a decline of pertussis vaccine coverage (Japan, United Kingdom) in a previously highly immunized population resulted in the resurgence of the disease (Mortimer 1988). By 1994, an estimated 71 million pertussis cases and 626 pertussis deaths were being prevented worldwide each year through immunization (Ivanoff & Robertson 1997). Current killed, whole cell pertussis vaccines are combined with diphtheria and tetanus toxoids and adsorbed onto aluminium salt, comprising DPT vaccine. The World Health Organization recommends three doses of DPT vaccine in infancy (Expanded Programme on Immunization 1995). Many countries have added a booster dose approximately one year later, and some countries schedule another dose at school entry. The immunization coverage with three doses of DPT vaccine in infants is very heterogeneous by region (Table 2.11), and even more varied by country and within countries. Of note, coverage with three doses of DPT vaccine has consistently remained much lower in the African Region, and remains below 50 per cent in some African countries. To effectively control pertussis in the world, all countries should use Table 2.11 Regional and global levels of coverage with three doses of diphtheria-pertussis-tetanus (DPT) vaccine among children by age 12 months for 1985, 1990 and 1995 Percentage of infants with three doses of DPT vaccine, by yeara Region 1985 1990 1995 Africa Americas Eastern Mediterranean Europe South-East Asia Western Pacific 20 52 45 77 35 66 58 70 81 80 82 93 58 83 71 83 91 91 Global 45 83 81 a Based on routine reporting systems data reported to the World Health Organization as of March 1996. Source: Expanded Programme on Immunization (1996). Pertussis 49 available pertussis vaccines in child immunization programmes. Since acellular pertussis vaccines are not generally available, the widespread use of DPT vaccine containing the whole-cell pertussis component should be continued. Efforts should be directed to increase or maintain coverage of infants with three doses of DPT vaccine at 90 per cent or higher in all districts. Surveillance of pertussis morbidity should be strengthened in all countries and, ideally, pertussis should be a reportable disease. More information on the epidemiological pattern of pertussis, especially the age distribution of pertussis cases in developing countries, is needed to develop recommendations on booster doses for children above one year of age. Discussion Based on data presented in this chapter, we estimate that in 1990 pertussis was responsible for 30 million cases and almost 350 000 deaths. Assuming that bronchopneumonia, convulsions and encephalopathy occur in 15 per cent, 1 per cent and 0.5 per cent of pertussis cases, respectively, we estimate that 4.5 million, 300 000, and 150 000 children suffered from these complications in 1990. There have been a number of previous attempts to quantify the burden of disease attributable to pertussis, but most estimates have referred to methods developed by the World Health Organization Expanded Programme on Immunization. A Rockefeller Foundation strategy paper on selective primary health care estimated that in 1977–1978 for developing countries alone pertussis accounted for 70 million infections and 250 000 to 450 000 deaths (Walsh & Warren 1979). During the late 1970s less than 5 per cent of children in developing countries were immunized against pertussis, so the Rockefeller Foundation estimates represent disease burden largely in the absence of vaccination. Their figures were based on World Health Organization estimates modified by extrapolations from published epidemiological studies performed in well defined populations. The Rockefeller Foundation strategy paper classified pertussis as a high priority for disease control because of its high morbidity, high mortality, and the availability of effective control with vaccines. By the early 1980s, the World Health Organization estimated that there were 60 million cases of pertussis worldwide each year with half a million to a million deaths (Muller, Leeuwenburg & Pratt 1986). These estimates took account of pertussis vaccine coverage, which was increasing gradually during this period, as well as work suggesting that in the absence of an immunization programme 80 per cent of surviving newborns would acquire pertussis in the first five years of life (Fine & Clarkson 1984). Conclusion Although in many countries pertussis has been successfully controlled by routine immunization of infants and children, it continues to cause exten- 50 Global Epidemiology of Infectious Diseases sive morbidity and mortality throughout the world. The disease burden is not sufficiently recognized because pertussis incidence and mortality are grossly underreported. Although we estimate some 30 million cases and 350 000 deaths occurred because of pertussis in 1990, very few of these cases and deaths were officially reported through routine surveillance systems. The number of pertussis cases reported annually to the World Health Organization decreased from between 2.1 and 2.8 million at the end of the 1970s to between 200 000 and 400 000 at the beginning of the 1990s. Although these numbers represent only a small proportion of the real morbidity attributable to pertussis, they indicate a genuine downward trend reflecting the increased delivery of pertussis vaccine to children around the world. Over the past 30 years, successful immunization programmes have been implemented in most countries of the world. In developing countries, immunization coverage of infants with a primary series of three doses of DPT vaccine reached 81 per cent in 1990, but there were considerable differences in coverage rates between regions and between and within countries. It is hoped that even higher levels of DPT vaccine coverage can be achieved; however, problems in reaching this goal have arisen in both developed and developing countries. In some developed countries pertussis vaccine coverage has dropped because of apathy and complacency on the part of physicians and parents, negative attitudes toward immunization spread by anti-immunization pressure groups, and litigation over liability for alleged vaccine-related injuries. In a number of developing countries, failure to reach and maintain high immunization coverage results from multiple factors, including weak management of immunization services, missed opportunities to immunize eligible children, and ineffective information and motivation of mothers to return to complete the immunization series. Pertussis is an exhausting illness that can last for several months. The disease is characterized by several immunological and epidemiological peculiarities compared with other communicable diseases of childhood. The newborn infant has little or no maternally transferred passive protection, and deaths from pertussis in the first year of life are inordinately frequent. Mortality rates are highest in infancy, especially in children under 6 months of age. Overall CFRs have been reported from 0.02 per cent to 0.4 per cent in developed countries and from 0.3 per cent to 16 per cent in developing countries. Among the major complications that follow pertussis illness, the most frequent are those affecting the respiratory tract. Bronchopneumonia has been reported in 14 to 31 per cent of pertussis cases. Young infants are especially vulnerable to pulmonary complications. Other major complications of pertussis include convulsions and encephalopathy. We estimate Pertussis 51 that in 1990 nearly 5 million children suffered from one of the serious complications of pertussis. To effectively control pertussis in the world, all countries should use available pertussis vaccines, aiming to achieve and sustain at least 90 per cent coverage of infants with three vaccine doses. References Adonajlo A (1975) Krztusiec [Pertussis]. Przeglad Epidemiologiczny, 29:93–94. Adonajlo A (1993) Krztusiec w 1991 roku [Pertussis in 1991]. Przeglad Epidemiologiczny, 47:67–70. Bergquist SO et al. (1987) Erythromycin in the treatment of pertussis: a study of bacteriologic and clinical effects. Pediatric Infectious Disease Journal, 6: 458–461. Binkin NJ et al. (1992) Epidemiology of pertusssis in a developed country with low vaccination coverage: the Italian experience. Pediatric Infectious Disease Journal, 11:653–661. Butler NR et al. (1982) Recent findings from the 1970 child health and education study: preliminary communication. Journal of the Royal Society of Medicine, 75:781–784. Bwibo NO (1971) Whooping cough in Uganda. Scandinavian Journal of Infectious Diseases, 3:41–43. Centers for Disease Control (1990) Pertussis surveillance—United States, 1986– 1988. Morbidity and Mortality Weekly Report, 39:57–58, 63–66. Cherry JD et al. (1988) Report of the Task Force on Pertussis and Pertussis Immunization—1988. Pediatrics, 81(Suppl):939–984. Clarkson JA, Fine PEM (1985) The efficiency of measles and pertussis notification in England and Wales. International Journal of Epidemiology, 14:153–168. Council for International Organizations of Medical Sciences, World Health Organization (1985). International nomenclature of diseases, Volume II : Infectious diseases, Part 1: Bacterial diseases. Geneva, Council for International Organizations of Medical Sciences. Crombie DL (1983) Whooping cough: what proportion of cases is notified in an epidemic. British Medical Journal, 287:760–761. Dittmann S (1996) Strategy for pertussis prevention in the World Health Organization’s European Region. Journal of Infectious Diseases, 174(Suppl 3):S291–S294. Expanded Programme on Immunization (1984) Outbreak of whooping cough, Indonesia. Weekly Epidemiological Record, 59:26–27. Expanded Programme on Immunization (1995) Immunization policy. Geneva, World Health Organization (WHO/EPI/GEN/95.03). Expanded Programme on Immunization (1996) EPI information system: global summary. Geneva, World Health Organization (WHO/EPI/CEIS/96.07). Farizo KM et al. (1992) Epidemiological features of pertussis in the United States, 1980–1989. Clinical Infectious Diseases, 14:708–719. 52 Global Epidemiology of Infectious Diseases Fendall NRE, Grounds JG (1965) The incidence and epidemiology of disease in Kenya. II. Some important communicable diseases. Journal of Tropical Medicine and Hygiene, 68:113–120. Fine PEM, Clarkson JA (1984) Distribution of immunity to pertussis in the population of England and Wales. Journal of Hygiene, 92:21–26. Galazka A (1991) Pertussis mortality: methods for estimation of number of deaths due to pertussis. Expanded Programme on Immunization, Research and Development Group Meeting, Geneva, 1–2 October 1991. Geneva,World Health Organization. (EPI/RD/91/WP.5/Oct.). Galazka A (1992) Control of pertussis in the world. World Health Statistics Quarterly, 45:238–247. Gordon J, Hood RI (1951) Whooping cough and its epidemiological anomalies. American Journal of the Medical Sciences, 222:333–361. Halperin SA et al. (1989) Persistence of pertussis in an immunized population: results of the Nova Scotia Enhanced Pertussis Surveillance Program. Journal of Pediatrics, 115:686–693. Halperin SA, Bortolussi R, Wort AJ (1989) Evaluation of culture, immunofluorescence, and serology for the diagnosis of pertussis. Journal of Clinical Microbiology, 27:752–757. Halperin SA, Marrie TJ (1991) Pertussis encephalopathy in an adult: case report and review. Reviews of Infectious Diseases, 13:1043–1047. Heimgartner E (1977) Observaciones epidemiológicas y clinicas de una epidemia de tosferina en la región mazahua, Estado de México [Epidemiological and clinical observations on an epidemic of whooping cough in the Mazahua Region, State of Mexico]. Revista de Investigación en Salud Publica, 37:101–111. Hinman AR, Koplan JP (1984) Pertussis and pertussis vaccine: reanalysis of benefits, risks, and costs. Journal of the American Medical Association, 251: 3109–3113. Huovila R (1982) Whooping cough in Finland 1920–1978: statistical and epidemiological studies. Acta Paediatrica Scandinavica. Supplement, 298:1–29. Isacson J et al. (1993) How common is whooping cough in a nonvaccinating country? Pediatric Infectious Disease Journal, 12:284–288. Ivanoff B, Robertson SE (1997) Pertussis: a worldwide problem. Developments in Biological Standardization, 89:3–13. Jenkinson D (1983) Whooping cough: what proportion of cases is notified in an epidemic? British Medical Journal, 287:185–186. Johnston IDA et al. (1985) Impact of whooping cough on patients and their families. British Medical Journal, 290:1636–1638. Johnston IDA et al. (1986) Effect of whooping cough in infancy on subsequent lung function and bronchial reactivity. American Review of Respiratory Disease, 134:270–275. Lau RCH (1989) Pertussis (whooping cough) toxin and Bordetella pertussis whole cell antibody levels in a healthy New Zealand population. New Zealand Medical Journal, 102:560–562. Pertussis 53 Mahieu JM et al. (1978) Pertussis in a rural area of Kenya: epidemiology and a preliminary report on a vaccine trial. Bulletin of the World Health Organization, 56:773–780. Mark A, Granstrom M (1992) Impact of pertussis on the afflicted child and family. Pediatric Infectious Disease Journal, 11:554–557. Mata LJ (1978) The Children of Santa Maria Cauque: a prospective field study of health and growth. Massachusetts Institute of Technology Press, Cambridge, Massachusetts. Miller CL, Fletcher WB (1976) Severity of notified whooping cough. British Medical Journal, 1:117–119. Miller E, Farrington CP (1988) The current epidemiology of pertussis in the developed world: UK and West Germany. Tokai Journal of Experimental and Clinical Medicine, 13(Suppl):97–101. Miller E, Vurdien JE, White JM (1992) The epidemiology of pertussis in England and Wales. Communicable Disease Report, 2:R152–R154. Morley D, Woodland M, Martin WJ (1966) Whooping cough in Nigerian children. Tropical and Geographical Medicine, 18:169–182. Mortimer EA (1988) Pertussis and pertussis vaccine in the industrialized world. Tokai Journal of Experimental and Clinical Medicine, 13 (Suppl):93–96. Mortimer EA (1990) Pertussis and its prevention: a family affair. Journal of Infectious Diseases, 161:473–479. Mortimer EA (1992) Pertussis (whooping cough). In: Krugman S, et al., eds. Infectious diseases of children. St. Louis, Mosby Year Book, pp. 299–313. Muller AS, Leeuwenburg J, Pratt DS (1986) Pertussis: epidemiology and control. Bulletin of the World Health Organization, 64:321–331. Nicoll A, Gardner A (1988) Whooping cough and unrecognised postperinatal mortality. Archives of Disease in Childhood, 63:41–47. Omololu A (1965) Child health in Western Nigeria. West African Medical Journal, 14:255–268. Onorato IM, Wassilak SGF (1987) Laboratory diagnosis of pertussis: the state of the art. Pediatric Infectious Disease Journal, 6:145–151. Ramkissoon A, Coovadia HM, Loening WEK (1989) Whooping cough—a neglected disease in southern Africa. South African Medical Journal, 75:560–561. Ramsay MEB, Farrington CP, Miller E (1993) Age-specific efficacy of pertussis vaccine during epidemic and non-epidemic periods. Epidemiology and Infection, 111:41–48. Romanus V, Jonsell R, Bergquist SO (1987) Pertussis in Sweden after the cessation of general immunization in 1979. Pediatric Infectious Disease Journal, 6: 364–371. Sokal DC, Soga G, Imboua-Bogui GS (1986) Rapport d’une enquête sur la coqueluche: épidémie du village de Dougba, Département de Tingrela, R.C.I., Mai-Août 1982. Organisation de Coordination et de Coopération pour la Lutte Contre les Grandes Éndemies. Bobo-Dioulasso, Haute Volta, No.7.998/ 82/Doc.Tech.OCCGE. 54 Global Epidemiology of Infectious Diseases Steketee RW et al. (1988) Evidence for a high attack rate and efficacy of erythromycin prophylaxis in a pertussis outbreak in a facility for the developmentally disabled. Journal of Infectious Diseases, 157:434–440. Strebel P et al. (1991) An outbreak of whooping cough in a highly vaccinated urban community. Journal of Tropical Pediatrics, 37:71–76. Strassburg MA (1984) Guidelines for the investigation and control of outbreaks of EPI target diseases. Geneva, World Health Organization (EPI/GEN/84.7/rev1.). Stroffolini T et al. (1989) Prevalence of pertussis IgG antibodies in children in Palermo, Italy. Infection, 17:280–283. Stroffolini T et al. (1991) Seroepidemiology of pertussis infection in an urban childhood population in Cameroon. European Journal of Epidemiology, 7:64–67. Sutter RW, Cochi SL (1992) Pertussis hospitalizations and mortality in the United States, 1985–1988: Evaluation of the completeness of national reporting. Journal of the American Medical Association, 267:386–391. Swansea Research Unit of the Royal College of General Practitioners (1981) Effect of a low pertussis vaccination uptake on a large community. British Medical Journal, 282:23–26. Swansea Research Unit of the Royal College of General Practitioners (1985) Respiratory sequelae of whooping cough. British Medical Journal, 290:1937–1940. Swansea Research Unit of the Royal College of General Practitioners (1987) Study of intellectual performance of children in ordinary schools after certain serious complications of whooping cough. British Medical Journal, 295:1044–1047. Thomas MG (1989) Epidemiology of pertussis. Reviews of Infectious Diseases, 11:255–262. Uttley KH (1960) The epidemiology and mortality of whooping cough in the Negro over the last hundred years in Antigua, British West Indies. West Indian Medical Journal, 9:77–95. Voorhoeve AM et al. (1978) Machakos project studies. Agents affecting health of mother and child in a rural area of Kenya: IV—The epidemiology of pertussis. Tropical and Geographical Medicine, 30:125–139. Walker E et al. (1981) Whooping cough in Glasgow 1969–80. Journal of Infection, 3:150–158. Walsh JA, Warren KS (1979) Selective primary health care: an interim strategy for disease control in developing countries. New England Journal of Medicine, 301:867–974. World Health Organization (1975) Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death, ninth revision. Geneva, World Health Organization. Wright PF (1991) Pertussis in developing countries: definition of the problem and prospects for control. Reviews of Infectious Diseases, 13(Suppl 6):S528–S534. Zellweger H (1959) Pertussis encephalopathy. Archives de Pédiatrie, 76:381–86. Chapter 3 Diphtheria Artur M Galazka, Susan E Robertson Introduction Before the advent of routine immunization, diphtheria was a common cause of morbidity and mortality among preschool-aged children. In temperate climates, more than 1 in 20 inhabitants suffered from clinical diphtheria during their lifetime and 5 to 10 per cent of these died of the disease (Griffith 1979). In developing countries, the disease appeared to be endemic. These countries have rarely reported highly visible outbreaks and epidemics, as were characteristic of the pre-vaccine era in industrialized countries. Recent epidemics of diphtheria in several eastern European countries have again drawn attention to this forgotten disease. Reports from developing countries suggest changes in the epidemiological patterns of the disease so that it now occurs in outbreaks, resulting in high case fatality rates and a large proportion of patients with complications (Galazka & Robertson 1995b). Causative agent Diphtheria is caused by the potent toxin produced by Corynebacterium diphtheriae, a gram-positive bacillus. C. diphtheriae strains may be either toxigenic or nontoxigenic. Toxin production results when bacteria are infected by a special phage (corynebacteriophage) carrying the toxin’s structural gene. Human beings are the only reservoir of infection. Transmission occurs via contact with a patient or a carrier, usually via airborne droplets. Untreated patients are infectious for 2–3 weeks, while antibiotic treatment usually renders patients noninfectious within 24 hours (Begg 1994). Immunization does not always prevent asymptomatic carriage of the organism. Less commonly, transmission has been documented by articles soiled by patient discharges and by milk products (Jones et al. 1985, Benenson 1995). 56 Global Epidemiology of Infectious Diseases Clinical features Forms of diphtheria There are five main forms of diphtheria: tonsillar (faucial); pharyngeal; laryngeal or laryngotracheal; nasal; nonrespiratory forms (cutaneous and other). Classical respiratory diphtheria is characterized by a gradual onset after an incubation period of 2 to 5 days. The patient develops pharyngitis, with a red throat slowly becoming covered with a patchy exudate. The exudate gradually spreads and coalesces into a membrane that may cover the tonsils and/or the pharynx. The membrane is distinctly gray, tough, and adherent; it can lead to mechanical obstruction of the airway, requiring tracheostomy or intubation. Attempts to lift a mature membrane may lead to local bleeding. There is a low-grade fever, and enlarged and tender cervical lymph nodes. Cervical node swelling may become so prominent that the patient develops a “bull neck” appearance. Laryngeal diphtheria occurs when the membrane extends into the larynx, and the patient exhibits hoarseness and stridor. Nasal diphtheria presents like a common cold and is characterized by unilateral or bilateral nasal discharge, which is initially clear and later purulent and bloody. The nasal discharge makes it difficult to see the underlying diphtheritic membrane. Cutaneous diphtheria results when C. diphtheriae infects a skin lesion, such as an abrasion, laceration, insect bite, or burn. Cutaneous diphtheria is more common in tropical areas, but is also found in temperate climates and among the homeless and other persons with poor hygiene. Rarely, diphtheria can infect the conjunctiva, ear, or genitalia. Diagnosis The clinical diagnosis of diphtheria is difficult, particularly in countries where the disease has largely been eliminated. The differential diagnosis includes bacterial and viral pharyngitis, infectious mononucleosis, candidiasis, oral syphilis, and ingestion of caustic chemicals (Begg 1994, Strassburg 1984). The presence of a smooth gray membrane should trigger suspicion of diphtheria. The clinical diagnosis should be confirmed by bacteriological examination. Guidelines on laboratory diagnosis of diphtheria have been issued by the World Health Organization (Brooks 1981, Efstratiou & Maple 1994). Swabs should be obtained from the throat, nasopharynx, and from the edge of the membrane, if it is present. These should be rapidly transported to the laboratory for inoculation Diptheria 57 onto the appropriate media. If specimens cannot be transported to the laboratory immediately, use of a transport medium should be considered. Diagnosis of diphtheria should not be made on direct microscopy of the smear, because of problems with false positives and false negatives. All laboratory isolates should be submitted for toxigenicity testing and strain confirmation at a qualified laboratory. Major complications of diphtheria The most common and most serious complications from diphtheria are those caused by the dissemination and effects of diphtheria toxin on the heart (myocarditis) and nervous system (polyneuritis). Myocarditis Cardiac manifestations usually appear in the second week of the disease. The greater the delay in instituting antitoxin therapy, the greater the likelihood of myocarditis (Begg 1994). Diphtheria toxin causes electrocardiographic (ECG) aberrations in 33 to 65 per cent of cases, and these are clinically significant in 10 to 20 per cent of cases (Boyer & Weinstein 1948, Morgan 1963, Desphande, Patel & Shetty 1970, Stockins et al. 1994). Myocarditis may present acutely with congestive failure and circulatory collapse, or more insidiously with progressive dyspnoea, weakness, diminished heart sounds, cardiac dilatation, and a galloping rhythm. Myocarditis is an important complication of diphtheria because it is responsible for approximately half of the mortality from the disease (Morgan 1963, Salih, Suliman & Hassan 1981). The number of deaths rises steadily with the increasing degree of ECG aberrations. In a large series of diphtheria cases, the case fatality rate (CFR) rose from 6 per cent in cases with slight ECG changes to 71 per cent in cases with marked ECG abnormalities (Boyer & Weinstein 1948). A patient with diphtheria may appear to be recovering from pharyngitis only to develop toxic myocarditis, which may result in congestive heart failure. Although the majority of patients who survive myocarditis recover completely, occasionally there may be permanent damage to the heart. Individuals who have apparently recovered from mild myocarditis have died suddenly several weeks later (Tillman 1980). Neurological complications Polyneuritis may involve cranial and peripheral nerve palsies. Peripheral nerve involvement is manifested primarily as a motor defect, usually bilateral. The severity may vary from mild weakness to complete paralysis. Soft-palate paralysis, which occurs during the third week, is the most common manifestation of diphtheritic neuritis. Other manifestations of neuritis include ocular palsy (paralysis of the muscle of accommodation, causing blurring of vision), paralysis of the diaphragm, and limb paralysis indistinguishable from the Guillain-Barré syndrome. The incidence of neurological complications is proportional to the 58 Global Epidemiology of Infectious Diseases severity of infection. Mild cases have a complication rate of 2 per cent while the rate in severe infections can be as high as 75 per cent (Hoeprich 1994). Although neuritis may persist for months, total resolution of all diphtheritic nerve damage is usual. In Sweden, during an outbreak of diphtheria in the 1980s, 6 out of 17 patients (35 per cent) had reversible paralysis (Rappuoli, Perugini & Falsen 1988). Studies of hospitalized patients in Sudan and the United States of America found a frequency of neurological complications between 8 and 10 per cent (Dobie & Tobey 1979, Salih, Suliman & Hassan 1981). Reported diphtheria morbidity and mortality Case definitions The ninth revision of the International Classification of Diseases (ICD-9) codes related to diphtheria are listed in Table 3.1 (World Health Organization 1977). The ICD-9 has a major category for diphtheria (032) which includes all types of infection by Cornybacterium diphtheriae. Subcategories code the anatomical location of the infection. In the early 1980s, the World Health Organization Expanded Programme on Immunization produced guidelines for investigation and control of outbreaks of the diseases targeted by immunization, including diphtheria (Strassburg 1984). This document recommended the following case definitions: Suspected diphtheria: • Acute pharyngitis, • Acute nasopharyngitis, • Acute laryngitis or with a pseudomembrane Probable diphtheria: • Suspected case and • Any one of the following: – typical findings on physical examination by a qualified health worker; Table 3.1 Ninth revision of the International Classification of Diseases (ICD-9) codes for diphtheria Code Description 032 032.0 032.1 032.2 032.3 032.8 032.9 Diphtheria (infection by Corynebacterium diphtheriae) Faucial diphtheria (diphtheritic membranous angina) Nasopharyngeal diphtheria Anterior nasal diphtheria Laryngeal diphtheria (diphtheritic laryngotracheitis) Other, cutaneous diphtheria Diphtheria, unspecified Source: World Health Organization (1977). Diptheria – – – – – – 59 airway obstruction; myocarditis or neuritis (paralysis) 1–6 weeks after onset; exposure to a case of diphtheria in the previous 2 weeks; epidemic of diphtheria currently in the area; death; common alternative diagnoses excluded by appropriate tests (for example; negative throat culture for group A streptococci or negative blood; tests for mononucleosis). Confirmed diphtheria: • Probable case; and • Positive culture of Corynebacterium diphtheriae. Demonstration of toxin production is recommended since there are nontoxigenic strains, but not required in typical cases. Microscopic examination of a smear of a clinical specimen is not sufficiently accurate to substitute for a culture. Global trends Since 1974, countries have submitted official reports of their annual incidence of diphtheria to the World Health Organization (Expanded Programme on Immunization 1996). In 1990, the total number of diphtheria cases reported was 22 624, with 21 442 from developing countries and 1182 from developed countries. The worldwide total number of reported cases fell from between 74 000 and 94 000 per year in the 1970s to an all-time low of between 20 000 and 25 000 per year in 1990–1992. This represents a global decrease in diphtheria morbidity of about 70 per cent over a 20-year period. The vast majority (95 per cent or more) of reported diphtheria cases were from developing countries during the period 1974 to 1990 (Table 3.2). However, the contribution of developing countries to the global total declined to 87 per cent in 1991 and fell further to 13 per cent in 1994. Since 1992, the global number of reported diphtheria cases has risen steeply, with totals of 31 953 in 1993 and 54 718 in 1994 (Expanded Programme on Immunization 1996). This is the result of a massive outbreak, which began in the Russian Federation and Ukraine and eventually spread to nearby countries. Regional trends There are notable differences in trends in reported diphtheria incidence among the World Health Organization regions (Figure 3.1). Since the mid-1980s, significant declines in diphtheria morbidity have occurred in the Region of the Americas and the Eastern Mediterranean and Western Pacific Regions. A less noticeable decline occurred in the South-East Asia Region. No clear change is apparent in the African Region, reflecting the later start of immunization programmes and the lower vaccine coverage levels achieved. In the European Region, diphtheria morbidity was stable for many years, but has increased dramatically since 1992. 60 Global Epidemiology of Infectious Diseases Table 3.2 Number of diphtheria cases reported to the World Health Organization from developing and developed countries, and percentage of total cases occurring in developing countries, 1974 –1994, based on reports received as of March 1996 Year Developing countries Developed countries World Percentage of cases from developing countries 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 75 318 92 767 72 008 85 612 78 302 76 911 97 155 73 922 59 919 49 652 45 146 41 954 28 957 27 883 28 461 29 728 21 442 21 553 14 609 12 388 6 910 1 827 1 549 1 646 1 141 953 822 656 725 1 040 1 545 1 704 1 637 1 256 1 163 966 887 1 182 3 210 5 835 19 565 47 808 77 145 94 310 73 654 86 753 79 255 77 733 97 811 74 647 60 959 51 197 46 850 43 591 30 213 29 046 29 427 30 615 22 624 24 763 20 444 31 953 54 718 98 98 98 99 99 99 99 99 98 97 96 96 96 96 97 97 95 87 71 39 13 Number of cases reported Source: Expanded Programme on Immunization (1996). Diphtheria incidence in developed countries Historical records document that severe epidemics of diphtheria have occurred in a cyclical manner since the 16th century (Table 3.3). In the mid-18th century, an epidemic raged in England, France, the New England colonies, and the West Indies. In the 19th century, diphtheria returned to America and Europe in a great pandemic. During and after the Second World War, a huge diphtheria epidemic occurred in Europe, with incidence rates higher than 100 per 100 000 population (Galazka, Robertson & Oblapenko. 1995a). In 1943 alone there were an estimated 1 million cases and 50 000 deaths attributable to diphtheria in Europe, excluding the USSR (Stowman 1945). The implementation of large-scale diphtheria immunization programmes has led to marked decreases in diphtheria—and in some countries its virtual elimination—over the past three decades. At the beginning of the 1980s, many developed countries were progressing towards the elimination of the disease. In some European countries, not a single case of diphtheria has been reported for more than 10 years. By 1995, coverage of infants Diptheria Figure 3.1 61 Annual diphtheria incidence reported to the World Health Organization, by region, 1974–1994, based on reports received as of March 1996 Africa Americas Eastern Mediterranean Europe South-East Asia Western Pacific Number of Reported Cases 100 000 10 000 1 000 100 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 Year Source: Expanded Programme on Immunization (1996). with three doses of diphtheria-pertussis-tetanus (DPT) vaccine reached 88 per cent among the developed countries (Expanded Programme on Immunization 1996). Widespread immunization has resulted not only in a considerable reduction of diphtheria incidence, but also in profound changes in the immune status of different age groups. Children acquire a high level of diphtheria immunity as the result of infant immunization. The level of immunity declines in late childhood and adolescence. High levels of immunity in children result in reduced incidence of the disease, and, as diphtheria has become rare, opportunities for acquiring or reinforcing natural immunity have also been reduced. In many industrialized countries, adults become susceptible to diphtheria as a consequence of reduced opportunities to boost immunity through subclinical infection. The changes in immunity profiles by age after introduction of routine immunization are presented schematically in Figure 3.2. Diphtheria cannot be considered a disease that has already been con- 62 Global Epidemiology of Infectious Diseases Table 3.3 Cyclic occurrence of diphtheria in the world, 16th–20th centuries Century Place, period Forms and extent 16th–17th Spain, 1583–1613, 1630, 1645, 1666 “Garotillo” or “morbus suffocans” 18th England, France, New England, West Indies, 1735–1740 “Throat distemper” with high case fatality rates 19th Europe and the Americas, 1857–1899 Pandemic with mortality up to 100 per 100 000 population 20th Europe during and after Second World War Virulent form of diphtheria, with incidence > 200 per 100 000 population Russian Federation and Ukraine, 1990–1994 Next round in the diphtheria cycle? quered. Since the 1990s, there has been a striking resurgence of diphtheria in several countries of Eastern Europe (Expanded Programme on Immunization 1993, 1994a, 1994c, Galazka, Robertson & Oblapenko 1995a, Hardy, Dittmann & Sutter 1996). In the Russian Federation, the number of reported cases of diphtheria increased dramatically from 1211 in 1990 to 39 703 in 1994; in Ukraine, reported cases increased from 109 in 1990 to 2990 in 1994 (Expanded Programme on Immunization 1996). The main reasons for the return of diphtheria in these countries were: decreasing immunization coverage among infants and children; waning immunity to diphtheria in adults; mass movements of the population secondary to economic and political changes; and an irregular supply of vaccines. By 1994, this outbreak had spread to nearby countries, including Azerbaijan, Figure 3.2 Age-specific immunity of a hypothetical population before and after widespread use of diphtheria vaccine 100 Pre-vaccine era Per cent immune 80 60 40 Post-vaccine era 20 0 0 5 10 15 20 25 30 35 Age (years) 40 45 50 55 60 Diptheria 63 Belarus, Georgia, Kazakhstan, Kyrgyzstan, Latvia, Lithuania, the Republic of Moldova, Tajikistan, and Uzbekistan. Diphtheria incidence in developing countries To date, diphtheria has not been a major problem in most developing countries. Immunization for infants and children was introduced with the Expanded Programme on Immunization in the late 1970s. Coverage of infants in developing countries with three doses of DPT vaccine rose gradually from less than 10 per cent in 1974 to 81 per cent in 1995 (Expanded Programme on Immunization 1996). In these countries, the process of maintaining immunity still operates through natural mechanisms, including frequent skin infections caused by C. diphtheriae. Socioeconomic changes, especially rapid urbanization with migration from rural areas, and sociocultural changes, including improved hygiene and different lifestyles, are changing the epidemiological patterns of diphtheria, so that in some developing countries epidemics are occurring, with more serious faucial and laryngeal forms. Algeria, China, Ecuador, Jordan, Lesotho, and Yemen have reported diphtheria outbreaks which occurred following a 5 to 10 year period of high immunization coverage. These outbreaks have been characterized by high case fatality rates (CFRs), a large proportion of patients with complications, and occurrence in both young and older age groups (Galazka & Robertson 1995). Death rates Before the era of immunization and modern treatment, diphtheria was a highly lethal disease. During a severe diphtheria outbreak in the New England colonies of the Americas in the mid-18th century, most families lost at least one child (English 1985). Statistics from Hampton Falls, New Hampshire, in 1735 show that the CFR was nearly 40 per cent in children younger than 10 years of age (Table 3.4). In 1881 in New York City, more than 1 per cent of children under 10 years of age died from diphtheria (Doull 1952). From 1880 to 1888, the annual diphtheria CFR in a large hospital in Boston ranged from 40 to 52 per cent (Kass 1993). In Europe during the late 19th century, Hamburg and London reported diphtheria mortality rates as high as 50 to 100 deaths Table 3.4 Mortality attributable to diphtheria, by age groups, Hampton Falls, New Hampshire, 1735 Population Age group in years < 10 10–20 > 20 Total Source: English (1985). Percentage Number of deaths Case fatality rate (%) 404 310 546 32.1 24.6 43.3 160 40 10 39.6 12.9 1.8 1 260 100.0 210 16.7 Number 64 Global Epidemiology of Infectious Diseases per 100 000 population, and the average mortality rate in Europe was approximately 40 per 100 000 population (Kostrzewski 1964). A high risk of death continues to be associated with diphtheria. CFRs of 5 to 10 per cent for noncutaneous diphtheria have changed little in 50 years (Benenson 1995). In the United States of America, from 1971 to 1981, more than 1000 cases of diphtheria were reported, with a relatively stable CFR close to the historically prevailing 10 per cent (Chen et al. 1985). In Poland, the CFR was 13.8 per cent in 1970 (Kostrzewski & Abgarowicz 1973). In the recent epidemic of diphtheria in the Russian Federation and Ukraine, CFRs ranged from 1.5 per cent to 4.8 per cent (Expanded Programme on Immunization 1993, 1994a). Late reporting of diphtheria cases to medical services and delays in diagnosis, hospitalization and treatment are important factors influencing the outcome of the disease. In St Petersburg, only 12 per cent of diphtheria cases were hospitalized during the first 3 days of the disease; the majority of cases were hospitalized 4 to 10 days after the onset of the disease (Rachmanova et al. 1993a). Among six patients who died from diphtheria, all were admitted later than the third day after onset, and three patients were hospitalized at 8, 10, and 40 days after onset (Rachmanova et al. 1993b). During the 1992–1993 epidemic in Ukraine, CFRs were several times higher in rural areas, where the access to medical services was limited and the quality of care was lower than in urban areas (Expanded Programme on Immunization 1994a) (Table 3.5). In developing countries in the 1960s to the 1980s, diphtheria CFRs were high among hospitalized cases. CRFs ranged from 6 to 20 per cent in India (Bhargava & Bhatt 1960, Basappa 1963, Garg & Sharma 1964, Desphande, Patel & Shetty 1970, Tibrewala et al. 1975), to 14 to 39 per cent in African countries (Bwibo 1969, Chintu, Bathirunathan & Patel 1978) and Yemen (Jones et al. 1985). CFRs depend on age. Hospital reports from developing countries reveal high rates (11 to 50 per cent) in infants, moderate rates (8 to 18 per cent) Table 3.5 Diphtheria case fatality rates in urban and rural areas, Ukraine, 1992 and 1993 Part of country 1992 1993 Urban Number of cases Number of deaths Case fatality rate (%) 1 239 35 2.8 2 497 46 1.8 Rural Number of cases Number of deaths Case fatality rate (%) 314 33 10.5 490 32 6.5 Total Number of cases Number of deaths Case fatality rate (%) 1 553 68 4.4 2 987 78 2.6 Source: Expanded Programme on Immunization (1994a). Diptheria 65 in preschool children, and low rates (0 to 5 per cent) in schoolchildren (Bassapa 1963, Bharghava & Bhatt 1960, Desphande, Patel & Shetty 1970, Garg & Sharma 1964, Tibrewala et al. 1975). Data for the United States for 1959–1970 showed the highest CFR (16 per cent) in children under five years of age (Munford et al. 1974). Completeness of routine reporting The true numbers of diphtheria cases and deaths are unknown. In developed countries, where diphtheria occurs in the form of single imported cases or, as recently in the Russian Federation and Ukraine, in definite outbreaks, the reporting may be assumed to be good. In developing countries, however, where the disease is usually endemic, reporting systems are weak and the impact of immunization on disease incidence is monitored primarily through national incidence figures. Such statistics are often inaccurate because they are based on incomplete data gathered by the routine surveillance systems, which are usually hospital-based. Completeness of reporting depends mainly on two elements. First, the public must have access to health services and use them. Second, the health services must report cases accurately and regularly to the appropriate public health authorities. A study in 13 developing countries that compared survey data with reported data found that only 2 to 5 per cent of tetanus cases and 1 to 26 per cent of poliomyelitis cases were detected and reported through routine surveillance systems (Expanded Programme on Immunization 1982). Overall reporting efficiency for the vaccine preventable diseases is estimated to be less than 10 per cent, although this estimate does not specifically address diphtheria case reporting (Expanded Programme on Immunization 1994b). Further efforts are needed to establish highly efficient surveillance systems capable of detecting most cases of targeted diseases. Estimated diphtheria morbidity and mortality Estimated total number of diphtheria cases and deaths The number of diphtheria cases reported from developed countries can be assumed to be relatively accurate; thus, we estimate that reporting was 60 per cent complete for developed countries in 1990. For developing countries, we estimate that reporting was much lower, with no more than 20 per cent of diphtheria cases reported. Thus, the total number of diphtheria cases worldwide in 1990 can be estimated to be about 106 000, with 104 000 from developing countries and about 2000 from developed countries. Assuming a case fatality rate of about 10 per cent for all parts of the world, the total number of deaths attributable to diphtheria in 1990 would be about 11 000. 66 Global Epidemiology of Infectious Diseases Estimated number of diphtheria cases and deaths by age In estimating the number of cases and deaths attributable to diphtheria by age, it is important to consider the differing age distributions of this disease in developing countries (mostly young children) and in developed countries (mostly adolescents and adults). Table 3.6 shows the estimated number of cases by age group occurring in 1990 in developing countries, developed countries, and worldwide. In estimating the number of diphtheria deaths by age, we applied different case fatality rates reported for different age groups. With these assumed rates, the number of estimated deaths attributable to diphtheria in 1990 is about 9000 in children under the age of 5 years and nearly 1600 in school-aged children (Table 3.7). Estimated number of diphtheria cases and deaths by region The estimated number of diphtheria cases and deaths by region is shown in Table 3.8. The highest estimated numbers of diphtheria cases in 1990 were from India, other Asian countries, and sub-Saharan Africa. Estimated number of diphtheria complications Assuming that clinically significant myocarditis occurs in 10 per cent of patients, we estimate that approximately 11 000 cases of myocarditis occurred in 1990. Most of these occurred in children under 5 years of Table 3.6 Estimated number cases of diphtheria, and proportion, by age group, in developing countries, developed countries, and worldwide, 1990 Developing countries Age group in years 0–4 5–14 15–49 >50 Total Number of cases 62 530 31 260 10 420 0 104 210 Table 3.7 Age group in years 0–4 5–14 15–49 >50 Total Developed countries World Proportion (%) Number of cases Proportion (%) Number of cases 60 30 10 0 178 446 1 069 89 10 25 60 5 62 708 31 706 11 489 89 100 1 782 100 105 992 Estimated number of diphtheria cases and deaths, by age group, worldwide, 1990 Estimated no. cases Estimated case fatality rate (%) Estimated no. deaths 62 708 31 706 11 489 89 14 5 2 5 8 779 1 585 230 4 105 992 10 10 598 Diptheria Table 3.8 67 Estimated number of diphtheria cases and deaths, by World Development Report region, 1990 Region Estimated number of cases Estimated number of deathsa Established Market Economies Formerly Socialist Economies of Europe India China Other Asia and Islands Latin America and the Caribbean Sub-Saharan Africa Middle Eastern Crescent Total 40 1 400 42 130 2 100 24 000 5 000 19 300 12 000 4 140 4 213 210 2 400 500 1 930 1 200 105 970 10 597 a. Assumed case fatality rate of 10 per cent for all regions. age and schoolchildren. About half of the myocarditis cases would end in death; only a small portion of the surviving patients would suffer from permanent heart damage. Neurological complications may be expected to occur in about 20 per cent of patients. This would amount to about 21 000 patients worldwide with neurological complications secondary to diphtheria in 1990. Prevention, treatment, and epidemic preparedness Routine immunization Diphtheria toxoid is one of the three components of diphtheria-pertussistetanus (DPT) vaccine. A primary series of three doses of DPT vaccine is given to children in their first year of life in all countries around the world. Global coverage of infants with three doses of DPT vaccine was 45 per cent in 1985, 83 per cent in 1990, and 81 per cent in 1995 (Expanded Programme on Immunization 1996). It needs to be emphasized, however, that overall coverage figures mask regional variations and intra-country variations in coverage (see Table 2.11, Chapter 2, Pertussis). Diphtheria toxoid is a formaldehyde-inactivated preparation of diphtheria toxin, adsorbed on aluminium salts to increase its antigenicity. This toxoid protects against the action of toxin, so when an immunized individual is infected by a toxin-producing strain of diphtheria, the systemic manifestations of diphtheria will not occur. There are no data from randomized controlled trials of the clinical efficacy of diphtheria toxoid, but outbreak investigations have shown a protective efficacy of over 87 per cent (Jones et al. 1985). After a primary series of three doses of DPT vaccine, diphtheria antitoxin levels wane gradually. In industrialized countries, more than half of all adults may lack immunity and many of these countries therefore recommend booster doses of diphtheria toxoid. The same situation does not appear to be true in developing countries, and 68 Global Epidemiology of Infectious Diseases it is thought that this may result from ongoing exposure to persons with cutaneous diphtheria. Thus, most developing countries do not recommend booster doses of diphtheria toxoid. This situation needs careful monitoring, and serosurveys may be appropriate as immunization programmes in these countries mature (Expanded Programme on Immunization 1995). Treatment When a case of diphtheria is identified, strict isolation procedures need to be instituted, including disinfection of all articles in contact with the patient (Benenson 1995). The outcome, including complications and mortality, depends on the speed with which antitoxin treatment is initiated. Antitoxin will only neutralize circulating toxin that is not yet bound to tissue. If diphtheria is strongly suspected, treatment with diphtheria antitoxin should be given immediately after bacteriologic specimens are taken, without waiting for laboratory results (Begg 1994, Farizo et al. 1993). Antibiotic therapy, consisting of intramuscular procaine penicillin G or parenteral erythromycin, should also be instituted, to prevent the patient from transmitting diphtheria to others. Once the patient can swallow, oral therapy with penicillin V or erythromycin should be continued for two weeks. During convalescence, the patient should receive a dose of diphtheria toxoid, because clinical diphtheria does not always confer immunity (Begg 1994). The risk of infection is directly related to the closeness and duration of contact. Anyone who has been in close contact with a case of diphtheria attributable to toxigenic C. diphtheriae in the previous seven days should be considered at risk. This includes household members, sexual contacts, school classroom contacts, those people with diphtheria. Close contacts should have a throat culture taken and receive a dose of intramuscular benzethine penicillin. They should be clinically assessed daily, including a check for pharyngitis, until seven days after their last contact with the case. If a positive culture is obtained in a close contact, they should receive antibiotic treatment with penicillin or erythromycin for two weeks, followed by another throat culture (Begg 1994). Epidemic preparedness Epidemic preparedness is an ever-important issue, brought to attention during the 1990s by the massive diphtheria epidemic affecting most countries of the former Soviet Union. Management of diphtheria outbreaks requires careful public health planning. In the outbreak situation, emphasis must be placed on rapid delivery of vaccine to the affected population, prompt diagnosis and management of diphtheria cases, and appropriate investigation and management of close contacts (Begg & Balraj 1995). In outbreaks with large numbers of adult cases, mass diphtheria immunization should be carried out for adult groups at highest risk, including health care workers, Diptheria 69 teachers, and military personnel. Detailed guidelines have been developed for immunization of adults (Galazka & Robertson 1996). Discussion Based on data presented in this chapter, we estimate that in 1990 diphtheria was responsible for a total of some 106 000 cases worldwide. Assuming a case fatality rate of about 10 per cent for all parts of the world, the total number of deaths attributable to diphtheria in 1990 would be about 11 000. Major complications of diphtheria result from the damaging effects of the bacterial toxin on heart and nerve tissues. For 1990, we estimate that 11 000 individuals suffered from myocarditis, with half of these cases ending in death. In the same year, an estimated 21 000 persons developed neurological complications of diphtheria, but most of these cases can be expected to have resolved spontaneously within a few months. These estimates take account of the disease reduction through vaccination, which had become widespread in both developing and developed countries by 1990. There have been few previous attempts to quantify the burden of disease due to diphtheria. A Rockefeller Foundation strategy paper on selective primary health care estimated that in 1977 for developing countries alone diphtheria accounted for 700 000 to 900 000 cases of disease and 50 000 to 60 000 deaths (Walsh & Warren 1979). During the late 1970s less than 5 per cent of children in developing countries were immunized against diphtheria, so the Rockefeller Foundation estimates represent disease burden largely in the absence of vaccination. A 1980 World Health Organization report estimated that worldwide diphtheria caused at least 100 000 deaths each year among children under five years of age (World Health Organization 1980). One medical textbook chapter on diphtheria published in 1992 states that in developing countries there are estimated 1 million deaths a year caused by diphtheria (Willett 1992). We believe this figure represents an extreme overestimate, which is nearly 100 times greater than our estimate of 11 000 deaths attributable to diphtheria in 1990. Unfortunately the author of this chapter provides no justification and no references for the estimate. The same textbook chapter states that DPT vaccine coverage of infants in developing countries was only 10 per cent (Willett 1992); however, based on official reports by governments to the World Health Organization, coverage of infants in developing countries with three doses of DPT vaccine reached 81 per cent worldwide by 1990 (Expanded Programme on Immunization 1996). Conclusion Diphtheria is still present in many parts of the world. The most effective preventive measure against diphtheria is active immunization of children 70 Global Epidemiology of Infectious Diseases and an adequate programme to maintain immunity against diphtheria throughout life. Many countries that have implemented an effective immunization programme report a continuous decline in the incidence of diphtheria or its virtual elimination. Despite this, countries cannot become complacent about this disease. In 1990, more than 90 per cent of diphtheria cases occurred in developing countries. After 1992, this pattern, seen for three decades, changed abruptly when a massive diphtheria outbreak surged in eastern Europe. This was the largest diphtheria epidemic since 1950 and it was declared by the World Health Organization to be a public health emergency. The history of diphtheria has been marked by sudden unexpected flares, which cause fear and panic because of the high mortality associated with this disease. Many of these epidemic surges have occurred during wars or periods of turmoil, thus exacerbating already difficult situations. The Eastern European outbreak of the 1990s serves as a reminder to all countries to be vigilant. There is a need to improve surveillance for diphtheria and maintain the laboratory capability to diagnose diphtheria even when the disease has declined to very low levels. In addition, diphtheria should be included as part of public health planning for epidemic preparedness. References Basappa K (1963) Epidemiological study of diphtheria in retrospect. Journal of the Indian Medical Association, 41:397–400. Begg N (1994) Manual for the management and control of diphtheria in the European Region. Copenhagen, Expanded Programme on Immunization, World Health Organization Regional Office for Europe (ICP/EPI/038). Begg N, Balraj V (1995) Diphtheria: are we ready for it? Archives of Disease in Childhood, 73:568–572. Benenson AS, ed. (1995) Control of communicable diseases in man, 16th ed. Washington, DC, American Public Health Association. Bhargava HS, Bhatt AN (1960) A study of 275 cases of diphtheria. Journal of the Indian Medical Association, 35:243–248. Boyer NH, Weinstein L (1948) Diphtheritic myocarditis. New England Journal of Medicine, 239:913–919. Brooks R (1981) Guidelines for the laboratory diagnosis of diphtheria. Geneva, World Health Organization (LAB/81.7). Bwibo NO (1969) Diphtheria in African children (report of 13 cases seen at Mulago Hospital). Journal of Tropical Pediatrics, 15:15–17. Chen RT et al. (1985) Diphtheria in the United States, 1971–81. American Journal of Public Health, 75:1393–1397. Chintu C, Bathirunathan N, Patel JM (1978) Diphtheria at the University Teaching Hospital Lusaka. East African Medical Journal, 55:36–38. Diptheria 71 Deshpande RS, Patel JI, Shetty SV (1970) Clinical features of diphtheria in Bombay. Journal of Postgraduate Medicine, 16:5–11. Dobie RA, Tobey DN (1979) Clinical features of diphtheria in the respiratory tract. Journal of the American Medical Association, 242:2197–2201. Doull JA (1952) The bacteriological era (1876–1920). In: The history of American epidemiology. Top FH, ed. St Louis, Mosby, pp 74–133. English PC (1985) Diphtheria and theories of infectious disease: centennial appreciation of the critical role of diphtheria in the history of medicine. Pediatrics, 76:1–9. Efstratiou A, Maple PAC (1994) Laboratory diagnosis of diphtheria. Copenhagen, Expanded Programme on Immunization, World Health Organization Regional Office for Europe (ICP/EPI/038C). Expanded Programme on Immunization (1982) The use of survey data to supplement disease surveillance. Weekly Epidemiological Record, 57:361–362. Expanded Programme on Immunization (1993) Outbreak of diphtheria, update, Russian Federation. Weekly Epidemiological Record, 68:134–140. Expanded Programme on Immunization (1994a) Diphtheria epidemic. Weekly Epidemiological Record, 69:253–258. Expanded Programme on Immunization (1994b) Poliomyelitis in 1993. Weekly Epidemiological Record, 69:169–175. Expanded Programme on Immunization (1994c) Diphtheria outbreak. Weekly Epidemiological Record, 69: 349–352. Expanded Programme on Immunization (1995) Immunization policy. Geneva, World Health Organization (WHO/EPI/GEN/95.3). Expanded Programme on Immunization (1996) EPI information system global summary. Geneva, World Health Organization. (WHO/EPI/CEIS/96.07). Farizo KM et al. (1993) Fatal respiratory disease due to Corynebacterium diphtheriae: case report and review of guidelines for management, investigation, and control. Clinical Infectious Diseases, 16:59–68. Galazka AM, Robertson SE (1995) Diphtheria: changing patterns in the developing world and the industrialized world. European Journal of Epidemiology, 11:107–117. Galazka AM, Robertson SE (1996) Immunization against diphtheria with special emphasis on immunization of adults. Vaccine, 14:845–857. Galazka AM, Robertson SE, Oblapenko GP (1995) Resurgence of diphtheria. European Journal of Epidemiology, 11:95–105. Garg BK, Sharma BK (1964) Diphtheria in Kanpur. Indian Pediatrics, 1:436–444. Griffith AH (1979) The role of immunization in the control of diphtheria. Developments in Biological Standardization, 43:3–13. Hardy IRB, Dittmann S, Sutter RW (1996) Current situation and control strategies for resurgence of diphtheria in newly independent states of the former Soviet Union. Lancet, 347:1739–1744. 72 Global Epidemiology of Infectious Diseases Hoeprich PD (1994) Diphtheria. In: Hoeprich PD, Jordan MC, Roland AR, eds. Infectious diseases: a treatise of infectious processes. Philadelphia, JB Lippincott Company, pp. 373–380. Jones EE et al. (1985) Diphtheria: a possible foodborne outbreak in Hodeida, Yemen Arab Republic. Bulletin of the World Health Organization, 63:287–293. Kass AM (1993) Infectious diseases at the Boston City Hospital: the first 60 years. Clinical Infectious Diseases, 17:276–282. Kostrzewski J (1964) Blonica (Diphtheria). In: Infectious diseases in Poland and their control in 1919–1962 [in Polish]. Warsaw, Panstwowy Zaklad Wydawnictw Lekarskich, pp. 96–117. Kostrzewski J, Abgarowicz A (1973) Blonica (Diphtheria). In: Infectious diseases in Poland and their control in 1961–1970 [in Polish]. Warsaw, Panstwowy Zaklad Wydawnictw Lekarskich, pp. 41–52. Morgan BC (1963) Cardiac complications of diphtheria. Pediatrics, 32:549–557. Munford RS et al. (1974) Diphtheria deaths in the United States, 1959–1970. Journal of the American Medical Association, 229:1890–1893. Rachmanova AG et al. (1993a) Diphtheria in St. Petersburg in 1992 [in Russian]. In: Aktualne problemy infekcjonnoj patologii, Part I. St Petersburg, Pasteur Institute, p. 159. Rachmanova AG et al. (1993b) Analysis of diphtheria deaths in 1992 [in Russian]. In: Aktualne problemy infekcjonnoj patologii, Part I. St Petersburg, Pasteur Institute, p. 160. Rappuoli R, Perugini M, Falsen E (1988) Molecular epidemiology of the 1984–1986 outbreak of diphtheria in Sweden. New England Journal of Medicine, 318: 12–14. Salih MA, Suliman GI, Hassan HS (1981) Complications of diphtheria seen during the 1978 outbreak in Khartoum. Annals of Tropical Paediatrics, 1:97–101. Stockins BA et al. (1994) Prognosis in patients with diphtheric myocarditis and bradyarrhythmias: assessment of results of ventricular pacing. British Heart Journal, 72:190–191. Stowman K (1945) Diphtheria rebounds. Epidemiological Information Bulletin, 1:157–168. Strassburg MA (1984) Guidelines for the investigation and control of outbreaks of EPI target diseases. Geneva, World Health Organization. (EPI/GEN/84.7/ rev1). Tibrewala NS et al. (1975) Age profile of diphtheria in Bombay: a study of 3,453 recent cases in children. Progress in Drug Research, 19:412–422. Tillman DB (1980) Corynebacterial infections. In: Yoshikawa TT, Chow AW, Guze LB, eds. Infectious diseases: diagnosis and management. Boston, Houghton Mifflin, pp. 336–340. Walsh JA, Warren KS (1979) Selective primary health care: an interim strategy for disease control in developing countries. New England Journal of Medicine, 301: 67–974. Diptheria 73 World Health Organization (1977) Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death, Ninth Revision, Volume 1. Geneva, World Health Organization. World Health Organization (1980) Sixth report on the world health situation 1973–1977, Part I: Global analysis. Geneva, World Health Organization. Willett HP (1992) Corynebacterium. In: Joklik WK et al., eds. Zinsser Microbiology, 20th ed. Norwalk, Connecticut, Appleton & Lange, pp. 487–496. Chapter 4 Measles CJ Clements, GD Hussey Introduction Measles has been called the greatest killer of children in history. Over the centuries, it has been responsible for severe epidemics throughout the world. Before measles vaccine became available, virtually all children contracted measles—an estimated 135 million cases and around 7–8 million deaths globally each year (assuming a case fatality rate of 7 per cent in the pre-immunization era in developing countries). Mortality rates from measles fell when socioeconomic conditions improved in industrial countries. Further gains were made against the disease when a vaccine was developed in the 1960s. However, despite its availability, the vaccine has not always been used to its full potential, and measles remains a major preventable cause of childhood mortality in developing countries. Towards the end of the 1970s, the Expanded Programme on Immunization (EPI) adopted measles vaccine as one of its six infant antigens. This resulted in a dramatic increase in immunization coverage, which has contributed significantly to reducing both measles morbidity and mortality. Even though developing nations did not all introduce measles vaccine into their routine immunization programmes until 1984–1985, many were quick to reach high coverage levels. Despite excellent progress, the World Health Organization estimated that as of August 1994, 45 million cases of measles still occurred in the world each year. Notwithstanding the severe underreporting of the disease, data received by the World Health Organization clearly demonstrated a dramatic downward trend in the number of measles cases (Figure 4.1). Some countries (both developed and developing) have experienced more than 99 per cent reduction in the incidence of reported cases. In recent years, however, the number of reported cases in some of those countries has temporarily increased, although still remaining far below pre-vaccine era levels. In developing countries, measles remains endemic in most areas, 76 Global Epidemiology of Infectious Diseases Number of cases of measles and measles vaccine coverage reported globally, 1974–1996 Number of cases % immunization coverage 100 7 6 Coverage (%) 80 5 60 4 40 3 2 20 1 0 73 75 77 79 81 83 85 87 89 91 93 95 0 Number of cases (millions) Figure 4.1 Y Year Source: WHO Expanded Programme on Immunization (1996) with the largest proportion of cases reported in children under 5 years, and over a quarter of all cases involving children under 9 months of age (Taylor et al. 1988). Definitions A case of measles. Until 1996, the World Health Organization used the following clinical case definition: • generalized maculopapular rash; and • history of fever 38°C (101°F) or more, (if not measured, “hot” to touch); and • at least one of the following: cough, coryza, or conjunctivitis. The current WHO case definition is: • any person in whom a clinician suspects measles infection; or • any person with fever and maculopapular rash and cough, coryza, or conjunctivitis. A measles death. A working definition of a measles death is “a death which occurs within one month of onset of measles.” The International Classification of Diseases (ICD) codes for measles are 055 (ICD-9) and B05 (ICD-10). It is probable that some measles deaths may occur beyond one month from onset of the disease. Decreased survival can be recognized epidemiologically following an outbreak of measles. Unless ways are devised of recognizing this phenomenon on a case-by-case basis, however, it is not realistic to count such cases as measles deaths. Wherever possible, a secondary cause of death is defined, such as diarrhoea, acute respira- Measles 77 tory infection or malnutrition. Even though a death may appear to be from other causes (e.g. diarrhoea, trauma), it is counted as a measles death if it occurs within one month of onset of measles. Case fatality rate (CFR). The number of individuals who contract mea- sles and die within one month of the attack, divided by the total number of individuals who contract measles, expressed as a percentage. Acute measles complication. The presence of a sign or symptom of a recognized complication of measles occurring within one month of the onset of measles illness. Delayed measles complication. A number of studies have indicated that children who have had measles have significantly increased morbidity and mortality in the ensuing months when compared with community controls. This is characterized by failure to thrive, recurrent infections, persistent pneumonia, and diarrhoea. Vitamin A deficiency will aggravate these problems. The survival rate of children during this phase is also significantly reduced (Aaby & Clements 1989, Hull, Williams & Oldfield 1983). Measurement of incidence Past attempts by the World Health Organization at measuring disease burden have been twofold. First, the number of cases reported through the routine surveillance system has been collected. These data suffer from gross incompleteness in that only 3–5 per cent of cases that are estimated to occur are actually reported. Sentinel sites reporting more complete data for their catchment area still provide incomplete national data. The very nature of measles means many cases are subclinical or so mild that they never come to the notice of health workers. Thus, recording the number of cases reported to health systems is not likely to provide an accurate picture of the situation, but may, over time, reflect a real trend. Second, the World Health Organization bases estimates of the number of cases occurring on coverage levels of measles vaccine. One of the assumptions made in calculating the estimate (see below) is that all children not immunized will eventually contract measles. This may have been near the truth in the pre-vaccine era, but is clearly not so in situations where little or no measles virus is circulating. For all its potential weaknesses, this method has provided a global estimate that has received general acceptance. Measurement of complication rates Complications from measles are listed in Table 4.1. Of these, respiratory complications and diarrhoea are undoubtedly the most frequent. Uncommon complications include acute encephalitis, nephritis, pneumomediastinum, myocarditis, pericarditis, hepatitis, ileocolitis, and appendicitis. Subacute sclerosing panencephalitis (SSPE) is a rare, long-term complication of measles. It is a degenerative disease of the brain attributable to a 78 Table 4.1 Global Epidemiology of Infectious Diseases Complications of measles Common Uncommon Pneumonia Encephalitis Diarrhoea Myocarditis Laryngotracheobronchitis (croup) Pericarditis Eye disease Pneumomediastinum Malnutrition Nephritis Otitis media Appendicitis Severe stomatitis Subacute sclerosing panencephalitis (SSPE) Nosocomial infections Source: Authors persistent measles virus infection and is the result of an inability of the immune system to clear the viraemia. Studies from five countries (Table 4.2) show that at least three-quarters of cases might be expected to have at least one complication, some with multiple system involvement. The frequency of complications varies in different parts of the world. For instance, a study from the United States of America in 1989 reported complications in 17.4 per cent of cases, including diarrhoea (6.4 per cent), otitis media (6 per cent), pneumonia (4.9 per cent), and encephalitis (0.2 per cent) (Centers for Disease Control 1990). The frequency of complications in developing countries is less well known because of less effective surveillance systems. Hospital-based surveys have nevertheless shown that the three major problems associated with a significantly higher mortality than others are pneumonia, diarrhoea, and croup. Table 4.2 Studies from five countries on the frequency of complications in measles (as a percentage) Country India South Africa Tanzania Thailand Zambia Cases < 12 months 28 69 26 15 17a Unimmunized NR NR 65 NR 47 Malnutrition 47 53 61 51 68 Pneumonia 50 76 75 75 21 Diarrhoea 15 78 21 19 88 Croup NR 32 22 NR NR Deaths NR 10 8 10 15 Deaths < 12 months NR 12 11 19 67a Number of cases 150 97 931 522 41 NR – not reported in this study. a. Less than 9 months. Sources: India: Raote & Bhave 1992, Tanzania: Burgess, Mduma & Josephson 1986, Thailand: Varavithya et al. 1985, South Africa: Hussey & Klein 1990, Zambia: Ministry of Health 1993. Measles 79 Measurement of deaths As with cases, enumerating measles deaths reported to health authorities is problematic. When deaths are recorded, they are frequently incorrectly attributed to diarrhoea, chest infection, or malnutrition when they are actually primarily measles deaths. This is because the underlying cause may have been measles, but complications did not set in for more than two weeks. At this time, those looking after the child may not have considered that measles itself was the cause of death, but rather a complication such as chest infection. In addition, measles causes increased mortality for up to a year after the illness, presumably from an altered immune system. Reports of such deaths are likely to ignore measles as the cause. Because reporting of measles deaths is subject to so much variation, the World Health Organization estimates measles deaths based on the number of cases estimated to occur multiplied by the case fatality rate (CFR). Even using this technique, both the CFR and the number of estimated cases may be largely unsubstantiated in some countries, leaving a measure of uncertainty as to the actual number of deaths attributable to measles. While the actual number of deaths remains uncertain, a number of well documented community and hospital studies reveal the complications from which measles patients die. These are principally acute lower respiratory infections (ALRI), diarrhoea, and others, many of which represent an extremely small percentage of the total. Markowitz & Nieberg (1991) reviewed published articles reporting ALRI-associated measles deaths. In an unpublished revision of these data, Gindler et al. extended the study to include deaths from all complications (see Table 3). In 17 studies reviewed, there were 3515 measles deaths. ALRI deaths reported in community studies were 10 per cent lower than in hospital studies, but it was not possible to draw any conclusions from community versus hospital-based diarrhoea deaths because of the small numbers. A child may have more than one complication from measles and, in the same way, death may be linked with more than one major pathology. From the review in Table 3, it is reasonable to make a general statement about the distribution of deaths. Around 67 per cent are associated with ALRI, 27 per cent with diarrhoea, and 5 per cent with another complication. Data on the prevalence of malnutrition are scant, but we estimate that up to 40 per cent of measles deaths are attributable to some level of malnutrition. Measurement of case fatality rates Difficulties related to measurement of case fatality rates (CFRs) include unreliable civil registration, poorly applied definition of a measles death, and the fact that many measles deaths occur at home, unrecorded by health facilities. Reported CFRs may be subject to bias, in that most information is usually obtained from sources that have an unrepresentative caseload. For example, hospitals tend to admit and report the most serious cases. Such hospitalized cases are at increased risk of dying compared to nonhospitalized cases. 80 Table 4.3 Global Epidemiology of Infectious Diseases Distribution of measles deaths by cause Country of region (study) Bangladesh (Koster et al. 1981) Bangladesh (Spika et al. 1989) Burma (Han, Khin & Hlaing 1990) England (Ellison 1932) Kenya (O’Donovan 1971) Kenya (Hayden 1974) Latin America (Puffer & Serano 1973) Nigeria (Obi 1979) Nigeria (Fagbule & Orifunmishe 1988) Nigeria (Adedoyin 1990) Nigeria (Ibia & Asinda 1990) Papua-New Guinea (Riley et al. 1986) Tanzania (Barclay, Foster & Sommer 1987) West Africa (Morley, Martin & Allen. 1967) Zaire (Markowitz et al. 1989) Zimbabwe (Kambarami et al. 1991) Type of studya Number of deathsb Percentage of deaths attributable for ALRIc Diarrhoea Otherd C C H H H C C 10 122 28 37 66 469 2 106 40 25 50 89 79 72 80 60 — 7 11 42 6 50 0 — 43 0 NEG 22 NEG H H 36 70 81 71 14 44 6 NEG H H C 71 80 22 73 70 64 37 29 — NEG 1 — H 29 48 14 38 H 266 74 56 NEG H H 71 20 66 55 — 30 — 15 a Type of study: C = Community, H = Hospital. b Data collected from death registries at 13 sites in 8 countries. c ALRI = acute lower respiratory infections. d Deaths not attributable to ALRI or diarrhoea. For studies where the total of ALRI and diarrhoea deaths exceeds 100 per cent, this is designated as “NEG.” Recognizing that many countries are uncertain of an accurate value for measles CFR, the World Health Organization/EPI reviewed published community studies of CFRs. Based on this review and on negotiation with countries, the World Health Organization now allocates CFRs to countries. This value ranges from 5 per cent in high-risk countries to 0.1 per cent in most industrialized countries. Such values must be viewed with caution in the light of some community studies which suggest that underestimation may occur (Aaby & Clements 1991) and that further work is needed to refine how CFRs should be calculated. For instance, an outbreak in Afghanistan resulted in a recorded CFR of 28 per cent for all ages and a CFR of 42 per cent for children aged 0–4 years (Wakeman 1978). Tables 4.4–4.6 list studies of case fatality rates, many of which were carried out before significant immunization began. The age-specific incidence of measles in these locations (and hence the CFR) will certainly have altered since then. Many countries have greatly improved their treatment capabilities over recent years and have reduced CFRs accordingly. Nonetheless, recently reported outbreaks confirm that alarmingly high CFRs continue Senegal (Sy 1967) Guinea-Bissau (Aaby et al. 1984a) Kenya (Burström et al. 1992) Kenya (Burström et al. 1992) Kenya (Leeuwenburg et al. 1984) Kenya (Leeuwenburg et al. 1984) Kenya (Muller et al. 1977) Mali (Morley 1973) Mali (Imperato 1975) Nigeria (Ogbeide 1967) Nigeria (Morley et al. 1967) Gambia (McGregor 1964) Some 56 None None None None Outbreak investigation Outbreak investigation Outbreak investigation Outbreak investigation Outbreak investigation Prospective community study Outbreak investigation None 2.5 5.0 4.7 2.8 2.5 None 60 29 20 3.5 3.5 None None Some 90 Median age at infection (years) None Household Survey Cameroon (World Health Organization 1979b) Ethiopia (Lindtjørn 1986) Gambia (Heyworth 1973) Gambia (Hull et al. 1983) Gambia (Lamb 1988) Immunization (per cent) Outbreak investigation Outbreak investigation Outbreak investigation Prospective community study Prospective community study Prospective community study Outbreak investigation Outbreak investigation Outbreak investigation Prospective community study Botswana (Osei & Maganu 1987) Rural Africa Type of study 9 7 2 13 12 8 34 22 9 0 27 79 222 532 134 98 331 101 259 77 26 63 Number of cases 0–4 years Case fatality rate (per cent) Measles case fatality rates in prospective community studies, Africa Country or Region (author) Table 4.4 5 1 38 50 25 2 8 9 6 24 10 5 0 4 0 Case fatality rate (per cent) continued 160 734 123 256 665 252 139 424 162 72 132 54 865 135 Number of cases All ages Measles 81 Outbreak investigation Prospective community study Outbreak investigation 5.0 None 4 6 2 Some 0 21 15 26 2 18 22 20 None Some 1.5 2.5 None Some 2.0 Some None None None 3.5 None 27 None 10 Case fatality rate (per cent) 27 3.6 Median age at infection (years) None None Immunization (per cent) 360 68 1069 78 356 124 119 600 537 87 44 106 160 966 Number of cases Source: derived from Aaby & Clements (1991) Note: Only deaths were registered. The case fatality rate is based on the assumption that all non-immunized children in the age group caught measles. Guinea-Bissau (Aaby et al. 1984b) Guinea-Bissau (Aaby et al. 1988) Nigeria (Hondius & Sutorius 1979) Nigeria (Rea 1968) Zaire (Kasongo Project Team 1981) Zambia (Rolfe 1982) Urban Africa Senegal (Sy 1967) Somalia (World Health Organization 1980) Senegal (Stephens 1986) Prospective community study Prospective community study Outbreak investigation Prospective community study Prospective community study Prospective community study Prospective community study Prospective community study Prospective community study Outbreak investigation Household Survey Senegal (Garenne & Aaby 1990) Senegal (Langaney & Pison 1979) Senegal (Pison 1982) Senegal (Pison & Bonneuil 1988) Senegal (Satge et al. 1965) Type of study 0–4 years Measles case fatality rates in prospective community studies, Africa (continued) Country or Region (author) Table 4.4 17 14 9 10 13 7 Case fatality rate (per cent) 459 161 266 221 68 1500 Number of cases All ages 82 Global Epidemiology of Infectious Diseases Prospective community study Prospective community study Outbreak investigation Outbreak investigation Guatemala (Gordon, Jansen & Ascoli 1965) Guatemala (Mata 1978) Mexico (Cuauhtli, Ruiz-Matus & Neiburg 1990) Mexico (Sanchez-Vargas, Lopez-Ortiz & Bustamante-Hernandez 1990) 39 54 None None Immunization (per cent) 5.0 2.0 3.5 10 4 5 Case fatality rate (per cent) 70 231 292 Number of cases 0–4 years Source: derived from Aaby & Clements (1991). Note: Only deaths were registered. The case fatality rate is based on the assumption that all non-immunized children in the age group caught measles. Type of study Median age at infection (years) Measles case fatality rates in prospective community studies, Americas Country or region (study) Table 4.5 4 6 2 4 Case fatality rate (per cent) 230 882 267 449 Number of cases All ages Measles 83 India (Narain et al. 1989) Bangladesh (World Health Organization 1986) Burma (Chin & Thaung 1985) India (Agarwal et al. 1986) India (Bhatia 1985) India (Chavan, Natu & Pratinidhi 1986) India (Cherian, Joseph & John 1984) India (Dhanoa & Cowan 1982) India (Indian Communicable Disease Bulletin 1985) India (Indian Communicable Disease Bulletin 1986) India (John et al. 1980) India (Lakhanpal & Rathore 1986) India (Lobo et al. 1987) Bangladesh (Koster et al. 1981) Afghanistan (Wakeman 1978) Bangladesh (Bhuiya et al. 1987) Rural Asia None None Outbreak investigation Outbreak investigation Prospective community study Outbreak investigation None None Outbreak investigation 7.0 3.0 2.6 2.9 None None Outbreak investigation Outbreak investigation 12 2 18 1 26 4 10 None 292 322 50 182 77 78 78 103 3 2 354 510 147 3 458 91 79 2 4 422 2 12 5 5.0 3.0 3.4 3.5 6.0 Number of cases 0–4 years Case fatality rate (per cent) None None None None None None None Median age at infection (years) Outbreak investigation Outbreak investigation Outbreak investigation Prospective community study Outbreak investigation Outbreak investigation Prospective community study Prospective community study Household survey Type of study Immunization (per cent) Measles case fatality rates in prospective community studies, Asia Country or Region (author) Table 4.6 7 2 14 1 21 24 8 4 3 2 4 28 771 436 65 241 102 46 182 112 515 3 026 896 327 Number of cases All ages Case fatality rate (per cent) 84 Global Epidemiology of Infectious Diseases Some None Outbreak investigation Outbreak investigation Some Household survey None None Outbreak investigation Outbreak investigation None Outbreak investigation None None None Outbreak investigation Outbreak investigation Outbreak investigation Some 2.5 4.0 5.0 5.5 6.0 3.0 3.5 4.0 2.6 4.0 None None None None None 2.5 None Survey Outbreak investigation Outbreak investigation Prospective community study Prospective community study Outbreak investigation Outbreak investigation 3 33 0 1 16 0 21 4 3 4 787 24 1 072 252 74 318 73 90 72 80 Source: derived from Aaby & Clements (1991). Note: Only deaths were registered. The case fatality rate is based on the assumption that all non-immunized children in the age group caught measles. Burma (Chin & Thaung 1985) India (Satpathy & Chakraborty 1990) India (Siddiqi, Ghosh & Berry 1974) India (Velhaf et al. 1989) Urban Asia Indonesia (World Health Organization 1979a) Indonesia (World Health Organization 1982) Marshall Islands (McIntyre et al. 1982) Papua (Jüptner & Quinnell 1965) Philippines (Almoiradie-Javonillo & Javonillo 1984) Sri Lanka (World Health Organization 1985b) Thailand (World Health Organization 1985a) India (Swami et al. 1987) India (Pereira & Benjamin 1972) India (Reddy et al. 1986) India (Salunke & Natu 1977) India (Sharma et al. 1984) India (Sinha 1977) 67 359 1 1 340 581 50 2 386 2 612 340 4 760 104 731 114 102 181 3 2 0 24 1 0 1 1 12 1 14 4 1 Measles 85 86 Global Epidemiology of Infectious Diseases to occur in some parts of the world during outbreaks. One outbreak in Niger in 1992 recorded CFRs of 18.2 per cent in children under 5 years of age (World Health Organization 1993b) while an earlier outbreak in Afghanistan resulted in CFRs between 28 and 42 per cent (Wakeman 1978). CFRs may vary during outbreaks, and may differ between studies in the community and those in hospitals. Outbreak investigations and retrospective community studies or “verbal autopsies” have been the most common methods used by researchers to measure CFRs. These have been used partly because routine data were not adequate, and partly because events, such as outbreaks, stimulated researchers to undertake the studies. There are difficulties and biases inherent in both these approaches, however. Retrospective surveys, for example, are likely to underreport deaths. Outbreak investigations are likely to be biased in two different ways. First, outbreaks with high mortality are probably more likely to be investigated. Second, medical intervention in the outbreak may tend to reduce the associated mortality. Investigations of measles outbreaks, especially in Africa, have generally reported surprisingly high CFRs. The most reliable estimates of CFRs for measles should therefore come from areas with existing longitudinal studies of child mortality. The World Health Organization has developed a standard protocol for use in the community to measure the CFR during an outbreak (World Health Organization 1993a). More refined estimates of acute measles mortality may not necessarily raise the global estimate of measles deaths significantly. The precise size of the burden of acute deaths from measles, however, remains uncertain. Risk factors for dying The burden of measles deaths falls on the very young and those children already weakened by malnutrition, especially vitamin A deficiency. The risk of dying is higher the younger the age of onset of measles. In the prevaccine era, the burden of cases (and therefore deaths) fell on the very young (Figure 4.2), with resulting high overall CFRs. As immunization programmes have taken effect, the age of onset of the disease has been pushed higher, helping to lower CFRs. Hospital studies indicate that the patient’s state of nutrition appears to be a major risk factor, as mortality during hospitalization is related to the weight-for-age at admission. In contrast, no community study with information on state of nutrition prior to infection has documented this to be a risk factor for outcome (Aaby 1988b). The correlation found in hospital studies may therefore reflect that children with the most severe and potentially fatal infections have lost more weight prior to admission. Whereas anthropometric indices may be of limited value in predicting the case fatality rate in measles infection, studies have implicated vitamin A deficiency as an important determinant of measles mortality (Markowitz et al. 1989). In Indonesia, even mild optical indicators of vitamin A deficiency, such as night blindness and Bitot’s spots, have been associated with mark- Measles Figure 4.2 87 Measles deaths in the Americas, 1971–1980, by age group 70 North Caribbean Central Tropical South Temperate South Deaths (percentage) 60 50 40 30 20 10 0 <1 1–4 5–9 10–14 15–19 20+ Age group edly increased morbidity and mortality in childhood. The incidence of both respiratory infection and diarrhoeal disease, two of the most debilitating complications of measles, were also considerably increased in children who were vitamin A deficient (Sommer, Katz & Tarwotjo 1984). Sex difference Garenne (1994) reported an excess of female measles mortality from birth to age 50. This is contrary to what might be expected in that males generally have higher mortality. Studies of the effect of high-titre measles vaccine used in children under 9 months of age revealed an unexpected and, as yet, unexplained lower survival in females compared with controls (Aaby et al. 1994). It is not clear at this point whether there is a link. Nor is it clear whether reported sex differences are a consistent phenomenon, are biologically or genetically based, or whether local social differences may account for some of the difference. Review of empirical databases by World Bank Region The World Health Organization estimates that by 1995 there were 2.4 million deaths annually from vaccine-preventable diseases and 1.16 million deaths from measles in the developing world. The degree to which measles virus inflicts death on a region or population (Table 4.7) varies widely between countries, between regions within countries, and even between epidemics in the same district. This is partly a result of the difference in age-specific attack rates, the underlying level of nutrition, the presence or absence of supportive treatment services, and other environmental and sociological factors. 24 389 10 680 4 805 23 615 10 985 17 785 18 865 22 280 Number of infants surviving to 12 months of age (thousands) 85 705 14 003 19 418 89 612 37 939 8 921 37 959 37 031 Number of reported cases 0.30 0.10 0.45 3.00 1.68 1.51 2.14 4.64 CFRa (per cent) 93 88 93 92 91 84 90 62 BCGb e. Tetanus toxoid, 2 doses d. Polio, 3 doses c. Diphtheria-pertussis-tetanus, 3 doses b. Bacillus Calmette-Guérin a. Case fatality rate 95 86 92 90 78 79 83 49 DPT-3c 95 85 93 90 87 79 88 49 Polio-3d 94 78 92 85 83 76 84 49 Measles Immunization coverage by vaccine (percentage) The estimated burden of measles by World Bank Region Source: World Health Organization, WHO/EPI information system, as of October 1994. CHI EME FSE IND LAC MEC OAI SSA World Bank Region Table 4.7 2 — — 78 47 57 68 36 TT2+e 88 Global Epidemiology of Infectious Diseases Measles 89 Using the number of cases reported is one method of assessing how many cases of measles are occurring. Table 4.7 shows the number of cases reported by World Bank Region. Although it is not obvious from the table, in general it can be said that developing countries with the highest immunization coverage rates experience the lowest incidence of cases. Assumptions made in calculating the burden of measles As previously mentioned, reported numbers of cases and deaths may be highly unreliable for a variety of reasons. The World Health Organization estimates cases and deaths using a number of assumptions. It is assumed that all susceptible members of the birth cohort will eventually contract measles by one year of age. Because the highest death rate in childhood is under one year of age, and because measles vaccine is generally administered in developing countries between 9 and 11 months of age, the number of children surviving to one year of age (SI) is taken as a component of the numerator. For the purposes of the calculation, measles vaccine efficacy (VE) is taken as 85 per cent (although when administered after one year of age, it may be 95 per cent or more). The following formulae are used: Estimated number = [1 – (Coverage × VE)] × SI of cases per year and Estimated number of deaths occurring per year = Estimated number of × CFR . cases per year These formulae have been used to estimate the numbers of cases and deaths shown in Table 4.8. China region As a result of China’s strong immunization programme, well over 90 per cent of children are immunized against measles. The surveillance system is able to detect a far greater proportion of cases than in many countries, resulting in a high number of reported cases. The burden of disease in this country, is, however, probably not as great as Table 4.8 would suggest, bearing in mind the size of its child population. Of those cases occurring, the CFR is only 0.3 per cent, indicating a strong health infrastructure that is able to treat cases successfully. Established market economies At 78 per cent, measles vaccine coverage is not as high as it could be. Limited resources and difficult access to services are not obstacles to immunization in these countries. Parental choice, is, however, a feature 90 Table 4.8 Global Epidemiology of Infectious Diseases Annual estimated burden of measles by World Bank Region, 1990, unadjusted for background mortality Region Estimated incidence of measles casesa Estimated CFR (percentage)b Estimated number of measles deathsa CHI EME FSE IND LAC MEC OAI SSA 4 902 189 3 559 160 4 233 124 6 553 162 3 235 082 6 295 890 5 395 390 12 993 696 0.30 0.10 0.45 3.00 1.68 1.51 2.14 4.64 14 707 3 559 19 049 196 595 54 349 95 068 115 461 602 907 a. Formula used for the calculation is explained in the text. b. Case fatality rates (CFRs) are estimated based on the World Health Organization's review of published studies and country information, summarised in Tables 4.4, 4.5 and 4.6. of the immunization programmes in many of these nations; when combined with mediocre promotional activities, the result is that too many children are not immunized. Significant numbers of cases continue to occur in pockets of the population, especially in the lower socioeconomic groups, and epidemics continue to be a feature of the disease. Largely excellent treatment services ensure a very low mortality rate. Former Socialist Economies of Europe An expectation that every child will be processed by child health services results in a high coverage in these countries. A strong reporting system means that a higher than average proportion of cases is reported. Nevertheless, treatment services are not quite as successful at preventing deaths as are some of the established market economies that are their European neighbours. India Given that India began using measles vaccine in its national programme only in 1985, the progress is truly impressive. The 85 per cent coverage rate for measles vaccine highlights the enormous numbers of children successfully immunized and protected to date. Higher case fatality rates suggest pockets of low coverage and areas with inadequate treatment facilities, poor access to treatment services and a high level of background ill health and malnutrition including vitamin A deficiency. Latin America and the Caribbean Following mass vaccination campaigns against measles in the past decade, the Latin American and Caribbean countries have been successful in reducing measles. This provides a model for other countries and regions to strive to emulate over the coming years. Measles 91 Middle Eastern Crescent While a few countries have average coverage and measles still occurs, some countries in this region have attained very high coverage and reduced measles cases and deaths close to zero. It is to be anticipated that these high-performance countries will be in a position to eliminate measles in the near future. Countries performing less well have the potential to improve performance rapidly and are expected to reduce the burden of disease accordingly over the next 5 years. Other Asia and Islands Some of the larger nations in this region are still suffering from a considerable amount of circulating measles virus, with significant morbidity and mortality. Other countries, particularly the smaller Pacific Island nations, have been able to reduce measles to very low levels. As with strongly performing countries in the Middle Eastern Crescent, such nations are in an excellent position to consider elimination in the next few years. Sub-Saharan Africa Children in the countries of this region suffer the greatest burden from measles. Underlying ill-health and malnutrition interact with measles infection to produce a high caseload and high CFR. Tragically, this is also the region where measles vaccine coverage is lowest and the health infrastructure is least able to cope with providing immunization and curative services. Administration of measles vaccine is one of the most cost-effective public health tools. A World Bank study calculated the cost-effectiveness of immunization in disability-adjusted life years (DALYs). The cost of gaining one DALY is less than US$10 for measles vaccine, to date the most costeffective of all public health interventions studied (World Bank 1993). Other considerations in calculating the burden of measles Crowded conditions have been found to be associated with higher case fatality rates (Aaby 1988a). Large urban agglomerations typically present ideal conditions for measles transmission, especially among the poor who live in close proximity to each other and are under-immunized. They also experience high rates of other infectious conditions as well as reduced access to medical care. Even in recent outbreaks in such conditions, the CFR can be as high as 20 per cent (World Health Organization 1993b). Transmission also occurs in both urban and rural areas in age groups below the recommended age of immunization. It was previously thought that the achievement of high coverage in older children would protect young children by herd immunity, but this has not always been observed because of the highly infectious nature of measles virus. The age at which children contract measles depends on a variety of local circumstances. Because maternal antibodies decay at variable rates in different parts of the world, different proportions of infants are protected 92 Global Epidemiology of Infectious Diseases against measles in the first few months of life. As maternal antibodies wane, the attack rates for measles increase each month to a maximum between one and two years of age in developing countries. A hospital-based study in Afghanistan reported a typical picture of the age distribution of cases, with 74 per cent of all cases occurring between 4 months and 3 years of age (Arya et al. 1987). Even so, measles is still responsible for more deaths than any other EPI target disease. Measles ranks as one of the leading causes of childhood mortality in the world. In one community study in Kenya (Spencer et al. 1987), measles accounted for 35 per cent of reported deaths in infants 1 to 12 months of age and for 40 per cent of deaths in children 1 to 4 years old. The impact of measles on the lives of millions of young children every year, particularly in developing countries, indicates the urgent need for further reduction in measles incidence. The risk of becoming ill or dying from measles is not evenly distributed in a population but occurs in high-risk groups which include the urban poor, school children (who represent cohorts from previous years when coverage was lower and who may not have been exposed to measles infection), ethnic minorities (who may have been underserved or may have rejected immunization for cultural reasons), hospitalized children (who are at high risk of nosocomial transmission), and children in refugee camps (where crowding facilitates the spread of the virus). A number of studies suggest that measles is more severe and is more likely to result in death in situations where intense exposure occurs, such as “secondary” transmission (transmission from an index case to other children in the same household), crowding and outbreaks. A study in the Machakos District, Kenya (Aaby & Leeuwenburg 1990), showed that in families with several cases of measles, secondary cases had a higher risk of dying than those who were infected from outside. Older children, with their increased mobility, appear to have an increased opportunity for acquiring infection outside the home but lower case fatality rates because of a lower infecting dose and a more mature immune system. Another study from the Gambia showed that measles mortality was strongly associated with more than 5 measles cases in a household (Hull 1988). This dynamic does not seem to affect the severity of disease in the same way in industrialized countries. Sutter et al. (1991) found no difference in severity between primary and secondary cases in a setting where the overall severity of measles was much less and mortality was minimal. Measles as a risk factor for other diseases As mentioned earlier, there are a number of potent co-factors which may act in synergy with measles infection to produce serious disease and complications. Among these are malnutrition (especially vitamin A deficiency) and high background morbidity and mortality from other infections such Measles 93 as diarrhoeal disease. Measles may cause a number of acute complications, described in detail below. Pneumonia Measles is a major cause of pneumonia in children. It has been estimated that measles accounts for 6–21 per cent of all cases of pneumonia and for 8–93 per cent of pneumonia-related deaths (Markowitz & Nieberg 1991). Studies done in the past two decades indicate that pneumonia is a major complication in hospitalized cases, occurring in about 60–80 per cent of cases, with CFRs varying from 5 per cent to 20 per cent (Table 4.9) (Tidstrom 1968, De Buse, Lewis & Mugerwa 1970, Commey & Richardson 1984, Burgess, Mduma & Josphson 1986, Hussey & Klein 1990, Lamabadusuriya & Jayantha 1992, Samsi et al. 1992). Rates during outbreaks may rise dramatically, as illustrated in the report by Narain et al. (1989) describing a measles outbreak in an unimmunized rural community in India. The reported incidence of pneumonia was 39.4 per cent and the case fatality rate was an exceptional 46.3 per cent. There are few published data on the etiology of measles-associated pneumonia. Results from three descriptive studies are shown in Table 4.10, indicating that Streptococcus pneumoniae and Staphylococcus aureus were the predominant isolates (Wesley, Sutton & Widrich 1971, Dover et al. 1975, Morton & Mee 1986). Ristori et al. (1962) reported that about 50 per cent of the fatal cases of pneumonia seen in Santiago in 1962 were attributable to S. aureus. Additional evidence for the important role of bacterial superinfection comes from postmortem studies that reported this problem in 25–50 per cent of fatal cases (Kaschula, Druker & Kipps 1983). The study from Nigeria (Morton & Mee 1986) showed that a higher bacterial isolation rate was found in severe cases of pneumonia (94.7 per cent) compared to milder cases (37.8 per cent, p < 0.01), and in more malnourished children (65 per cent) than in better nourished children (42 per cent, p = 0.07). Limited studies suggest that viruses are also a significant cause of measles-associated pneumonia. Postmortem studies (histology and virology) from Cape Town indicate that measles virus, herpes virus or adenovirus are each responsible for approximately 25 per cent of cases (Kipps & Kaschula 1976). Serological studies from Nigeria (Morton & Mee 1986) and Colombia (Dover et al. 1975) found respectively, that 3 out of 7 cases (42.9 per cent) and 5 out of 21 cases (23.8 per cent) were positive for adenovirus. Croup Recent data from both the United States of America and Africa indicate that croup occurs in approximately 10 to 25 per cent of hospitalized cases of measles (Table 4.9) (Tidstrom 1968, De Buse, Lewis & Mugerwa 1970, Hancock 1972, Commey & Richardson 1984, Burgess, Mduma & b. CFR = case fatality rate. a. IR = incidence rate. 1980–81 1987–88 1988–89 1982–86 1985–87 1990 1948–62 1973–82 1986–87 1972 1990 1981–83 1967–68 Year — 77 — 75 — — 26 63 16 — 68 75 79 IRa (%) — 7 — 10 — — 1 22 — — 4 8 20 CFRb (%) Pneumonia — 13 22 — — 19 4.1 11 — 10 — 18 14 IRa (%) Croup — 7 7 — — 1 — 20 — 39 — 23 21 CFRb (%) 79 80 — 3 71 — — 9 — — 13 13 — IRa (%) — 0 — 8 8 — — 21 — — 0 0 — CFRb (%) Diarrhoea Incidence rates of pneumonia, croup and diarrhoea and respective case fatality rates in hospitalized patients with measles Bangkok (El Behairy et al. 1986) Cape Town (Hussey & Klein 1990) Houston (Ross et al. 1992) Jakarta (Samsi et al. 1992) Lima (Greenberg et al. 1991) Los Angeles (Hancock 1972) Denmark (Tidstrom 1968) Ghana (Commey & Richardson 1984) India (Raote & Bhave 1992) Kenya (Labay et al. 1985) Sri Lanka (Lamabadusuriya & Jayantha 1992) Tanzania (Burgess, Mduma & Josephson 1986) Uganda (De Buse Lewis & Mugerwa 1970) Country or city Table 4.9 94 Global Epidemiology of Infectious Diseases Measles Table 4.10 95 Frequency of bacterial isolates and complications following lung and tracheal aspirates in measles-associated pneumonia Colombia (Dover et al. 1975) Colombia (Dover et al. 1975) Nigeria (Morton & Mee 1986) South Africa (Wesley, Sutton & Widrich, 1971) Aspirate Lung aspirate Tracheal aspirate Lung aspirate Tracheal aspirate Number 21 21 56 22 10 5 0 5 — — — — 62 29 10 19 — — — — 55 55 12 9 3 — 3 — 68 0 46 0 0 14 0 27 Pneumothorax — — 21 23 Case fatality rate 15 — 18 41 Country (study): Bacterial isolates (percentage) Positive cultures S. pneumoniae S. aureus H. influenzae S. pyogenes S. viridans Klebsiella spp E. coli Complications (percentage) Josephson 1986, Hussey & Kelin 1990, Fortenberry et al. 1992, Ross et al. 1992). Case fatality rates vary from about 1 per cent in the United States of America (Ross et al. 1992) up to an exceptional 40 per cent in some areas of Africa (Hancock 1972). There are no recently published data on the incidence of measles-associated croup from Asia or South America. There are few data on the aetiology of measles-associated croup. Many children with measles present with croup early in the clinical course, this complication probably being caused by the measles virus itself. Unpublished data from Cape Town suggest that herpes virus is a common cause of post-measles croup. Some cases may be associated with oropharyngeal lesions. Herpes virus (from the oropharynx) and adenovirus (from the lung) were cultured in 2 of 6 patients with croup in one study from the United States of America (Ross et al. 1992). In the Cape Town study, 13 of 189 children (6.9 per cent) were presumed to have early onset measles-croup and none required airway intervention. In contrast, late onset croup, presumed to be herpes-associated, occurred in 40 of 189 patients (21.2 per cent) and 12.5 per cent required intubation (Hussey & Klein 1990). Bacterial tracheitis is an uncommon cause of stridor and is usually due to S. aureus, S. pneumoniae or Haemophilus influenzae infection (Dover et al. 1975, Labay et al. 1985). Its presentation in measles is not much different from that in non-measles cases (Conley, Beste & Hoffman 1993). Experience indicates that it should be considered in patients whose symptoms, in addition to stridor, include toxicity, purulent sputum, or associated pneumonia. 96 Global Epidemiology of Infectious Diseases Diarrhoea It has been estimated that measles-associated diarrhoea accounts for 1–7 per cent of all diarrhoeal episodes and 9–77 per cent of diarrhoeal deaths (Feachem & Koblinsky 1983). The proportion of measles cases presenting with diarrhoea varies greatly as does the CFR (Table 4.9). In a measles outbreak in an unimmunized rural community in India the reported incidence (and CFRs) of diarrhoea and dysentery were, respectively, 32.2 per cent (4.8 per cent), and 10.8 per cent (13.3 per cent) (Narain et al. 1989). Few studies have evaluated the role of measles in persistent diarrhoea. A prospective, community-based study from Bangladesh (Koster et al. 1981) reported that 25 per cent of measles-associated diarrhoea lasted for 7 or more days. The CFR was significantly higher when associated with prolonged diarrhoea (i.e. lasting 7 or more days); 11.9 per cent compared to 1 per cent in those with diarrhoea lasting less than 7 days. In addition, in children with acute diarrhoea, those with measles-associated diarrhoea had a higher case fatality rate (1 per cent) than those without measles (0.1 per cent). In contrast, a prospective hospital-based study (Dutta et al. 1991) on persistent diarrhoea reported that no such cases occurred in children who had reported having had measles in the previous 6 months. The aetiology of measles-associated diarrhoea has been well defined in four case-controlled studies (Varavithya et al. 1985, 1989, Greenberg et al. 1991, Sang et al. 1992). A few distinct trends are noted when comparing the cases (measles-associated diarrhoea) and controls (non-measles-associated-diarrhoea). The frequency with which no pathogen was identified was higher in cases than in controls in all of the studies. This would suggest that many cases of measles-associated diarrhoea (varying from 8 to 46 per cent) were attributable to measles virus per se. Rotavirus was isolated in 16 to 30 per cent of controls in three studies where it was looked for, while it was virtually never isolated in cases of measles (Varavithya et al. 1989, Greenberg et al. 1991, Sang et al. 1992). Parasitic infections were more common in the two studies (Varavithya et al. 1989, Greenberg et al. 1991) where this was looked for. The reason for this phenomenon may be related to the immune suppression of measles. No difference was noted in bacterial isolation rates, except for one study (Greenberg et al. 1991) indicating a higher rate of campylobacter isolation. Additional support for the belief that measles virus may be a cause of mild diarrhoea comes from a descriptive study from Egypt (El Behairy et al. 1986) which reported that the bacterial stool cultures of children with mild diarrhoea (n = 33) were negative in 42 per cent of cases compared to 27 per cent in children with moderate diarrhoea (n = 11) and none in severe diarrhoea (n = 6). Malnutrition Malnutrition has generally been held to be the major predictor of mortality in cases of measles. Data derived mainly from hospital surveys show Measles 97 that children who are significantly malnourished have a higher morbidity and mortality rate (Table 4.11) (Burgess, Mduma & Josephson 1986, Samsi et al. 1992, Alwar 1992). This, however, does not necessarily imply a causal relationship. It has been shown that measles is an extremely catabolic event and that children lose a significant proportion of weight following infection. The amount of weight loss may be related to severity of infection (infecting dose). The presence of dehydration resulting from diarrhoea and reduced fluid intake exacerbates weight loss and complicates the assessment of nutritional status. Most studies have not controlled for dehydration. An exception is a study from Kenya (Sang et al. 1992) which also reported a longer hospital stay for severely malnourished children. If malnutrition is a risk factor, then the effects are probably mediated via immune suppression. In a review of risk factors for measles, Aaby (1991) reported that no difference in mortality was noted in relation to nutritional status in community based studies (except for one in Bangladesh). Conjunctivitis and keratitis Conjunctivitis and keratitis are hallmarks of measles and mild cases usually resolve within a few days of onset. Treatment is supportive and topical antibiotics are frequently prescribed, although their efficacy in preventing secondary bacterial infections has not been tested. Significant corneal damage in children with measles can result in blindness (Foster & Sommer 1987). The causes of corneal damage include measles virus infection, secondary herpes or bacterial infection, and chemical conjunctivitis as a result of harmful eye practices such as herbal remedies and vitamin A deficiency (xerophthalmia). Table 4.11 Association between nutritional status and complications Percentage of expected weight Country of city (study) All >80 60–80 <60 Pneumonia 74.9 61.9 84.6 100.00 Diarrhoea 18.6 13.7 19.9 39.2 Convulsions 7.5 8.2 7.7 2.2 Otitis media 5.9 6.7 5.9 2.2 23.1 Morbidity Jakarta (Samsi et al. 1992) Mortality 10.3 5.9 12.7 Kenya (Foster & Sommer 1987) 1.8 0.8 1.4 4.0 Tanzania (Burgess, Mduma & Josephson 1986) 8.0 3.6 7.3 24.5 Jakarta (Samsi et al. 1992) 98 Global Epidemiology of Infectious Diseases Xerophthalmia The association between infections and xerophthalmia is a well-recognized observation. In a global review of xerophthalmia, Oomen, McLaren & Escapani (1964) stated that “there appears to be a universal relationship between infectious diseases and xerophthalmia. This relates especially to measles….” The results from case control studies, however, are conflicting. Studies from Ethiopia (De Sole, Belay & Zegeye 1987) and Malawi (Tielsch et al. 1986) reported significant risk ratios (4.6 and 1.6 respectively) but studies from the Philippines (Solon et al. 1978) and Bangladesh (Stanton et al. 1986) found no significant association. A prospective study from India (Reddy et al. 1986) reported that conjunctival signs of xerophthalmia developed in 1.1 per cent of children (3 out of 281) in the 6-month period following measles, compared to 0.5 per cent of children (4 out of 819) without measles. The relative risk of developing xerophthalmia was 2.19 times greater following measles, although the 95 per cent confidence intervals were wide, 0.49 to 9.71, p = 0.25. Children presenting with acute corneal xerophthalmia during the early stage of measles (10 of 318) were excluded from the analysis. Acute encephalitis Central nervous system (CNS) manifestations may be a consequence of a number of measles-associated complications including hyperpyrexia, hypoxaemia associated with pneumonia and croup, and dehydration and electrolyte disturbance associated with diarrhoea. In a 1963 communitybased survey in the United Kingdom (Miller 1964), neurological disturbances, which include a range of CNS problems including encephalitis, were reported in 0.4 per cent of cases; and 0.1 per cent were regarded as having encephalitis. In a subsequent hospital-based study in Denmark (Tidstrom 1968), encephalitis was seen in 1.4 per cent of patients with a CFR of 9 per cent. At follow-up, 32 per cent of survivors had some neurological dysfunction. The epidemiology of this complication in developing countries has not been well described. Reported CNS complications in developing countries (Varavithya et al. 1985, Wahab et al. 1988, Samsi et al. 1992, Raote & Bhave 1992, Lamabadusuriya & Jayantha 1992) are shown in Table 4.12. The major area of concern relates to the diagnostic criteria for encephalitis in developing countries. The lack of standard case definitions makes comparison of data difficult. Otitis media and stomatitis Although otitis media and severe stomatitis are well recognized complications of measles, there are very few publications dealing with their aetiology, natural history, complications, and management. A study from South Africa reported that herpes virus infection was detected in 43 per cent of hospitalized cases and 37 per cent of outpatient children with measles (Orren et al. 1981). Measles Table 4.12 99 Frequency of central nervous system complications in measles Location (study) Year Clinical criteria Percentage Bangkok (Varavithya et al. 1985 1980–81 Encephalitis Ghana (Commey & Richardson 1984) 1973–82 Coma or convulsions 3.2 India (Raote & Bhave 1992) 1986–87 Encephalitis Meningitis Convulsions 8.0 3.3 2.6 1990 Encephalitis 4.3 Sudan (Wahab et al. 1988) 1988 Convulsions Thailand (Samsi et al. 1992) 1982–86 Sri Lanka (Lamabadusuriya & Jayantha 1992) Convulsions Encephalopathy 13.9 7.2 17.0 Nosocomial infections Children with measles are prone to secondary infections, particularly when hospitalized. Data are, however, lacking on the subject. A retrospective case-controlled study in South Africa reported that nosocomial bacteraemia was six times more common in children with measles than in general paediatric patients (Hussey & Simpson 1990). The predominant organisms were gram-negative ones (Klebsiella and Salmonella), accounting for 86.5 per cent of the isolates, and 23 per cent were multiple drug resistant. Morbidity from delayed complications A number of studies have indicated that children who have had measles have significantly increased morbidity and mortality in the ensuing months when compared with community controls. A case-controlled study in Lima (Greenberg et al. 1991) reported that the number of episodes of diarrhoea in the month following measles was significantly greater than in the absence of measles, as was the duration of episodes. No differences in weight gain pattern were, however, observed, which may be ascribed to good home-based follow up. A prospective, community-based study in Bangladesh (Koster et al. 1981) reported that children with measles complicated by diarrhoea for 7 days or more failed to improve their nutritional status compared to controls with diarrhoea or to children with measles complicated by a brief attack of diarrhoea. In India, in a follow-up study of children for a period of 6 months after hospitalization, Bhaskaram et al. (1984) found that children gained hardly any weight in the first month but thereafter they gained weight, although more slowly than the controls. In addition, the frequency of other infections was 10 times that of the controls during the period. In a prospective, community-based study of 281 cases of measles and 819 controls, Reddy et al. (1986) found that 34 per cent of the children with measles developed chest infections in the 6 months following infection, 100 Global Epidemiology of Infectious Diseases compared to 6 per cent of the control children. In addition, the incidence of hospitalization was, respectively, 2.9 per cent versus 0.4 per cent. It has generally been believed that measles predisposes people to the development of tuberculosis or its reactivation as a consequence of immune suppression. A comprehensive review by Flick (1976) concluded that the evidence for this assumption was not very strong. Decreased tuberculin skin reactivity should not be equated with reduced tuberculosis immunity or increased risk of tuberculosis. No recent data support or refute the hypothesis. Measles has been associated with recurrent chest infections and bronchiectasis. The magnitude of the problem has, however, not been quantified. In South Africa, a number of cases have been shown to follow measles (Kaschula, Druker & Kipps 1983). Studies in West Africa have reported that children with measles had a fivefold to tenfold greater risk of dying in the year following an attack of measles, compared to community controls (Hull 1988). The effect was most noticeable in those under one year of age. A recent study in Kenya (Burström et al. 1992) reported a mortality rate of 4.3 per cent in children aged 8–35 months living in compounds where there was a measles outbreak the previous year. The mortality rate in the compounds where there was no measles was much lower (1.1 per cent), but this was not statistically significantly because of the small numbers. The reasons for the increased morbidity and mortality following measles are unclear, but may theoretically be attributable to viral persistence (although this has never been demonstrated), persistent immune suppression or vitamin A deficiency, or more likely a combination of these. Certain conditions are known to result from measles infection and cause long term disability. These include blindness, mental retardation, brain damage (including brain abscess, deafness, and fits), and vitamin deficiencies. Conclusion Case fatality rates have dropped dramatically over the past 20 years, especially in countries with good primary health care facilities. Simple inexpensive treatments for complications prevent death. These include the administration of oral rehydration fluids, antibiotics, nutritional management, and vitamin A. Case fatality rates remain high in areas where these measures are not available, where communities are not provided with the services, not educated to avail themselves of the services or are unable to afford them, or where there is an underlying problem of malnutrition or vitamin A deficiency. Burden with no intervention If there were no immunization with measles vaccine, no administration of vitamin A and no treatment of complications, measles would remain Measles 101 one of the greatest killers of children. Virtually every child in the world would eventually contract measles. At least 90 per cent of children under 5 years of age could be expected to contract the disease. In industrialized countries, at least 10 per cent of children would develop complications, but only about 0.1 per cent of those would die (mostly from encephalitis). In developing countries, however, with high transmission rates in very young children, complication rates could be as high as 70–90 per cent, with case fatality rates at 6 per cent, and as high as 25 per cent in some outbreaks. A conservative estimate would be 16 million deaths; but taking into account very high local case fatality rates and the possibility of delayed mortality, that figure could easily be doubled. Cost of interventions High measles vaccine coverage throughout a population, followed by mass campaigns, has been shown capable of eliminating measles in the Americas. In the early 1990s, measles vaccines were inexpensive at about US$0.15 a dose and were at least 85 per cent effective when administered correctly. There is strong evidence suggesting that the measles virus could be eliminated from large portions of the world with existing technologies. Global eradication will be a greater challenge but is considered by many to be at least theoretically possible. In the meantime, properly applied simple treatments could reduce deaths to as low as 0.1 per cent of cases (complications such as encephalitis are likely to prevent case fatality rates from reaching zero). Measles immunization is probably the most cost-effective public health intervention in the world. Estimates suggest that the cost per life-year gained for expanding measles coverage in a routine programme from 50 per cent to 80 per cent is US$2.53 (at a discounted rate) in areas with high disease incidence and high measles case fatality rates. The cost of immunizing a child against measles in a mass campaign is assumed to be in the order of US$0.60 if conducted jointly with a poliomyelitis national immunization day, or US$1.20 for a stand-alone measles campaign. These figures can be combined with modelling results to calculate the cost per death prevented. This suggests that adding mass measles campaigns to a strategy of improving routine coverage is highly cost-effective. For a baseline campaign aimed at children aged 9–59 months, the cost per death prevented is US$40 if measles vaccine is given as part of a poliomyelitis national immunization day activity. The overall cost per death prevented by a campaign aimed at children aged 9 months to 14 years is still excellent value at around US$166. Update Following the assessment of the burden of measles in the 1980s, global measles vaccine coverage flattened out at around 80 per cent. This overall 102 Global Epidemiology of Infectious Diseases figure conceals many areas and districts with much lower coverage and proportionately higher risks of measles virus circulation. A supreme effort will clearly be needed if coverage is to be raised above this level globally, and especially in some trailing countries. In the late 1990s, when stagnation in coverage levels seemed likely, countries in the Americas stepped up their efforts at measles control by achieving high coverage for routine vaccination, following this up with mass campaigns for children aged 9 months to 14 years. Island nations such as Cuba showed that it was possible to interrupt measles virus transmission in this way. Following this example, mass campaigns aimed at elimination were held in all countries of Latin America. Indigenous measles has now been eliminated from the Americas. In other parts of the world, mass campaigns are now being carried out in many countries and situations. As well as drives towards elimination, campaigns are being conducted in situations where there are epidemiological grounds for thinking that an outbreak is expected. Studies show that mass campaigns performed after an outbreak is under way have virtually no impact and are expensive (Aylward, Clements & Olivé 1997). Preventive campaigns using measles-containing vaccines (such as measles-rubella or measles-mumps-rubella vaccines) have been implemented in, for instance, the United Kingdom with great success. In addition, mass campaigns have been used in high-risk situations such as refugee populations or in the urban environment when intense transmission is anticipated. In more than one hundred countries, vitamin A supplementation has been provided with national immunization days for poliomyelitis eradication. While not yet eliminating vitamin A deficiency, this has undoubtedly reduced the burden of disease from lack of the micronutrient, and it is anticipated that a major impact will be demonstrable over the long term in reducing the mortality and morbidity from measles infection as a result. The latter part of the 1990s saw a dramatic change in the epidemiology and control of measles. The successes of the smallpox and the poliomyelitis eradication initiatives have demonstrated the human capacity for controlling infectious diseases. It will soon become clear whether the global community is prepared to pay the price to eradicate measles from the globe. References Aaby P (1988a) Introduction to community studies of severe measles: comparative test of crowding/exposure hypothesis. Reviews of Infectious Diseases, 10:451. Aaby P (1988b) Malnutrition and overcrowding-exposure in severe measles infection: a review of community studies. Reviews of Infectious Diseases, 10:478–491. Aaby P (1991) Determinants of measles mortality: host or transmission factors. In: de la Maza LM, Medical Virology 10, New York: Plenum, pp. 83–116. Aaby P, Clements CJ (1989) Measles immunization research: a review. Bulletin of the World Health Organization, 67:443–448. Measles 103 Aaby P, Clements CJ (1991) Mortality from measles: measuring the impact. Geneva, World Health Organization (EPI/RD/91/WP.6 APR). Aaby P, Leeuwenburg J (1990) Patterns of transmission and severity of measles infection: a reanalysis of data from the Machakos Area, Kenya. Journal of Infectious Diseases, 161:171–174. Aaby P et al. (1984a) Determinants of measles mortality in a rural area of GuineaBissau: crowding, age and malnutrition. Journal of Tropical Pediatrics, 30: 164–168. Aaby P et al. (1984b) Overcrowding and intensive exposure as determinants of measles mortality. American Journal of Epidemiology, 120:49–63. Aaby P et al. (1988) Further community studies on the role of overcrowding and intensive exposure on measles mortality. Reviews of Infectious Diseases, 10: 474–477. Aaby P et al. (1994) Sex-specific differences in mortality after high-titre measles immunization in rural Senegal. Bulletin of the World Health Organization, 72: 761–770. Adedoyin MA (1990) The pattern of measles in Ilorin. West African Journal of Medicine, 9:103–7. Agarwal DK et al. (1986) Natural history of measles in a rural and urban community of Varanasi. Journal of Communicable Disease, 8:289–298. Almoiradie-Javonillo I, Javonillo TO (1984) Profile of a measles epidemic in a remote Philippine barrio. Journal of the Philippine Medical Association, 60:103–108. Alwar AJE (1992) The effect of protein energy malnutrition on morbidity and mortality due to measles at Kenyatta National Hospital, Nairobi, Kenya. East African Medical Journal, 69:415–418. Arya LS et al. (1987) Spectrum of complications of measles in Afghanistan: a study of 784 cases. Journal of Tropical Medicine and Hygiene, 90:117–122. Aylward RB, Clements CJ, Olivé JM (1997) The impact of immunization control activities on measles outbreaks in middle and low income countries. International Journal of Epidemiology, 26:662–669. Barclay AJG, Foster A, Sommer A (1987) Vitamin A supplements and mortality related to measles: a randomised clinical trial. British Medical Journal, 294: 294–296. Bhaskaram P et al. (1984) Effect of measles on the nutritional status of preschool children. Journal of Tropical Medicine and Hygiene, 87:21–25. Bhatia R (1985) Measles outbreak in village Tophema in Nagaland. Journal of Communicable Disease, 17:185–189. Bhuiya A et al. (1987) Measles case fatality among under-fives: a multivariate analysis of risk factors in a rural area of Bangladesh. Social Science and Medicine, 24:439–443. Burgess W, Mduma B, Josephson GV (1986) Measles in Mbeya, Tanzania— 1981–1983. Journal of Tropical Pediatrics, 32:148–153. Burström B et al. (1992) Severe measles outbreak in western Kenya. East African Medical Journal, 69:419–423. 104 Global Epidemiology of Infectious Diseases Centers for Disease Control (1990) Measles—U.S., 1989 and first 20 weeks 1990. Morbidity and Mortality Weekly Report, 39:353–355, 361–363. Chavan H, Natu M, Pratinidhi A (1986) Outbreak of measles—epidemiological study. Journal of the Indian Medical Association, 84:273–274. Cherian T, Joseph A, John TJ (1984) Low antibody response in infants with measles and children with subclinical measles virus infection. Journal of Tropical Medicine and Hygiene, 87:27–31. Chin J, Thaung UM (1985) The unchanging epidemiology and toll of measles in Burma. Bulletin of the World Health Organization, 63:551–558. Commey JOO, Richardson JE (1984) Measles in Ghana—1973–1982. Annals of Tropical Paediatrics, 4:189–194. Conley SF, Beste DJ, Hoffman RG (1993) Measles-associated bacterial tracheitis. Pediatric Infectious Disease Journal, 12:414–415. Cuauhtli M, Ruiz-Matus C, Neiburg P (1990) Analysis of data from ten measles outbreaks during the 1988–1989 epidemic in Mexico. 12th Meeting of the International Epidemiology Association, Los Angeles, 1990, p. 150. De Buse PJ, Lewis MG, Mugerwa JW (1970) Pulmonary complications of measles in Uganda. Journal of Tropical Pediatrics, 16:197–203. De Sole G, Belay Y, Zegeye B (1987) Vitamin A deficiency in southern Ethiopia. American Journal of Clinical Nutrition, 45:780–784. Dhanoa J, Cowan B (1982) Measles in the community: a study in non-hospitalised young children in Punjab. Journal of Tropical Pediatrics, 28:59–61. Dover AS et al. (1975) Pneumonia associated with measles. Journal of the American Medical Association, 234:612–614. Dutta P et al. (1991) Prospective hospital based study on persistent diarrhoea. Gut, 32:787–790. El Behairy F et al. (1986) Clinico-bacteriological study of diarrhoea complicating measles. Journal of Tropical Pediatrics, 32:44–46. Ellison JB (1932) Intensive vitamin therapy in measles. British Medical Journal, October 15:708–711. [Anonymous] Epidemiological investigation of measles (1985). Indian Communicable Diseases Bulletin, 2:18. Fagbule D, Orifunmishe F (1988) Measles and childhood mortality in semi-urban Nigeria. African Journal of Medical Science, 17:181–185. Feachem RG, Koblinsky MA (1983) Interventions for the control of diarrhoeal disease among young children: measles immunization. Bulletin of the World Health Organization, 61:641–652. Flick JA (1976) Does measles really predispose to tuberculosis? American Review of Respiratory Disease, 114:257–265. Fortenberry JD et al. (1992) Severe laryngotracheobronchitis complicating measles. American Journal of Diseases of Children, 146:1040–1043. Foster A, Sommer A (1987) Corneal ulceration, measles, and childhood blindness in Tanzania. British Journal of Ophthalmology, 71:331–343. Measles 105 Garenne M (1994) Sex differences in measles mortality: a world review. International Journal of Epidemiology, 23:632–642. Garenne M, Aaby P (1990) Pattern of exposure and measles mortality in Senegal, Journal of Infectious Diseases, 161:1088-1094. Gindler JS et al. Measles and the impact of measles vaccine on acute respiratory infection in children. Unpublished. Gordon JE, Jansen AAJ, Ascoli W (1965) Measles in rural Guatemala. Journal of Pediatrics, 66:779–786. Greenberg BL et al. (1991) Measles-associated diarrhoea in hospitalized children in Lima, Peru: pathogenic agents and impact on growth. Journal of Infectious Diseases, 163:495–502. Han AM, Khin MM, Tin A, Hlaing T (1990) Measles-associated diarrhoea in the Infectious Disease Hospital, Rangoon. Journal of Tropical Medicine and Hygiene, 93:205–209. Hancock BD (1972) Tracheostomy in measles laryngo-tracheo-bronchitis. Journal of Laryngology and Otology, 86:23–31. Hayden RJ (1974) The epidemiology and nature of measles in Nairobi before the impact of measles immunization. East African Medical Journal, 51:199–205. Heyworth B (1973) Pathogenesis of measles. British Medical Journal, 3:693. Hondius AJK, Sutorius DM (1979) Severe measles in Nigeria. Utrecht, Stichting WSO. Hull H (1988) Increased measles mortality in households with multiple cases in the Gambia, 1981. Reviews of Infectious Diseases, 10:463–467. Hull HF, Williams PJ, Oldfield F (1983) Measles mortality and vaccine efficacy in rural West Africa. Lancet, 1:972–975. Hussey GD, Klein M (1990) A randomized, controlled trial of Vitamin A in children with severe measles. New England Journal of Medicine, 323:160–164. Hussey G, Simpson J (1990) Nosocomial bacteremias in measles. Pediatric Infectious Disease Journal, 9:715–717. Ibia EO, Asinda AA (1990) Measles in Nigerian children in Calabar during the era of Expanded Programme on Immunization. Tropical Geographical Medicine, 42:226–232. Imperato PJ (1975) A wind in Africa: a story of modern medicine in Mali. St Louis, Warren H. Green. John TJ et al. (1980) Epidemiology and prevention of measles in rural south India. Indian Journal of Medical Research, 72:153–158. Jüptner H, Quinnell C (1965) Epidemiological observations during a measles epidemic in the Trobriand Islands (Papua). Medical Journal of Australia, 1: 538–540. Kambarami RA et al. (1991) Measles epidemic in Harare, Zimbabwe, despite high measles immunization coverage rates. Bulletin of the World Health Organization, 60:213–219. 106 Global Epidemiology of Infectious Diseases Kaschula ROC, Druker J, Kipps A (1983) Late morphological consequences of measles: a lethal and debilitating lung disease among the poor. Reviews of Infectious Diseases, 5:395–404. Kipps A, Kaschula ROC (1976) Virus pneumonia following measles: a virological and histological study of autopsy material. South African Medical Journal, 50: 1083–1088. Kasongo Project Team (1981) Influence of measles vaccination on survival pattern of 7–35 month-old children in Kasongo, Zaire. Lancet, 1:764–767. Koster FT (1988) Mortality among primary and secondary cases of measles in Bangladesh. Reviews of Infectious Diseases, 10:471–473. Koster FT et al. (1981) Synergistic impact of measles and diarrhoea on nutrition and mortality in Bangladesh. Bulletin of the World Health Organization, 59: 901–908. Labay MV et al. (1985) Membranous laryngotracheobronchitis, a complication of measles. Intensive Care Medicine, 11:326–327. Lakhanpal U, Rathore MS (1986) Epidemiology of measles in rural area of Punjab. Journal of Communicable Disease, 18:185–188. Lamabadusuriya SP, Jayantha UK (1992) An outbreak of measles in the Southern Province. Ceylon Medical Journal, 37:46–48. Lamb WH (1988) Epidemic measles in a highly immunized rural West African (Gambian) village. Reviews of Infectious Diseases, 10:457–462. Langaney A, Pison G (1979) Rougeole et augmentation temporaire de la masculinité des naissances: coincidence ou causalité? [Measles and temporary increase in masculine births: coincidence or causality?] Comptes rendus des séances de l’Académie des sciences, 289(Série D):1225–1228. Leeuwenburg J et al. (1984) The epidemiology of measles. In: JK van Ginneken, Muller AS, eds. Maternal and child health in rural Kenya. London, Croom Helm, pp. 77–94. Lindtjørn B (1986) Severe measles in the Gandulla area of South-West Ethiopia. Journal of Tropical Pediatrics, 32:234–239. Lobo J et al. (1987) Epidemiology of measles in a rural community. Indian Journal of Pediatrics, 54:261–265. Markowitz LE, Nieberg P (1991) The burden of acute respiratory tract infection due to measles in developing countries and the potential impact of measles vaccine. Reviews of Infectious Diseases, 13(Suppl 6):S555–561. Markowitz LE et al. (1989) Vitamin A levels and mortality among hospitalised measles patients, Kinshasa, Zaire. Journal of Tropical Pediatrics, 35:109–112. Mata LJ (1978) The children of Santa Maria Cauqué: a prospective field study of health and growth. Cambridge (Massachusetts), MIT Press. McGregor IA (1964) Measles and child mortality in the Gambia. West African Medical Journal, 14:251–257. McIntyre RC et al. (1982) Measles and measles vaccine efficacy in a remote island population. Bulletin of the World Health Organization, 60:767–775. Measles 107 [Anonymous] (1986) Measles outbreak in South Arcot district. Indian Communicable Disease Bulletin, 2:1–6. Miller DL (1964) Frequency of complications of measles. British Medical Journal, 2:75–78. Ministry of Health, Zambia (1993) Infectious Disease Newsletter, No. 3. Morley D (1973) Paediatric priorities in the developing world. London, Butterworth. Morley DC, Martin WJ, Allen I (1967) Measles in West Africa. West African Journal of Medicine, 16:24–31. Morton R, Mee J (1986) Measles pneumonia: lung puncture findings in 56 cases related to chest X-ray changes and clinical features. Annals of Tropical Paediatrics, 6:41–45. Muller AS et al. (1977) The impact of measles in a rural area of Kenya. East African Medical Journal, 54:364–372. Narain JP et al. (1989) Epidemic measles in an isolated unvaccinated population, India. International Journal of Epidemiology, 18:952–958. Obi JO (1979) Measles in hospital practice in Nigeria. Journal of Tropical Pediatrics and Environmental Child Health, 25:30–33. O’Donovan C (1971) Measles in Kenyan children. East African Medical Journal, 48:526–532. Ogbeide MI (1967) Measles in Nigerian children. Journal of Pediatrics, 71:737–741. Oomen HAP, McLaren DS, Escapani H (1964) Epidemiology and public health aspects of hypovitaminosis A: a global survey of xerophthalmia. Tropical and Geographical Medicine, 4:271–315. Orren A et al. (1981) Increased susceptibility to herpes simplex virus infections in children with acute measles. Infection and Immunity, 31:1–6. Osei W, Maganu ET (1987) Measles vaccination in Botswana: does high vaccination coverage ensure protection? 3ème Séminaire International sur les vaccinations en Afrique. Niamey (Niger), 29–31 Janvier 1987. Pereira, SM Benjamin V (1972) Measles in a South Indian community. Tropical and Geographical Medicine, 24:124–129. Pison G (1982) Dynamique d’une population traditionelle: les Peul Bande (Sénégal oriental). Institut national d’études démographiques. Cahier no. 99. Paris, Presses Universitaires de France. Pison G, Bonneuil N (1988) Increased risk of measles mortality for children with siblings among the Fula Bande, Senegal. Reviews of Infectious Diseases, 10:468–470. Puffer RR, Serano CV (1973) Measles. In: Patterns of mortality in childhood: report of the inter-American investigation of mortality in childhood. Washington, DC, Pan American Health Organization (Scientific Publication No. 262). Raote GJ, Bhave SY (1992) Clinical profile of measles: a prospective study of 150 hospital based children. Indian Pediatrics, 29:45–48. Rea JN (1968) Measles in Africa. Lancet, 1:356. 108 Global Epidemiology of Infectious Diseases Reddy V et al. (1986) Relationship between measles, malnutrition, and blindness: a prospective study in Indian children. American Journal of Clinical Nutrition, 44:924–930. Riley ID et al. (1986) Pneumococcal vaccine prevents death from acute lower-respiratory-tract infections in Papua New Guinean children. Lancet, 2:877–881. Ristori C et al. (1962) Medical importance of measles in Chile. American Journal of Diseases of Children, 103:6–71. Rolfe M. (1982) Measles immunization in the Zambian copperbelt: cause for concern. Transactions of the Royal Society of Tropical Medicine and Hygiene, 76:529–530. Ross LA et al. (1992) Laryngotracheobronchitis as a complication of measles during an urban epidemic. Journal of Pediatrics, 121:511–515. Salunke SR, Natu M (1977) Epidemiological investigations of measles: an outbreak in Ajiwali. Indian Pediatrics, 14:519–521. Samsi TK et al. (1992) Risk factors for severe measles. Southeast Asian Journal of Tropical Medicine and Public Health, 23:497–503. Sánchez-Vargas MC, López-Ortiz O, Bustamante-Hernández A (1990) Brote de sarampión en la comunidad indigena de Yaitepec, Oaxaca 1988 [Measles outbreak in the Indian community of Yaitepec, Oaxaca 1988]. Boletin Médico del Hospital Infantil de México, 47:489–494. Sang FC et al. (1992) E. coli associated with acute measles and diarrhoea at Kenyatta National Hospital, Kenya. East African Medical Journal, 69:135–139. Satge P et al. (1965) Etude d’une épidémie de rougeole en milieu rural au Sénégal. Bulletin de la société médicale d’Afrique noire de langue française, Mars:83–87. Satpathy SK, Chakraborty AK (1990) Epidemiological study of measles in Singur, West Bengal. Journal of Communicable Disease, 22:23–26. Sharma RS et al. (1984) An epidemiological investigation of measles outbreak in Alwar, Rajasthan. Journal of Communicable Disease, 16:299–303. Siddiqi N, Ghosh S, Berry AM (1974) The natural history of measles in a low-income urban community in South Delhi. Indian Pediatrics, 11:557–562. Sinha DP (1977) Measles and malnutrition in a West Bengal village. Tropical and Geographical Medicine, 29:125–34. Solon FS et al. (1978) Vitamin A deficiency in the Phillipines: a study of xerophthalmia in Cebu. American Journal of Clinical Nutrition, 31:360–8. Sommer A, Katz J, Tarwotjo I (1984) Increased risk of respiratory disease and diarrhoea in children with pre-existing mild vitamin A deficiency. American Journal of Clinical Nutrition, 40:1090–1095. Spencer HC et al. (1987) Impact on mortality and fertility of a community-based malaria control programme in Saradidi, Kenya, Annals of Tropical Medicine and Parasitology, 81 (Suppl. 1):36–45. Spika JS et al. (1989) Acute lower respiratory infections: a major cause of death in children in Bangladesh. Annals of Tropical Paediatrics, 9:33–39. Stanton BF et al. (1986) Risk factors for developing mild nutritional blindness in urban Bangladesh. American Journal of Diseases of Children, 140:584–588. Measles 109 Stephens PW (1986) Reliability of lay-reporting of morbidity and cause-of-death data: an evaluation of reported cases and deaths from measles in rural Senegal. International Union for the Scientific Study of Population, Seminar on comparative studies of mortality and morbidity: old and new approaches to measurement and analysis, Siena, 1986. Sutter RW et al. (1991) Measles among the Amish: a comparative study of measles severity in primary and secondary cases in households. Journal of Infectious Diseases, 163:12–16. Swami SS et al. (1987) Epidemiology of measles in western Rajasthan. Journal of Communicable Disease, 19:370–372. Sy I (1967) Etude de la rougeole en milieu rural Sénégalais [thesis]. Dakar, Faculté Mixte de Médicine et de Pharmacie. Taylor WR et al. (1988) Measles control efforts in urban Africa complicated by high incidence of measles in the first year of life. American Journal of Epidemiology, 127:788–794. Tidstrom B (1968) Complications in measles with special reference to encephalitis. Acta Medica Scandinavica, 184:411–415. Tielsch JM et al. (1986) Prevalence and severity of xerophthalmia in southern Malawi. American Journal of Epidemiology, 124:561–568. Varavithya W et al. (1985) Diarrhoea in measles. Journal of the Medical Association of Thailand, 68:298–302. Varavithya W et al. (1989) Etiology of diarrhoea in measles. Journal of the Medical Association of Thailand, 72(Suppl 1):151–154. Velhaf GD et al. (1989) Measles survey in peri-urban slum of Bombay. Commununicable Diseases Bulletin, 6:14–16. Wahab MMA et al. (1988) Measles in Equatorial region: south Sudan. Annals of Tropical Paediatrics, 8:31–34. Wakeman PF (1978) Severe measles in Afghanistan. Journal of Tropical Pediatrics and Environmental Child Health, 24:87–88. Wesley AG, Sutton JB, Widrich AJ (1971) The aetiology of pneumonia associated with measles in Bantu children. South African Medical Journal, 45:1402–1404. World Bank (1993) World development report 1993: investing in health. New York, Oxford University Press. World Health Organization (1979a) Measles epidemiology. Weekly Epidemiological Record, 54:73–76. World Health Organization (1979b) Measles surveillance: United Republic of Cameroon. Weekly Epidemiological Record, 54:273. World Health Organization (1980) Epidemiology of measles in a rural community. Weekly Epidemiological Record, 55:85–87. World Health Organization (1982a) Measles in Indonesia. Geneva, (EPI Library 250.03/82/012). World Health Organization (1985a) Measles outbreak among hill tribes. Weekly Epidemiological Record, 60:79. 110 Global Epidemiology of Infectious Diseases World Health Organization (1985b) Public health importance of measles. Weekly Epidemiological Record, 60:95–97. World Health Organization (1986) Public health importance of measles. Weekly Epidemiological Record, 61:89–90. World Health Organization (1993a) Generic protocol for determining measles case fatality rates in a community, either during an epidemic or in a highly endemic area. Geneva (WHO/EPI/GEN/93.3). World Health Organization (1993b) High measles case-fatality rates during an outbreak in a rural area (Niger). Weekly Epidemiological Record, 68:142–145. World Health Organization Expanded Programme on Immunization (1996). EPI information system: global summary. Geneva (WHO/EPI/CEIS/96.07) Chapter 5 Poliomyelitis Kenji Shibuya, Christopher JL Murray Introduction Poliovirus is an enterovirus in the family of Picornaviridae and consists of three antigenic types, 1, 2, and 3. All three types cause paralysis and the poliomyelitis vaccine is intended to be antigenic for all of these. Poliovirus type 1 causes paralysis most often and is responsible for most epidemics in unvaccinated populations (Assaad & Ljungars-Esteves 1984, Sen, Sharma & Sarthanam 1985, Kadi et al. 1988, Khare et al. 1991). Some outbreaks due to poliovirus type 1 have been reported among unimmunized populations in regions with high vaccine coverage (Kim-Farley et al. 1984, Sutter et al. 1991, Otten et al. 1992). In contrast, the poliomyelitis cases in vaccinated populations are usually caused by types 2 and 3 (Assaad & Ljungars-Esteves 1984). Immunity to poliovirus is lifelong. Poliovirus infections occur only in humans and there is no animal reservoir. This is one of the key characteristics for possible eradication of poliomyelitis. Spread of infection by faecal–oral routes is most common in developing countries with poor hygiene and sanitation, while oral–oral (respiratory) infection may be more common in industrialized countries and during outbreaks. Perinatal transmission from mothers to newborn infants also occurs. After oral ingestion of poliovirus, the virus first multiplies in the lymph nodes in the pharynx and gastrointestinal tract. The incubation period is 4 to 35 days, most cases occurring 1–3 weeks after exposure. Once the virus enters the human body, it invades local lymphoid tissues, enters the bloodstream, and then may invade anterior horn cells of the spinal cord. As it multiplies intracellularly, poliovirus may cause temporary or permanent damage of the nerve cells due to the inflammation process. Other areas of both the spinal cord and brain may be affected as well. Most poliovirus infections are asymptotic. Non-septic illness with lowgrade fever and sore throat (minor illness) occurs in 4 to 8 per cent of infections. Aseptic meningitis occurs in 1 to 5 per cent of patients a few 112 Global Epidemiology of Infectious Diseases days after the minor illness has resolved. Severe illness with rapid onset of asymmetric, acute flaccid paralysis (AFP) resulting from damage of lower motor neurons, with areflexia of the involved limb and without loss of sensation occurs in 0.1 to 2 per cent of infections. Bulbar paralysis and paralysis of respiratory muscles may occur. Incidence of residual paralytic disease ranges from 1 per 1000 to 1 per cent of infections, and thus most cases are asymptomatic. Although cases are rarely recognized (Ledinko 1951, Debre & Thieffery 1955, Melnick & Nathanson & Martin 1979), incidence is known to be higher among older patients (Halstead et al. 1985). Despite the fact that, of those who survive an acute attack, most will recover some degree of function in the affected muscles, the true burden of paralytic poliomyelitis is that the affected patients may suffer from disability and handicap for the rest of their life and only a minority will recover completely. Patients with a history of paralytic poliomyelitis, in particular those who were older at the onset of polio infection and had an initially severe presentation of poliomyelitis, may present years later with complaints of muscle pain and increased weakness in the previously affected limb: late effects or post-paralytic-poliomyelitis syndrome (Halstead, Wiechers & Rossi 1985). Although it has a good prognosis, it may alter the sequelae of the disease burden. It is well known that there are temporal and epidemiological changes in the pattern of transmission of poliovirus (Anderson & May 1993, Horstmann 1982, Melnick 1990, Nathanson & Martin 1979, Paul et al. 1952). Melnick (1990) has suggested that poliomyelitis can be viewed as having three major phases: endemic, epidemic, and vaccine-era. Because of poor sanitation and hygiene in most developing countries, polioviruses circulate freely among infants and young children. Therefore, almost all children develop antibodies to at least one of the three poliovirus types before 5 years of age. Consequently, clinical poliomyelitis in tropical developing countries is almost exclusively an infantile disease, with more than 90 per cent of cases occurring under the age of 5 years because of age-dependent acquisition of immunity (Babaniyi & Parakoyi 1991, Heymann et al. 1983, Mahadevan et al. 1989, Ofosu-Amaah, Kratzer & Nicholas 1977, Onadeko & Familusi 1990, Prabhakar et al. 1981, Sabin & Silva 1983, Thuriaux 1982). As hygiene and sanitation improve, paralytic poliomyelitis will become epidemic and affect young adolescents and the older population, where it is more likely to cause severe illness (Anderson & May 1993, Nathanson & Martin 1979, Paul 1955). It has been shown that the age distribution of reported cases during 1944–1950 in England and Wales moved towards an older age group, with 30 per cent of reported cases occurring over 15 years of age (Table 5.1) (Freyche & Nielsen 1955). A similar age shift in the incidence of poliomyelitis was observed in a review of cases in the United States from 1969 to 1981 (Moore et al. 1982). These two phases sometimes coexist and are seen in developing countries where vaccine coverage has not been adequate. Poliomyelitis 113 Table 5.1 Percentage distribution of notified cases of poliomyelitis by age-group and median age of notified cases: England and Wales, 1912–19 and 1944–50 Year Sex 0–5 5–14 15 and over Total (per cent) Median age (years) 1912–19 Male Female 64 66 28 28 8 6 100 100 3.93 3.90 1944–50 Male Female 34 33 37 33 29 34 100 100 8.46 9.15 Percentage distribution by age-group (years) Review of past attempts at measuring the burden of poliomyelitis Paralytic poliomyelitis causes significant lifelong disability and handicap in affected persons. Since the early 1970s, lameness surveys to determine the prevalence of residual paralysis resulting from poliomyelitis have been carried out, particularly in developing regions (a detailed list of lameness surveys and their poliomyelitis prevalences is available from the authors on request). As a consequence of high immunization coverage and specific programmes to eradicate poliomyelitis, the global incidence of poliomyelitis has fallen dramatically since the initiation of the poliomyelitis eradication initiative in 1988 (Figures 5.1 and 5.2). It is suggested, however, that under the current reporting system, actual prevalence or incidence of poliomyelitis in the world is still underreported, particularly in developing countries (Tangermann et al. 1997, World Health Organization 1994). The reported cases in the African Region in 1990 was initially only 572, but was revised and given as 4408 cases in the official figure. WHO estimated that the actual incidence of poliomyelitis in the world was approximately Figure 5.1 Global immunization coverage (per cent) of OPV3 for children in the first year of life 90 80 Percentage 70 60 50 40 30 20 10 0 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year 114 Global Epidemiology of Infectious Diseases Figure 5.2 Reported incidence of indigenous poliomyelitis 70 60 Percentage 50 40 30 20 10 0 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year 148 000 in 1990, based on the assumption of a vaccine efficacy of 80 per cent and a paralysis rate of 5 per 1000 infections (Expanded Programme on Immunization 1993). Similarly, although a total of 7898 cases of poliomyelitis were reported in 1993, WHO estimated that 110 000 cases of poliomyelitis occurred, suggesting that reporting efficiency was less than 10 per cent (World Health Organization 1994). It is well known that indirect estimates of incidence derived from lameness surveys have suggested a much heavier disease burden than reported cases would suggest. There is a considerable difference between low incidence of reported cases and the relatively high incidence derived from lameness surveys (Table 5.2) (Evans 1984, Melnick 1990). The prevalence of paralytic poliomyelitis in lameness surveys ranges from less than 1 per 1000 to more than 20 per 1000 and is 3–4 cases per 1000 on average among Table 5.2 Annual incidence of poliomyelitis per 100 000 total population during the period 1976–1980 Country Burma Cameroon Côte d’Ivoire Egypt Ghana India Indonesia Malawi Thailand Yemen Reported Incidence Estimate from lameness survey 1.1 1.2 1.2 1.8 2.3 2.1 0.1 1.2 1.7 3.3 18 24 34 7 31 18 13 28 7 14 Poliomyelitis 115 school-age children in most developing countries (Bernier 1984). There is some variation within countries (urban or rural) and between regions. A review of surveys reveals considerable variation in both the choice and use of survey methods and in the assumption made in the analysis and interpretation of findings (Bernier 1984). The true magnitude of the burden of poliomyelitis in 1990, just before the wide dissemination of the key strategies for global poliomyelitis, was thus not well documented. Definition and Measurement The incidence of poliomyelitis can be estimated in a variety of ways: surveillance of clinical diagnoses, such as diagnosis of acute flaccid paralysis (AFP); laboratory confirmation of poliovirus isolated from cases; indirect estimation through lameness surveys of prevalence; and serosurveys for poliomyelitis neutralising antibodies (Pan American Health Organization 1993). A definitive diagnosis is made by laboratory confirmation. Virus isolation from stool culture is the most reliable, and is thus a standard for diagnosis of poliomyelitis incidence. At the population level, poliomyelitis incidence has been estimated either by lameness surveys to identify cases with residual paralysis or by AFP surveillance of acute cases. AFP surveillance has been proposed as one of the key strategies for the eradication of poliomyelitis and implemented more widely since the early 1990s (Hull et al. 1994). Before the use of AFP surveillance became widespread, indirect estimation of poliomyelitis incidence from lameness surveys was more common (Bernier 1983). In serosurveys for poliomyelitis, antibodies are of little use for estimating incidence since it is almost impossible to distinguish antibodies formed in response to vaccination from those formed in response to wild poliovirus (Pan American Health Organization 1993). Surveillance and reporting by health care providers is most often used in many countries, but the standardized and systematic surveillance of poliomyelitis was not feasible until the 1990s in most developing countries with relatively poor information systems, with the exception of Latin America and the Caribbean (de Quadros et al. 1997, Pan American Health Organization 1993). It is suggested that, in the early 1990s, many cases of paralytic poliomyelitis were not reported (Robertson et al. 1990, Ward et al. 1993, World Health Organization 1994). The process of developing a highly efficient surveillance system capable of detecting all cases of poliomyelitis is one of the critical steps leading to eradication. The World Health Organization has recommended using standardized case definitions for active surveillance of AFP (Andrus et al. 1992, Biellik et al. 1992, de Quadros et al. 1991, Hull et al. 1994, Pan American Health Organization 1988). Whether the surveillance is effective or not depends largely upon the sensitivity and specificity of the case definition. In the early years of efforts to eradicate the poliomyelitis when wild poliovirus was prevalent, AFP cases were “confirmed” as poliomyelitis if they met one of the following 116 Global Epidemiology of Infectious Diseases criteria: laboratory confirmation, epidemiological linkage to another case of AFP or to a confirmed case of poliomyelitis, residual paralysis 60 days after onset of symptoms, death or lack of appropriate investigation of a reported AFP case (de Quadros et al. 1997, Pan American Health Organization 1993). Therefore, AFP surveillance to detect paralytic poliomyelitis had relatively high sensitivity but its specificity was low, resulting in a large number of false positives. In 1990, of 2497 cases of AFP that were reported, only 18 had wild poliovirus isolates from the stool (Dietz et al. 1992). In the AFP surveillance in China during the early 1990s, Chiba et al (1992) showed that 25 of 130 AFP cases were actually poliomyelitis, which 40 per cent of cases were attributable to either Guillain-Barré syndrome or transverse myelitis. In contrast, it has been suggested that, among children with AFP, the use of criteria such as age under 6 years and either presence of fever at the onset of paralysis or a less than four-day period for complete development of paralysis, result in a sensitivity of 96 per cent and an increase in specificity to 49 per cent (Andrus et al. 1992, Dietz et al. 1992). Even though data may be comparable and internally consistent, a relatively small number of reported cases might be attributable to biases in selection and observation attributable to qualification of health personnel and other social, cultural and economic factors. As the programme for the eradication of poliomyelitis progresses, with increasing vaccine coverage and reported AFP cases, the criteria for confirmation have become more specific (de Quadros et al. 1997). In 1990, the definition of confirmed cases was modified and the criteria for final classification were shifted from a clinical classification scheme to greater reliance on the virological investigation (World Health Organization 1998). Accordingly, an AFP case is confirmed only if it is associated with wild poliovirus isolation from either the patient’s stool or a specimen obtained from a contact of the patient. Under this classification system, AFP cases that resulted in residual paralysis or death or that could not be fully investigated or neither be confirmed nor discarded are classified as “compatible” cases with poliomyelitis. Such cases indicate a failure of the AFP surveillance system to collect adequate specimens in a timely manner (de Quadros et al. 1997). Following successful use in Latin America and the Caribbean, the experiences and the success in the LAC Region, AFP surveillance emerged as one of the key strategies for polio eradication and was more widely used in developing countries during the 1990s (Pan American Health Organization 1988, 1994). The efficiency of surveillance is, however, higher in countries with a low incidence of poliomyelitis, and lower in countries with high levels of this disease. This was true, in particular during the early years of AFP surveillance around 1990. Thus, data on incidence of AFP cases are inadequate for the estimation of incidence of poliomyelitis in 1990 in established market economies and in Latin America and the Caribbean (Birmingham et al. 1997b, Centers for Disease Control and Prevention 1994). Before the introduction of AFP surveillance, indirect estimation of Poliomyelitis 117 incident cases from lameness surveys was more widely used, particularly in the endemic regions. The case definition used in the lameness surveys includes: acute onset of the disease; no progression; intact sensation; atrophy of muscle; and flaccid paralysis (Singh & Kumar 1991). In paralytic poliomyelitis, muscles in the proximal lower limbs are usually involved. Clinical data from several developing countries, where more than 95 per cent of poliomyelitis cases occurred in children under 5 years of age, showed evidence of involvement of the lower limbs in 78 to 86 per cent of cases (LaForce et al. 1980). La Force and colleagues (1980) have suggested that any prevalence study of poliomyelitis in developing countries that uses a lameness survey as a criterion for diagnosis would identify approximately 80 per cent of cases and the total prevalence of residual paralysis attributable to poliomyelitis could be estimated by multiplying the prevalence by 1.25 or Total cases 100 = Lower limb cases 80 . In contrast, studies in the established market economies have shown a slightly higher rate of residual paralysis because of the difference in agecomposition of cases (Nathanson & Martin 1979). For example, 203 paralytic poliomyelitis cases were reported from 1969 to 1981 in the United States, and lower limb involvement was present in 180 patients (93 per cent) among 193 patients for whom the site of spinal paralysis was specified (Moore et al. 1982). To estimate incidence from prevalence in a lameness survey, it must be assumed that the region is in a stable, endemic condition. Under this assumption, and if, as in most developing countries, most cases occur in children under 5 years of age, the paralysis attributable to poliomyelitis in the 5–9 year age group may reflect the residual paralytic cases which occurred in the 0–4 year age group from 1 to 5 years ago. The following simple formula can be used to estimate the incidence of paralytic poliomyelitis from 1 to 5 years of age: I= P5 − 9 × 1.2 25 5 where I is the annual incidence of paralytic poliomyelitis and P5–9 is the prevalence of lameness in the 5–9 year age group (LaForce et al. 1980). Some studies show that majority of cases occur before 4 years of age. Therefore, the denominator here should be altered according to the epidemiological pattern of the region (Bernier 1984). This case definition assumes that the sequelae of residual paralytic poliomyelitis are observable and distinctive enough to be reliably attributed to poliomyelitis. But, the assumption may not hold, since it is difficulty to make an accurate etiological diagnosis long after the onset of illness (Pan American Health Organization 1993). The problems of sensitivity and specificity of case 118 Global Epidemiology of Infectious Diseases definition and sample size of a lameness survey still remain as in AFP surveillance. First, there are a number of possible outcomes after poliomyelitis infection. These outcomes are important since it is only the child with residual paralysis who can be identified at a later date as having had poliomyelitis. According to the study conducted in Greater Bombay, India, from 1971 to 1975 (Table 5.3), 74 per cent of the cases showed residual effects when assessed 2 months after the initial onset of paralysis and 26 per cent of the cases, including deaths and patients who recovered completely, would not have been detected in a later clinical survey (Chodankar & Dave 1979, LaForce et al. 1980). Thus, prevalence surveys of residual paralysis due to poliomyelitis will always underestimate the true incidence of the disease, ignoring fatal cases and complete recovery cases. A rough estimate of all clinical cases of poliomyelitis could be made by multiplying the prevalence rate of residual paralysis attributable to poliomyelitis by 1.33 (LaForce et al. 1980). Second, schoolchildren have often been selected as the sample for lameness surveys because of the feasibility of such studies in terms of cost and time (Belcher et al. 1979, Joseph et al. 1983). The population of schoolage children in the catchment area might, however, not be representative of the population at risk since not all school-age children attend school, particularly lame children, children living in remote rural areas, or female children in certain societies. Consequently, it is difficult to generalize the results to other populations. As demonstrated in the surveys in several developing countries (Belcher et al. 1979, Heymann et al. 1983, Heymann 1984), the sensitivity of school-based surveys is not always as high as that of community-based surveys with house-to-house case finding and may lead to underestimation of the true magnitude of the burden of poliomyelitis. House-to-house surveys, although expensive and time-consuming, are a sensitive method of identifying lame children in both urban and rural regions. School surveys and reviews of hospital and clinic registers, while equally sensitive in urban regions, are less sensitive in rural regions and may significantly underestimate the annual incidence of paralytic poliomyelitis (Heymann et al. 1983). Moreover, as mentioned above, the prevalence of lameness attributable to poliomyelitis is always an underestimation of the true total morbidity and mortality associated with poliomyelitis. Finally, the sensitivity of lameness survey varies significantly among Table 5.3 Poliomyelitis cases in Greater Bombay: clinical outcome Outcome Number Percentage No recovery Partial recovery Complete recovery Death 342 1 895 506 255 11 63 17 9 Total 2 898 100 Poliomyelitis 119 surveys and is often lower than that of AFP surveillance, ranging from 34 per cent in Bangladesh to 86 per cent in Madagascar (Snyder et al. 1981, World Health Organization 1982, Bernier 1984). For example, a survey in Ghana showed that poliomyelitis was the leading cause of lameness (63.3 per cent) and motor neuron disorders of the central nervous system, such as cerebral palsy; and encephalitis was second (11.7 per cent) (Nicholas et al. 1977). In most lameness surveys, the specificity of the survey has not been fully evaluated. Box 5.1 lists the major limitations encountered in lameness surveys, which may lead to biases in the estimation of annual incidence of paralytic poliomyelitis. International Classification of Diseases (ICD) codes for poliomyelitis ICD-9 codes for poliomyelitis In the basic tabulation of the ICD-9, acute poliomyelitis (045) is classified in the category of poliomyelitis and other non-arthropod-borne viral diseases of the central nervous system (045–049), and its four-digit codes are defined based on the clinical and anatomical diagnosis of poliomyelitis. It should be noted that the code 045 excludes the late effects of acute poliomyelitis, which are classified in one of the additional codes for late effects of infectious and parasitic diseases (137–139). Encephalitis attributBox 5.1 Limitations of lameness surveys Unreliable case findings because of: lack of sensitivity or specificity of case definition inter-observer variations Restricted study population: school catchment area house-to-house survey Sample size considerations Potential biases: change of epidemiological patterns in the past extent of vaccine coverage change in population Inadequate information concerning: age distribution sex differences baseline mortality rate remission rate case-fatality rate vaccine coverage and efficacy Source: Heymann (1984) 120 Global Epidemiology of Infectious Diseases able to poliovirus infection is also listed under encephalitis, myelitis and encephalomyelitis (323.2*). The ICD-9 codes are: 045 Acute poliomyelitis (323.2*) 045.0 Acute poliomyelitis specified as bulbar (323.2*) Infantile paralysis (acute) specified as bulbar Poliomyelitis (acute) (anterior) Polioencephalitis (acute) (bulbar) Polioencephalomyelitis (acute) (anterior) (bulbar) 045.1 Acute poliomyelitis with other paralysis (323.2*) Paralysis: – acute atrophic, spinal – infantile, paralytic Poliomyelitis (acute) with paralysis except bulbar – anterior – epidemic 045.2 Acute nonparalytic poliomyelitis (323.2*) Poliomyelitis (acute) specified as nonparalytic – anterior – epidemic 045.9 Acute poliomyelitis, unspecified (323.2*) –Infantile paralysis Poliomyelitis (acute) unspecified whether – anterior paralytic or nonparalytic – epidemic 138 Late effects of acute poliomyelitis The late effects include conditions specified as such, or as sequelae, or as attributable to old or inactive poliomyelitis, without evidence of active disease. ICD-9 codes consistent with acute flaccid paralysis Acute flaccid paralysis (AFP) surveillance has become a crucial strategy since the late 1980s. The fact that AFP includes various types of diseases sometimes reduces the specificity of AFP surveillance for poliomyelitis cases. The possible ICD-9 codes for AFP other than poliomyelitis include: 094.1 323.0–323.9 335.2 342.0 344.0–344.9 355.0–355.9 357.0 General paresis Encephalitis, myelitis and encephalomyelitis Motor neurone disease Flaccid hemiplegia Arm or leg paralytic syndromes Mononeuritis of lower limb Acute infective polyneuritis (Guillain-Barré Syndrome) Poliomyelitis 781.2 781.4 121 Gait disorders Transient paralysis of limb ICD-10 codes for poliomyelitis In accordance with the rapid decline in global incidence of poliomyelitis and the substantial progress toward its eradication, the emphasis of the classification scheme in the ICD-10 for poliomyelitis (A80) shifted from the clinical and anatomical diagnosis in ICD-9 to the source of poliovirus infection. The late effects of poliomyelitis are now classified in the category of sequelae of poliomyelitis (B91). The ICD-10 codes are as follows: A80 Acute poliomyelitis A80.0 Acute paralytic poliomyelitis, vaccine-associated A80.1 Acute paralytic poliomyelitis, wild virus, imported A80.2 Acute paralytic poliomyelitis, wild virus, indigenous A80.3 Acute paralytic poliomyelitis, other and unspecified A80.4 Acute nonparalytic poliomyelitis A80.9 Acute poliomyelitis, unspecified B91 Sequelae of poliomyelitis Estimation of Disability-Adjusted Life Years (DALYs) To calculate the burden of disease attributable to poliomyelitis requires detailed estimates of the age-specific and sex-specific epidemiology of poliomyelitis (Murray & Lopez 1996). Valid community-based epidemiological studies that could inform these estimates do not exist for many of the variables in many regions. For this reason, the burden of paralytic poliomyelitis was calculated by age, sex, and region from various sources of available data. Incidence Annual incidence rates of paralytic poliomyelitis were estimated from information on prevalence of residual paralysis, since comparable incidence data based on acute flaccid paralysis (AFP) surveillance were not available before 1990, except from the Latin America and Caribbean (LAC) region. A significant amount of information on the prevalence of residual paralysis has been reported to WHO and is available for India (IND), Other Asia and Islands (OAI), and sub-Saharan Africa (SSA) regions. To estimate annual incidence of paralytic poliomyelitis in 1990 in these regions, the most recent and representative house-to-house lameness surveys, or schoolbased surveys in some cases, were selected. China, where there have been several epidemics in the southern provinces, has improved its reporting system of the AFP cases; but AFP surveillance before 1990 was incomplete and the assessment of residual paralysis was not reported. In these 122 Global Epidemiology of Infectious Diseases circumstances, the estimate of prevalence in China in 1990 was based on regions that have similar epidemiological profiles, including neighbouring OAI countries where poliomyelitis was endemic. For other regions, especially those with lower incidence, such as Established Market Economies (EME), Former Socialist Economies of Europe (FSE) and LAC regions, lameness surveys have not been conducted recently because of the low incidence of residual paralysis and because AFP surveillance of sporadic cases is more important for the eradication process. The estimation of incidence in EME and FSE was primarily based on the number of poliomyelitis cases reported to the World Health Organization. These reported cases included a sporadic outbreak in EME among the unvaccinated population and a slight increase in paralytic poliomyelitis cases in FSE because of social instability and vaccine shortage, although the number of such cases was small. For the LAC region, data both from lameness surveys in the 1980s and AFP surveillance around 1990 can be used to estimate the annual incidence in 1990. Since our main focus is the residual paralysis attributable to poliomyelitis, and there was a large discrepancy between the relatively high incidence of paralytic poliomyelitis and the low numbers of reported cases, lameness surveys are a more appropriate method to estimate the burden in 1990 than the reported cases of AFP. Since lameness surveys are most often aggregated for the children aged 5 to 9 years, age-specific- and sex-specific estimation of annual incidence for developing regions should be made by incorporating information on sex differences and age distributions. Some epidemiological studies have reported the male-to-female ratio in prevalence of paralytic poliomyelitis. In general, the ratio varies significantly from one study to another, even in the same region. Since such variations could be attributable to potential biases or chance alone, we apply a ratio of 1.4 to 1 of males to females, which is an average ratio in the selected studies when excluding extreme values (Jamison et al. 1993). In developing regions, the majority of poliomyelitis cases occurred among children under 5 years of age, following the classical endemic pattern poliomyelitis infection ( Nicholas et al. 1977, Nathanson & Martin 1979, LaForce et al. 1980, Thuriaux 1982, Heymann et al. 1983, Khajuria et al. 1989, Onadeko & Familusi 1990, Mandke et al. 1991) from the prevalence data on residual paralysis of the specific age-group studied. The age distribution of incidence was estimated. First, using the formula proposed by LaForce et al. (1980), the prevalence of lameness in the age group 5–9 years, obtained from relatively large house-to-house surveys in the 1980s, was converted to the incidence rates of the age group 0–4 in the study years. Since the age-patterns of incidence by age group were not available in all regions, where necessary all the rates for a region were estimated from other regions with similar epidemiological patterns by adjusting for the vaccine coverage, as described below. Second, the agespecific incidence rates of paralytic poliomyelitis were assumed to follow Poliomyelitis 123 a polynomial distribution function which fitted well the endemic pattern of poliomyelitis in developing regions ( Paul et al. 1952, Nathanson & Martin 1979, Anderson & May 1993). The fitted polynomial function for developing regions has the following distribution fraction by age group: 97 per cent for age 0–4 years; 5.4 per cent for age 5–14 years; 1.6 per cent for age 15–44 years; 0.8 per cent for age 45–59 years; and 0.06 per cent for age 60 years and over. To estimate age-specific incidence for each age group in developing regions, the estimated incidence of age 0–4 was then multiplied by the factor derived from the ratio of the distribution fraction of each age group to that of age 0–4: 1.0000 for age 0–4; 0.056 for 5–14; 0.0165 for 15–44; 0.0082 for 45–59; 0.0006 for 60 and over. The estimated age-specific and sex-specific incidence rates in the reference years of lameness surveys in developing regions, which were used to compute internally consistent epidemiological parameters, are described in the following section. The estimated age- and sex-specific incidence will thus pertain to a specific time period with a particular level of coverage with the vaccine (Table 5.4). To derive incidence in 1990 from data in the study years of lameness surveys, adjustment for the change in vaccine coverage over the previous decade was made using the exponential change in incidence rate for each year. Suppose we want to estimate the incidence in year (tt + n) but we only have the incidence of year (t) and vaccine coverage of both years. Then the rate of change (r) can be calculated using the formula: 100 − Ct + n r = ln 100 − Ct where Ct and Ct+n are the percentages of vaccine coverage in years t and (tt + n), respectively. Then, using this rate, the incidence in year (tt + n) is written in the form: It + n = I ne rn where It+n is the incidence in year (tt + n) and In is the incidence in year t. Table 5.4 Change in OPV coverage (percentage) by region Region 1981 1985 1989 1990 CHI IND OAI SSA FSE EME LAC MEC 90 7 35 21 95 83 45 45 NA 35 39 30 88 88 70 60 95 82 73 58 94 88 82 87 98 93 86 69 93 85 87 88 Global 53 59 82 87 124 Global Epidemiology of Infectious Diseases Finally, internal consistency between incidence, remission, case-fatality rates, duration of paralysis and prevalence estimates was ascertained by the DisMod software specifically developed for the Global Burden of Disease Study (Murray & Lopez 1996). For incidence data on paralytic poliomyelitis estimated from lameness surveys, we assume that remission and case-fatality rates of those with residual paralysis are zero because deaths and patients with complete recovery would not have been detected in later clinical surveys. Mortality To estimate the burden of premature deaths from paralytic poliomyelitis, both vital registration data and an epidemiological model based on casefatality rates were employed. For the regions where a vital registration system is established (i.e. EME, FSE and LAC), the reported number of deaths attributable due to poliomyelitis was used. For most developing regions where vital data are not available or incomplete (China, India, MEC, OAI, and SSA), mortality was estimated from incidence of paralytic poliomyelitis, multiplied by the correction factor of 1.33 for total cases (LaForce et al. 1980), and case-fatality rates. There are some variations in the reported case-fatality rates of poliomyelitis among studies, but most very from 5 per cent to 15 per cent ( Freyche and Nielsen 1955, Chodankar & Dave 1979, LaForce et al. 1980, Moore et al. 1982, Hajar et al. 1983, Mahadevan et al. 1989, Khare et al. 1991, Jamison et al. 1993). In general, mortality data are influenced by the extent and completeness of notification and registration of deaths. The differences in study years, treatment facilities and results, and the time interval between onset of the disease and death in fatal cases may explain these variations as well as true differences in case-fatality rates. Since age-specific and sex-specific mortality data were not available, and since, as reported in most studies, most cases occur under 5 years of age in developing regions, for the main analysis we used a case-fatality rate of 10 per cent as a plausible value in developing regions and employed a constant case-fatality rate for all age groups. Case-fatality rates may be affected by the age composition of the study population because poliovirus infection in older people is more likely to produce serious illness. Case-fatality rates in older patients may be higher than those of young children in developed regions, but the bias would be minimal in developing regions since the majority of cases and deaths occur among children under 5 years of age, and samples of the epidemiological studies on poliomyelitis in developing regions most often include children aged under 15 years (Basu & Sokhey 1984, Chodankar & Dave 1979, Hajar et al. 1983, Khare et al. 1991, LaForce et al. 1980, Mahadevan et al. 1989, Maru et al. 1988). Disability Weights In the calculation of disability-adjusted life years (DALYs) of poliomyelitis, the disability weights were derived from studies on the distribution of dis- Poliomyelitis 125 abilities associated with poliomyelitis in developing regions that report on the severity of disabilities within each age group ( Nicholas et al. 1977, Basu 1981, Basu & Sokhey 1984, Heymann 1984, Deivanayagam & Nedunchelian 1992). We assume that all patients affected by paralytic poliomyelitis become disabled to some extent, since the estimated incidence was derived from the lameness surveys of residual paralysis. Furthermore, the percentage distribution of the degree of disability, adjusted to be consistent with the lameness surveys and their definition of disability, is assumed to be constant among lame patients over each age group, even though serious illness more likely results when infection occurs in older individuals. Other Considerations in Calculating Disease Burden To estimate incidence from prevalence of residual paralysis of poliomyelitis, it is assumed that each region is in a stable, epidemic or poliomyelitis free situation. Lameness surveys most often target children 5 to 9 years of age in developing countries where poliomyelitis is endemic. As a consequence, residual paralysis in children of 5 to 9 years of age reflects the paralytic poliomyelitis cases that occurred from 1 to 10 years before the survey. Therefore, if a sudden increase in incidence (poliomyelitis outbreaks) occurred some years prior to the lameness survey, the actual incidence would be underestimated. In contrast, if a reduction of poliomyelitis incidence had been achieved through effective interventions, such as vaccination and improvement of sanitation, the actual incidence would be overestimated. Although changes in vaccine coverage and subsequent changes in incidence have been taken into account in our estimation, some of the limitations of the original surveys should be noted. Since our estimate of the incidence of residual paralysis is based on lameness surveys for developing regions, sporadic cases other than indigenous poliomyelitis cases including imported and vaccine-associated cases are not captured in our calculation of the burden of poliomyelitis in those regions. Disability weights for paralytic poliomyelitis were derived from the studies in developing regions. It is suggested that disability attributable to poliomyelitis in developed regions is more severe, reflecting the increase in mean age of infection (Nathanson & Martin 1979). For example, in the review of the poliomyelitis cases in the United States, the proportion of severe paralysis was high: 11 per cent of patients had minor paralysis whereas 79 per cent had significant or severe residual paralysis that required mechanical aids (Moore et al. 1982). Although the application of the same disability weights to all regions would underestimate the burden in developed regions, the bias would be minimal since the number of reported incident cases in developed regions (EME and FSE) was quite small. Neither late effects of paralytic poliomyelitis (post-paralytic poliomyelitis syndrome) nor partial remission resulting from an effective rehabilitation programme that might alter the sequelae of disease are taken into 126 Global Epidemiology of Infectious Diseases account in our calculation of the disease burden. It is assumed that the severity of residual paralysis remains constant over the rest of life. Burden of Poliomyelitis in 1990 Tables 5.5a and 5.5b represent the detailed epidemiological estimates (incidence, prevalence, number of deaths, years of life lost, years lived with disability, and disability adjusted life years) for poliomyelitis by age, sex, and region in 1990. The distribution of estimated incidence of poliomyelitis by region is given in Figure 5.3. Among the estimated global total of 212 000 paralytic poliomyelitis cases including complete recovery and fatal cases, India accounted for 47 000 cases, or 39 per cent of the global total in 1990. It was followed by SSA with 38 000 cases (18 per cent), MEC with 30 000 cases (14 per cent), OAI with 26 000 cases (12 per cent), China with 24 000 cases (11 per cent), and LAC with 12 000 cases (6 per cent). The majority of reported incidence in FSE were the outbreak cases in several southern former Soviet republics. The two poliomyelitis-endemic regions—India and SSA—accounted for more than half of the global incidence of poliomyelitis in 1990. The estimated number of deaths attributable to poliomyelitis in 1990 was approximately 27 000, with a range of 20 000 to 35 000 depending on the case-fatality rate used in the analysis. The burden of disease, measured by years of life lost (YLLs), years lived with disability (YLDs) and disability-adjusted life years (DALYs), shows a similar pattern. The global burden of poliomyelitis was 3 360 000 DALYs, which amounted to 0.24 per cent of the global burden of disease from all causes in 1990. The proportion of burden from poliomyelitis was highest in India (0.5 per cent), followed by MEC, OAI and SSA. On average, YLDs accounted for 75 per Figure 5.3 Annual incidence of paralytic poliomyelitis, 1990 250 000 Annual Incidence 200 000 150 000 100 000 50 000 0 EME FSE LAC MEC SSA OAI WDR Region IND CHI Total Number (’000s) Rate per 100 000 Incidence 0 0 0 0 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0–4 5–14 15–44 45–59 60+ All Ages India (IND) 0–4 5–14 15–44 45–59 60+ All Ages 46 0 1 0 0 48 0 0 0 0 0 0 77.50 0.2 0.7 0.3 0.0 11.00 0.0 0.0 0.0 0.0 0.0 0.0 Formerly Socialist Economies (FSE) 0–4 5–14 15–44 45–59 60+ All Ages 141 690 1 500 362 227 2 920 0 0 0 0 0 0 0 0 0 0 0 0 Number (’000s) 235.8 678.1 748.0 760.0 762.6 664.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Rate per 100 000 Prevalence 2.5 9.7 29.8 52.2 69.7 3.5 — — — — — — — — — — — — Avg. Age at Onset (years) 58.7 56.4 38.7 20.0 9.8 58.0 — — — — — — — — — — — — Average Duration (years) Epidemiological estimates for poliomyelitis, 1990, males Established Market Economies (EME) Age group (years) Table 5.5a 6 0 0 0 0 6 0 0 0 0 0 0 0 0 0 0 0 0 10.40 0.0 0.0 0.0 0.0 1.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.0 Rate per 100 000 Deaths Number (’000s) 208 1 1 0 0 210 0 0 0 0 0 0 0 0 0 0 1 2 531 3 14 1 0 549 0 0 0 0 0 0 0 0 0 0 0 0 YLDs (’000s) YLLs (’000s) DALYs continued 739 4 15 1 0 758 0 0 0 0 0 0 0 0 1 0 1 2 (’000s) Poliomyelitis 127 13 0 1 0 0 14 Number (’000s) 14 0 1 0 0 15 0–4 5–14 15–44 45–59 60+ All Ages 21 0 1 0 0 22 Sub–Saharan Africa (SSA) 0–4 5–14 15–44 45–59 60+ All Ages 45.10 0.2 0.5 0.3 0.0 8.8 31.20 0.2 0.5 0.3 0.0 4.3 20.80 0.1 0.3 0.2 0.0 2.4 Rate per 100 000 Incidence 58 213 339 68 36 714 37 180 376 83 50 725 32 106 355 89 60 642 Number (’000s) 122.1 303.2 326.0 336.0 339.0 283.0 83.0 214.0 234.0 243.1 245.0 211.4 53.1 109.3 115.9 122.5 123.0 109.7 Rate per 100 000 Prevalence 2.4 9.8 29.5 52.2 69.5 3.2 2.5 10.0 29.8 52.3 70.1 4.4 2.5 10.0 29.9 52.3 69.6 5.1 Avg. Age at Onset (years) 50.6 50.9 36.1 19.7 9.5 50.2 60.8 57.8 40.1 21.2 10.1 59.4 64.5 59.0 40.5 20.7 9.6 62.3 Average Duration (years) Epidemiological estimates for poliomyelitis, 1990, males (continued) Other Asia and Islands (OAI) 0–4 5–14 15–44 45–59 60+ All Ages China (CHI) Age group (years) Table 5.5a 3 0 0 0 0 3 2 0 0 0 0 2 1 0 0 0 0 1 5.6 0.0 0.0 0.0 0.0 1.1 4.2 0.0 0.0 0.0 0.0 0.6 2.3 0.0 0.0 0.0 0.0 0.2 Rate per 100 000 Deaths Number (’000s) 90 1 2 0 0 93 61 1 3 0 0 65 47 0 1 0 0 49 232 2 5 0 0 240 158 2 9 0 0 169 148 1 11 1 0 160 YLDs (’000s) YLLs (’000s) DALYs 323 3 7 0 0 335 219 3 11 1 0 234 195 2 12 1 0 209 (’000s) 128 Global Epidemiology of Infectious Diseases 7 0 0 0 0 8 25.80 0.1 0.2 0.1 0.0 3.5 0–4 5–14 15–44 45–59 60+ All Ages World 0–4 5–14 15–44 45–59 60+ All Ages 118 1 5 0 0 124 17 0 1 0 0 18 36.70 0.1 0.4 0.2 0.0 4.7 40.30 0.2 0.5 0.2 0.0 6.7 Middle Eastern Crescent (MEC) 0–4 5–14 15–44 45–59 60+ All Ages Latin America and the Carribean (LAC) 333 1 448 3 076 707 439 6 002 45 172 321 65 40 643 20 87 186 41 26 360 103.0 262.7 246.1 226.3 200.6 226.2 109.3 263.2 281.8 291.0 293.0 250.8 69.6 166.9 178.4 182.0 182.7 162.4 2.5 9.8 29.8 52.3 69.8 3.8 2.5 9.8 29.8 52.3 70.1 3.5 2.5 9.8 29.8 52.3 70.9 3.5 58.9 56.6 39.5 20.7 9.9 58.0 61.6 58.8 40.8 21.5 10.1 60.8 64.0 59.5 41.8 23.0 10.9 63.1 150 0 0 0 0 150 2 0 0 0 0 2 1 0 0 0 0 1 4.7 0.0 0.0 0.0 0.1 0.6 5.2 0.0 0.0 0.0 0.0 0.8 3.1 0.0 0.0 0.0 0.0 0.4 508 3 9 1 1 521 72 0 0 0 0 73 30 0 0 0 0 30 1 348 11 46 2 0 1 408 192 1 6 0 0 200 87 1 2 0 0 90 1 857 14 55 3 1 1 929 264 2 6 0 0 273 117 1 3 0 0 120 Poliomyelitis 129 Number (’000s) Rate per 100 000 Incidence 0 0 0 0 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0–4 5–14 15–44 45–59 60+ All Ages India (IND) 0–4 5–14 15–44 45–59 60+ All Ages 340 0 1 0 0 350 0 0 0 0 0 0 59.80 0.2 0.5 0.3 0.0 8.6 0.0 0.0 0.0 0.0 0.0 0.0 Formerly Socialist Economies (FSE) 0–4 5–14 15–44 45–59 60+ All Ages 103 498 1 056 270 170 2096 0 0 0 0 0 0 0 0 0 0 0 0 Number (’000s) 181.0 525.0 576.3 586.0 587.7 511.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Rate per 100 000 Prevalence 2.5 9.7 29.8 52.3 70.1 3.5 — — — — — — — — — — — — Avg. Age at Onset (years) 59.6 57.8 40.7 21.5 10.1 58.9 — — — — — — — — — — — — Average Duration (years) Epidemiological estimates for poliomyelitis, 1990, females Established Market Economies (EME) Age group (years) Table 5.5b 5 0 0 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0 Number (’000s) 8.3 0.0 0.0 0.0 0.0 1.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 Rate per 100 000 Deaths 160 1 0 0 0 161 0 0 0 0 0 0 0 0 0 0 0 0 Number (’000s) YLLs 389 2 10 1 0 402 0 0 0 0 0 0 0 0 0 0 0 0 Number (’000s) YLDs 549 3 11 1 0 564 0 0 0 0 0 0 0 0 0 0 1 1 Number (’000s) DALYs 130 Global Epidemiology of Infectious Diseases 9 0 1 0 0 10 100 0 1 0 0 110 160 0 0 0 0 170 34.70 0.1 0.4 0.2 0.0 6.5 24.00 0.1 0.4 0.2 0.0 3.2 16.00 0.1 0.3 0.1 0.0 1.9 0–4 5–14 15–44 45–59 60+ All Ages 5 0 0 0 0 6 19.90 0.1 0.2 0.1 0.0 2.6 Latin America and the Carribean (LAC) 0–4 5–14 15–44 45–59 60+ All Ages Sub–Saharan Africa (SSA) 0–4 5–14 15–44 45–59 60+ All Ages Other Asia and Islands (OAI) 0–4 5–14 15–44 45–59 60+ All Ages China (CHI) 18 119 280 63 46 526 44 163 267 57 33 565 27 132 287 66 43 555 24 76 253 61 49 463 65.0 234.5 269.0 269.7 271.0 236.2 93.6 234.0 251.3 259.0 261.0 219.0 64.3 164.6 179.8 187.0 189.0 163.4 41.4 84.1 89.1 94.0 94.8 84.4 2.5 9.8 29.9 52.4 71.6 3.6 2.4 9.8 29.5 52.3 70.0 3.3 2.5 10.0 29.8 52.4 70.8 4.5 2.5 10.0 29.9 52.4 70.6 5.0 68.2 63.4 45.0 25.3 11.6 67.3 53.8 53.5 38.2 21.1 10.0 53.3 64.9 61.4 43.3 23.6 10.8 63.4 67.5 62.3 43.6 23.3 10.6 65.4 1 0 0 0 0 1 2 0 0 0 0 2 1 0 0 0 0 1 1 0 0 0 0 1 2.2 0.0 0.0 0.0 0.0 0.3 4.1 0.0 0.1 0.0 0.0 0.8 3.0 0.0 0.0 0.0 0.0 0.4 1.8 0.0 0.0 0.0 0.0 0.2 20 0 0 0 0 21 65 1 2 0 0 68 43 1 2 0 0 46 35 0 0 0 0 36 65 0 2 0 0 68 181 1 4 0 0 187 119 1 7 0 0 127 110 1 8 0 0 120 continued 86 0 2 0 0 89 247 2 6 0 0 255 162 2 9 1 0 173 146 1 9 0 0 156 Poliomyelitis 131 Number (’000s) Rate per 100 000 Incidence 0–4 5–14 15–44 45–59 60+ All Ages World 0–4 5–14 15–44 45–59 60+ All Ages 870 1 3 0 0 920 120 0 0 0 0 130 28.30 0.1 0.3 0.1 0.0 3.5 31.00 0.1 0.4 0.2 0.0 5.2 249 1 114 2 377 566 376 4 682 33 125 233 50 35 476 Number (’000s) 80.0 212.0 198.3 182.0 139.7 179.1 84.0 201.6 217.3 224.3 225.0 193.0 Rate per 100 000 Prevalence 2.5 9.8 29.8 52.4 70.5 3.8 2.5 9.8 29.9 52.4 70.8 3.5 Avg. Age at Onset (years) 61.2 59.2 42.5 22.7 10.5 60.3 64.4 61.7 43.6 23.7 10.8 63.6 Average Duration (years) 11 0 0 0 0 12 2 0 0 0 0 2 3.6 0.0 0.0 0.0 0.0 0.4 4.0 0.0 0.0 0.0 0.0 0.7 Rate per 100 000 Deaths Number (’000s) Epidemiological estimates for poliomyelitis, 1990, females (continued) Middle Eastern Crescent (MEC) Age group (years) Table 5.5b 378 2 7 1 0 388 54 0 0 0 0 54 Number (’000s) YLLs 1 010 8 35 2 0 1 055 145 1 4 0 0 150 Number (’000s) YLDs 1 388 10 41 3 1 1 442 198 1 5 0 0 205 Number (’000s) DALYs 132 Global Epidemiology of Infectious Diseases Poliomyelitis 133 cent of total disease burden from poliomyelitis. As shown in Figures 5.4 and 5.5, India and SSA accounted for approximately 60 per cent of total deaths and disability-adjusted life years from poliomyelitis in 1990. Burden and Intervention The global incidence of paralytic poliomyelitis was dramatically reduced with the introduction of the Salk inactivated poliomyelitis vaccine (IPV) in 1955 and the Sabin live-attenuated oral vaccine (OPV) in 1961. Mass campaigns in addition to routine immunization with OPV in the 1960s Figure 5.4 Annual number of deaths due to poliomyelitis, 1990 Number of Deaths 15 000 12 000 9 000 6 000 3 000 0 EME FSE LAC MEC SSA OAI IND CHI Total WDR Region Figure 5.5 DALYs lost due to paralytic poliomyelitis, 1990 5 000 000 DALYs Lost 4 000 000 3 000 000 2 000 000 1 000 000 0 EME FSE LAC MEC SSA OAI WDR Region IND CHI Global 134 Global Epidemiology of Infectious Diseases virtually eliminated paralytic poliomyelitis from developed regions. Both vaccines have advantages and limitations for developing countries, as summarized in Box 5.2 (Melnick 1978,1990). Because of its low cost, ease of administration, superiority in inducing secretory immunity in the Box 5.2 Advantages and disadvantages of poliomyelitis vaccines Advantages Disadvantages Inactivated Poliomyelitis Vaccine (IPV) Confers humoral immunity in vaccinees if sufficient numbers of doses of potent vaccine are given. Can be incorporated into regular immu- nization programme, such as diphtheriapertussis-tetanus Absence of living virus excludes po- tential for mutation and reversion to virulence. Absence of living virus permits its use in immunodeficient or immunosuppressed individuals and their households. Has greatly reduced the spread of po- lioviruses in some regions where it has been properly used. May prove useful in certain tropical areas where live virus vaccine has failed to take in some young infants. Early studies indicated a disappointing record in percentage of vaccines developing antibodies after three doses (but more immunopotent antigens are now produced). Generally, repeated boosters have been required to maintain detectable antibody levels. Does not induce appreciable local in- testinal immunity in the vaccinee; hence vaccines do not serve as a block to transmission of wild polioviruses by the faecal–oral route More expensive than live virus vaccine. Use of virulent polioviruses as vaccine seed creates potential for tragedy. This problem can, however, be overcome by use of attenuated strains for production. Oral Poliomyelitis Vaccine (OPV ) Like the natural infection, confers both humoral and intestinal immunity. Vaccine virus may mutate, and in very Induces antibody very quickly in a large proportion of vaccinees. rare instances has reverted toward neurovirulence sufficient to cause paralytic poliomyelitis in recipients and their contacts. Oral administration is more acceptable Vaccine progeny virus spreads to house- Immunity induced appears to be lifelong. to vaccinees than injection, and easier to accomplish. Administration does not require use of highly trained personnel. When properly stabilized, can retain potency under difficult field conditions with little refrigeration and no freezing. Under epidemic conditions, not only induces antibody quickly, but also rapidly infects the alimentary tract, blocking epidemic spread of virus. Is relatively inexpensive, both to produce and to administer, and does not require continuing booster doses. Source: Melnick (1978, 1988). hold contacts. Vaccine virus also spreads to people in community. In certain tropical countries, induction of antibodies in a satisfactorily high proportion of vaccinees has been difficult to accomplish. Contraindicated in those with immuno- deficiency diseases, and in their household associates, as well as in people undergoing immunosuppresive therapy and their households. Poliomyelitis 135 intestines, and ability to infect household and community contacts, OPV has been used by WHO for the purpose of eradicating wild poliovirus (Wright et al. 1991, Hull et al. 1994, 1997). Initially, the WHO Expanded Programme on Immunization (EPI) promoted three doses of OPV for all children in the first year of life as the routine immunization programme. Since the foundation of the EPI programme, WHO, UNICEF, and other international donors have contributed to establishing an effective routine immunization system. Vaccine coverage for poliomyelitis with OPV has increased and reported incidence has substantially decreased (Figures 5.1 and 5.2). Although the primary strategy of EPI was to raise the vaccine coverage of OPV and sustain the higher level, it has been suggested that poliomyelitis eradication may not be achieved through a routine immunization programme alone (Sutter et al. 1991) and three doses of OPV have proven inadequate in some endemic countries (Montefiore et al. 1963, John 1976, Bottiger et al. 1981, Chopra, Kundu & Chowdhury 1989, Patriarca 1993). Some studies have shown that the seroconversion rate and the efficacy are lower in the tropics (Patriarca, Wright & John 1991, Sutter et al. 1991) and an additional fourth dose at birth (or five doses) has been recommended in these areas ( John 1976, Dong et al. 1986, Chopra, Kundu & Chowdhury 1989, Hull & Ward 1992). For example, despite the relatively high vaccine coverage rates, several outbreaks among children fully immunized with three doses of OPV have been reported in India, China, Other Asia and Islands (OAI), Middle East Crescent (MEC) and Former Socialist Economies of Europe (FSE) regions (Expanded Programme on Immunization 1993, Lasch et al. 1984, Samuel, Sutter et al. 1991, Hull & Ward 1992, Balraj & John1993). Extremely high routine immunization coverage has failed to prevent epidemics in China, which reported epidemics of poliomyelitis in 1989 and 1990 despite the 98 per cent coverage with three doses of OPV achieved in 1990 (Expanded Programme on Immunization 1993). In some developing regions where OPV alone would not decrease incidence of poliomyelitis further, combined OPV/IPV regimens have been used and proven successful (Magnus & Peterson 1984, Melnick 1988, Tulchinsky et al. 1989). The vaccines do not interfere with each other and each compensates for some of the other’s limitations. It has been suggested that failure to vaccinate is much more critical than vaccine failure for an outbreak to take place (Anderson & May 1985, Kim-Farley et al. 1984, Patriarca, Sutter & Oostovogel 1997). In fact, in China, most of the paralytic poliomyelitis cases occurred in unimmunized or incompletely immunized children as shown in Figure 5.6 (Expanded Programme on Immunization 1993). Routine vaccination may not reach unregistered populations, certain religious groups, ethnic minorities, illegal immigrants or migratory populations. Other factors, such as false contraindication, poor access, and attitude and knowledge of health personnel, may lead to missed opportunities (Pan American Health Organization 1994). Although a role for IPV deserves further exploration, especially in 136 Global Epidemiology of Infectious Diseases Figure 5.6 Immunization status of poliomyelitis cases in Shandong Province, 1990 Number of Cases 150 120 90 60 30 0 Not immunized (0) Partially immunized (1–2) Fully immunized (3+) Unknown Immunization Status developed regions where transmission has been interrupted, improving vaccine delivery rather than changing immunization regimen should be the priority for poliomyelitis eradication since many cases are attributable to missed opportunities (John 1976, Wright et al. 1991, Hull et al. 1994). WHO currently recommends implementation of mass campaigns in addition to maintaining a high coverage rate of routine vaccination with four doses of OPV (Hull et al. 1994, 1997). In Brazil, a policy of holding a national immunization day, when all children under 5 years of age are vaccinated regardless of prior immunization status, has led to the cessation of poliovirus transmission (Expanded Programme on Immunization 1993). This policy was implemented to compensate for the failures of routine vaccination, such as thermal inactivation and lower seroconversion rate. The success of the approach has made it the model for poliomyelitis eradication in the Latin America and Caribbean (LAC) Region. WHO has endorsed this strategy and recommends that all children under 5 years of age receive an additional two doses of OPV during the national immunization days (NIDs), preferably during the low season for poliovirus transmission (Wright et al. 1991, Hull & Ward 1992, Centers for Disease Control and Prevention 1993, Ward et al. 1993). There was a substantial increase in the number of countries with endemic poliomyelitis that conducted national immunization days from 16 countries in 1988 to 62 countries in 1995, resulting in tremendous successes in LAC and China, and some parts of OAI, MEC and sub-Saharan Africa (SSA) regions (Hull et al. 1994, Birmingham et al. 1997a, Cochi et al. 1997). What would be the current burden attributable to paralytic poliomyelitis Poliomyelitis 137 if immunization coverage were zero? We have estimated the current burden of poliomyelitis in the absence of immunization for polio. Before the introduction of OPV, in the early 1950s, poliomyelitis was highly endemic all over the world. Figures for the incidence of poliomyelitis at that time were available only in EME (Nathanson & Martin 1979) and some parts of today’s FSE (Freyche & Nielsen 1955). Lameness survey data before dissemination of OPV, before 1980, are available for some developing regions, including India, LAC, MEC, SSA, and OAI, with some studies reporting vaccine coverage (Foster et al. 1984, Khajuria et al. 1989, Nicholas et al. 1997). Where data for the 1950s were not available, indirect methods were used, based on the historical trend of epidemiological and demographic transition, and including lameness surveys and vaccine coverage. Because of the epidemiological shift of disease occurrence in developed regions (EME and FSE), the following age-distribution proportions of poliomyelitis incidence were used for developed regions: 47 per cent for 0–4, 24 per cent for 5–14, 28 per cent for 15–44, 0.06 per cent for 45–59 and 0.02 per cent for 60 and over ( Paul 1955, Moore et al. 1982). For developing regions, where the majority of cases occur under 5 years of age, the agedistribution proportions for each age group were assumed to follow the endemic pattern: 97 per cent for age 0–4; 5.4 per cent for age 5–14; 1.6 per cent for age 15–44; 0.8 per cent for age 45–59; and 0.06 per cent for age 60 and over. Figure 5.7 illustrates that over 800 000 cases would have occurred annually in the absence of current vaccination programmes, and over half a million cases could be prevented with current immunization coverage (Figure 5.8). It should be noted, however, that an improved standard of living, including better sanitation and hygiene, might have reduced the incidence of poliomyelitis during this period. Nonetheless, the result illustrates the significant impact of immunization programmes, especially in developing regions where these programmes, along with case manageFigure 5.7 Reduction of polio incidence by region, 1990 900 000 1990 (no vaccine) 1990 (vaccine available) Annual Incidence 800 000 700 000 600 000 500 000 400 000 300 000 200 000 100 000 0 EME FSE LAC MEC SSA OAI WDR Region IND CHN Global 138 Global Epidemiology of Infectious Diseases Figure 5.8 Anticipated reduction of polio incidence with full vaccine coverage 250 000 1990 (actual coverage) 1990 (100% coverage) Annual Incidence 200 000 150 000 100 000 50 000 0 EME FSE LAC MEC SSA OAI IND CHN Global WDR Region ment and surveillance systems, have been implemented and improved both through the commitment of governments and with the financial and technical support of international community. The lessons learned from smallpox eradication suggest that poliomyelitis eradication depends on effective disease surveillance to guide vaccine strategy, verify outcome and certify success, particularly in the final stage toward eradication (Pan American Health Organization 1993). Based on the experiences in the LAC Region, the focus of EPI shifted in the 1990s from monitoring vaccine coverage to assessing disease incidence through surveillance (Birmingham et al. 1997b). WHO has recommended the strengthening of the following elements in the poliomyelitis eradication programme in addition to the routine OPV vaccination and mass campaigns (Hull et al. 1994, 1997): improve the acute flaccid paralysis (AFP) surveillance system to the point where each country can detect all cases of paralysis which might be poliomyelitis; build a global network of laboratories with the technical competence to reliably identify poliovirus from specimens collected from these cases; and conduct supplemental immunization activities including outbreak response immunization and mopping-up immunization. As discussed earlier in this chapter, AFP surveillance was widely used during the 1990s as a critical component of poliomyelitis eradication strategies (Box 5.3) (de Quadros et al. 1997, Hull et al. 1997). Since there are no absolute criteria that will permit poliomyelitis to be identified on a clinical basis, WHO recommends laboratory-based AFP surveillance Poliomyelitis Box 5.3 139 Key interventions for poliomyelitis eradication Interventions Indicators High routine immunization coverage At least 90 per cent of infants should be vaccinated against poliomyelitis by one year of age through routine four doses of OPV immunization. National immunization days All children under 5 years of age should be immunized regardless of prior immunization status during two rounds of mass campaign. Laboratory-based surveillance of acute flaccid paralysis (AFP) surveillance All AFP cases in children under 15 years of age should be reported immediately; clinical and epidemiological investigation should begin within 48 hours; two stool samples should be collected; stool from contacts of cases should also be tested. Mopping-up immunization All children under 5 years of age should be immunized, regardless of prior immunization status, in the targeted district. Source: (Hull et al. 1994, 1997). for the purpose of poliomyelitis eradication. The performance of AFP surveillance in each country is evaluated by the following five indicators: at least 80 per cent of representative health units within the reporting network should report regularly (weekly or monthly); at least one case of non-poliomyelitis AFP should be detected annually per 100 000 population under 15 years of age; at least 80 per cent of all AFP cases reported should be investigated within 48 hours and have two stool specimens taken for virus culture within two weeks of onset of acute paralysis; at least 80 per cent of all AFP cases should have a follow-up examination of residual paralysis at 60 days after the onset of paralysis; and all virological studies on AFP cases must be performed in a laboratory of certified competence that is part of the WHO’s global poliomyelitis laboratory network (Hull & Dowdle 1997, World Health Organization 1998). Many endemic countries, including China, India, and countries in the SSA, MEC, and OAI regions, started national immunization days and AFP surveillance in the 1990s. It is suggested, however, that the efficiency and quality of surveillance vary considerably within and between countries (Centers for Disease Control and Prevention 1994). For example, the assessment of performance of AFP surveillance in SSA, MEC, and FSE in the early 1990s showed great variations between countries and evidence of underreporting of AFP cases in these regions (Birmingham et al. 1997b). For example, the non-poliomyelitis AFP rate, which is an indicator of AFP surveillance sensitivity, varies significantly among regions, from less than 0.1 in SSA to 1.2 in LAC, CHN and a part of OAI (Tangermann et al. 1997). Developing a surveillance system is a long-term commitment that must be maintained until global eradication is certified (Eichner & 140 Global Epidemiology of Infectious Diseases Dietz 1996), and AFP surveillance poses special difficulties in countries with inadequate health infrastructure (Hull & Dowdle 1997). In countries where poliovirus circulation has been limited to focal areas, and surveillance is adequate, the next step towards eradicating the transmission of wild poliovirus completely and permanently from the region is a moppingup campaign (Pan American Health Organization 1993). A mopping-up campaign comprises a response to an additional outbreak by immunizing children living in a community where a case of suspected poliomyelitis has occurred and carrying out a localized immunization campaign targeting high-risk population. As shown in World development report 1993: investing in health (World Bank 1993), immunization is one of the most cost-effective health interventions. In a study of poliomyelitis immunization in Brazil, intensification strategies such as mass campaigns had lower average costs than routine immunization programmes (Creese 1984). Although Shepard (1989) has suggested that most studies do not report the incremental cost-effectiveness of additional expenditures on routine vaccination, mass campaign programmes are likely to have a greater impact per dollar spent than are routine vaccination since they have lower incremental costs for additional vaccines and their efficacy has been proven to be greater (Richardson et al. 1995, Reichler et al. 1997). Moreover, such a comprehensive effort could be extended to combine other interventions, such as vitamin A supplementation, that are considered to be cost-effective. A particular feature of the economic analysis of the health interventions for poliomyelitis is the assessment of benefits of eradication. If an eradication programme is feasible and its benefit is discounted at low rates, it yields almost infinite benefits, from a cost-effectiveness viewpoint, and any amount of finite costs can be undertaken to eradicate a disease (Acharya & Murray 1998). Such a calculation nevertheless depends on the timespan of the global poliomyelitis eradication processes in different regions. In countries where immunization and surveillance are inadequate, extending the implementation of additional strategies for poliomyelitis eradication would be quite costly in the short run and the current expenditure for poliomyelitis eradication would be cost-effective only when benefits in the distant future were taken into account. In contrast, in countries where poliovirus transmission has been interrupted and active surveillance has been established, immediate eradication rather than delayed eradication proves more cost-effective (Musgrove 1988, Bart, Foulds & Patriarca 1996). Furthermore, since the eradication of poliomyelitis in a region also depends on poliovirus transmission in other regions, it is more cost-effective for such a region to have others engaged in intensive strategies to prevent the reintroduction of wild poliovirus (Hall et al. 1998). Therefore, global poliomyelitis eradication programmes should be cost-effective from a global viewpoint, even though they may not be cost-effective for a single country in the short run. Since poliomyelitis eradication strategies would impose a heavy financial burden Poliomyelitis 141 on the health care resources of many endemic countries, the international community has a significant role to play in providing resources to sustain global poliomyelitis eradication efforts ( Hull et al. 1997, Satcher 1999, Aylward et al. 2000). Conclusions Reported cases of poliomyelitis have been dramatically decreased since the global poliomyelitis eradication initiative in 1988. We expect that progress will continue until the eradication of poliomyelitis early in the 21st century. As our estimates for 1990 have shown, however, paralytic poliomyelitis still causes more burden than would be expected from the reported incidence, especially in India and sub-Saharan Africa. The relatively low efficacy of the surveillance and reporting systems in developing regions, in which poliomyelitis is still endemic, may underestimate the true magnitude of paralytic poliomyelitis in past decades. Although rising coverage of routine oral poliomyelitis vaccination (OPV) has contributed to the significant decline in the incidence of paralytic poliomyelitis, it is suggested that polio eradication may not be achieved through the current vaccination programme alone. Implementation and establishment of mass campaigns and laboratory-based surveillance are critical to the final stage of the eradication. The benefits of eradicating poliomyelitis include: reduction in the burden of life-long disability and premature death; savings in costs of prevention, treatment and rehabilitation, and other non-medical costs; and gains in productivity. Such benefits could exceed the additional costs for expanding the key strategies, not only for countries free from poliovirus in the short run, but even for endemic poor countries in the long run. Under the WHO initiative, national and international financial and technical support for the endemic countries is essential to sustain the vaccine supply and to improve the surveillance system. Update: Epidemiological trends in the 1990s Since the adoption of the global poliomyelitis eradication initiative in 1988, WHO has been leading a global partnership of international organizations and governments to implement strategies to eradicate poliomyelitis for poliomyelitis eradication in all remaining endemic countries. This chapter estimates the global burden of paralytic poliomyelitis in 1990; but there has subsequently been substantial progress towards eradication. The number of known or suspected poliomyelitis-endemic countries has decreased from over 120 to less than 50 and is expected to continue to decline (World Health Organization 1999). The reported incidence of paralytic poliomyelitis declined from 35 251 in 1988 to 6349 in 1998, more than 85 per cent reduction over the decade. Nevertheless, the estimated incidence was approximately 80 000, assuming that less than 10 per 142 Global Epidemiology of Infectious Diseases cent of cases were reported (Hull et al. 1997). Tremendous achievements were observed in the Americas, where eradication of wild poliovirus was announced in 1994 (Robbins & de Quadros 1997). The Western Pacific Region, which includes the world’s most populous country, China, was recently certified free of indigenous wild poliovirus transmission (World Health Organization 2000). Success in these regions has proved the effectiveness of the strategies recommended by WHO, and the number of countries implementing the strategies has increased rapidly since the mid-1990s. Global routine OPV coverage peaked in 1990 at 85 per cent and stabilized at around 80 per cent thereafter (Figure 5.1). The number of countries conducting national immunization days (NIDs), often coupled with other primary health care interventions, has increased to over 100 including many endemic countries in Africa and South-East Asia (Aylward et al. 2000). AFP surveillance has been implemented in many countries in which poliomyelitis is or recently was endemic (Tangermann et al. 1997). Despite these efforts, poliomyelitis is still endemic in some 50 countries, including India, sub-Saharan Africa, and parts of the MEC and OAI Regions (Satcher 1999). In order to achieve the goal of eradication, the rapid development of complete and timely AFP surveillance and the continuation of national immunization days are a priority (Hull et al. 1998). Compared to implementation of national immunization days, the establishment of AFP surveillance has been delayed. In 1998, the global average rate for non-poliomyelitis AFP exceeded the targeted level at 1.15 per 100 000 population but the rates in endemic areas were still much lower. Eradication efforts should now focus on the remaining poliomyelitisendemic countries, especially those in sub-Saharan Africa and the Indian Subcontinent that are affected by war or are politically isolated. From the remaining reservoirs in those countries wild poliovirus can continue to spread into bordering or even distant poliomyelitis-free regions (Centers for Disease Control and Prevention 1995, Tangermann et al. 1997, 2000, Wise 1999). The major obstacle facing eradication is inadequate political and financial commitment, the success of the initiative rests on political commitment within the endemic countries together with support from international organizations and donor countries (Centers for Disease Control and Prevention 1998, Hull et al. 1998, Satcher 1999). Although the goal of eradicating of poliomyelitis by the year 2000 was not achieved, enough resources should continue to be mobilized through international cooperation until poliomyelitis is eradicated. The Global Alliance for Vaccines and Immunizations (GAVI), which aims to mobilize enough resources to reach children who are currently missed in immunization efforts, has been launched to that end. Poliomyelitis 143 References Acharya AK, Murray CJL (1998) Economic appraisal of eradication programs: the question of infinite benefits. In: Dowdle WR, Hopkins DR, eds. The eradication of infectious diseases: Dahlem Workshop reports. Chichester, John Wiley, 75–90. Aylward RB et al. (2000) Disease eradication as a public health strategy: a case study of poliomyelitis eradication. Bulletin of the World Health Organization, 78:285–297. Anderson RM, May RM (1985) Vaccination and herd immunity to infectious diseases. Nature, 318:323–329. Anderson RM, May RM (1993) Infectious diseases of humans: dynamics and control. New York, Oxford University Press. Andrus JK et al. (1992) Screening of case of acute flaccid paralysis for poliomyelitis eradication: ways to improve specificity. Bulletin of the World Health Organization, 70:591–596. Assaad F, Ljungars-Esteves K (1984) World overview of poliomyelitis: regional patterns and trends. Reviews of Infectious Diseases, 6(Suppl 2):S302–307. Babaniyi O, Parakoyi B (1991) Cluster survey for poliomyelitis and neonatal tetanus in Ilorin, Nigeria. International Journal of Epidemiology, 20:515–520. Bart KJ, Foulds J, Patriarca P (1996) Global eradication of poliomyelitis: benefitcost analysis. Bulletin of the World Health Organization, 74:35–45. Basu RN (1981) Magnitude of problem of poliomyelitis in India. Indian Pediatrics, 18:507–511. Basu RN, Sokhey J (1984) Prevalence of poliomyelitis in India. Indian Journal of Pediatrics, 51:515–519. Belcher DW et al. (1979) Comparison of methods for estimating the frequency of paralytic poliomyelitis in developing countries. Bulletin of the World Health Organization, 57:301–307. Bernier RH (1983) Prevalence survey techniques for paralytic polio: an update. Geneva, World Health Organization. (EPI/GAG 83/WP10). Bernier RH (1984) Some observations on poliomyelitis lameness surveys. Reviews of Infectious Diseases, 6(Suppl 2):S371–375. Biellik RJ et al. (1992) Poliomyelitis case confirmation: characteristics for use by national eradication programmes. Bulletin of the World Health Organization, 70:79–84. Birmingham ME et al. (1997a) National immunization days: state of the art. Journal of Infectious Diseases, 175(Suppl 1):S183–188. Birmingham ME et al. (1997b) Poliomyelitis surveillance: the compass for eradication. Journal of Infectious Diseases, 175(Suppl 1):S146–150. Bottiger M et al. (1981) Antibodies against poliomyelitis and measles viruses in immunized and unimmunized children, Ghana, 1976–78. Bulletin of the World Health Organization, 59:729–736. 144 Global Epidemiology of Infectious Diseases Centers for Disease Control and Prevention (1993) Progress toward global eradication of poliomyelitis, 1988–1993. Morbidity and Mortality Weekly Report, 43:499-503. Centers for Disease Control and Prevention (1994) Emerging polio-free zone: Southern Africa, 1990–1994. Morbidity and Mortality Weekly Report, 43: 768–771. Centers for Disease Control and Prevention (1995) Progress toward global poliomyelitis eradication, 1985–1994. Morbidity and Mortality Weekly Report, 44: 273–275, 281. Centers for Disease Control and Prevention (1998) One thousand days until the target date for global poliomyelitis eradication. Morbidity and Mortality Weekly Report, 47:234. Chiba Y et al. (1992) Outbreaks of paralytic poliomyelitis and polio surveillance in Shandong Province of China. Japanese Journal of Medical Science and Biology, 45:255–266. Chodankar VP, Dave KH (1979) Pattern of prevalence of paralytic poliomyelitis in Greater Bombay, India. Journal of the Indian Medical Association, 73:25–29. Chopra K, Kundu S, Chowdhury DS (1989) Antibody response of infants in tropics to five doses of oral polio vaccine. Journal of Tropical Pediatrics, 35:19–23. Cochi SL et al. (1997) Commentary: the unfolding story of global poliomyelitis eradication. Journal of Infectious Diseases, 175(Suppl 1):S1–3. Creese AL (1984) Cost effectiveness of alternative strategies for poliomyelitis immunization in Brazil. Reviews of Infectious Diseases, 6(Suppl 2):S404–407. de Quadros CA et al. (1991) Eradication of poliomyelitis: progress in the Americas. Pediatric Infectious Disease Journal, 10:222–229. de Quadros CA et al. (1997) Eradication of wild poliovirus from the Americas: acute flaccid paralysis surveillance, 1988–1995. Journal of Infectious Diseases, 175(Suppl 1):S37–42. Debre R, Thieffery S (1955) Symptomatology and diagnosis of poliomyelitis. In: R, Debree, ed. Poliomyelitis, Geneva, World Health Organization, pp. 109–136, World Health Organization Monograph Series, No. 26. Deivanayagam N, Nedunchelian K (1992) Epidemiological and clinical features of acute poliomyelitis children admitted in an urban hospital. Indian Pediatrics, 29:25–28. Dong DX et al. (1986) Immunisation of neonates with trivalent oral poliomyelitis vaccine (Sabin). Bulletin of the World Health Organization,64:853–860. Dietz V et al. (1992) Predictors of poliomyelitis case confirmation at initial clinical evaluation: implications for poliomyelitis eradication in the Americas. International Journal of Epidemiology, 21:800–806. Eichner M, Dietz K (1996) Eradication of poliomyelitis: when can one be sure that polio virus transmission has been terminated? American Journal of Epidemiology, 143:816–822. Evans AS (1984) Criteria for assessing accomplishment of poliomyelitis control. Reviews of Infectious Diseases, 6(Suppl 2):S571–576. Poliomyelitis 145 Expanded Programme on Immunization (1992) Poliomyelitis in 1988, 1989, and 1990. Weekly Epidemiological Record, 67:113–117. Expanded Programme on Immunization (1993) Poliomyelitis in 1991. Weekly Epidemiological Record, 68:17–21. Foster SO et al. (1984) Control of poliomyelitis in Africa. Reviews of Infectious Diseases, 6(Suppl 2):S433–437. Freyche M-J, Nielsen J (1955) Incidence of poliomyelitis since 1920. In: Debre R. ed. Poliomyelitis. Geneva, World Health Organization, pp. 59–106. World Health Organization Monograph Series, No. 26. Hajar MM, et al. (1983) Prevalence, incidence, and epidemiological features of poliomyelitis in the Yemen Arab Republic. Bulletin of the World Health Organization, 61:353–359. Hall RG et al. (1998) Group report: what are the criteria for estimating the costs and benefits of disease eradication? In WR Dowdle, DR Hopkins, eds. The eradication of infectious diseases. Chichester, UK, John Wiley and Sons, pp. 107–124. Halstead LS, Wiechers DO, Rossi CD (1985) Late effects of poliomyelitis: a national survey. In: LS Halstead, DO Wiechers , eds. Miami, Fla. Late effects of poliomyelitis. Symposia Foundation, pp.11–32. Heymann DL (1984) House-to-house and school lameness surveys in Cameroon: a comparison of two methods for estimating the prevalence and annual incidence of paralytic poliomyelitis. Reviews of Infectious Diseases, 6(Suppl 2): S376–378. Heymann DL et al. (1983) Estimation of incidence of poliomyelitis by three survey methods in different regions of the United Republic of Cameroon. Bulletin of the World Health Organization, 61:501–507. Horstmann DM (1982) Control of poliomyelitis: a continuing paradox. Journal of Infectious Diseases, 146:540–551. Hull BP, Dowdle WR (1997) Poliovirus surveillance: building the global Polio Laboratory Network. Journal of Infectious Diseases, 175(Suppl 1):S113–116. Hull HF, Ward NA (1992) Progress towards the global eradication of poliomyelitis. World Health Statistics Quarterly, 45:280–284. Hull HF et al. (1997) Progress toward global polio eradication. Journal of Infectious Diseases, 175(Suppl 1):S4–9. Hull HF et al. (1998) Perspectives from the global poliomyelitis eradication initiative. Bulletin of the World Health Organization, 76(Suppl 2):42–46. Hull HF et al. (1994) Paralytic poliomyelitis: seasoned strategies, disappearing disease. Lancet, 343:1331–1337. Jamison DT et al. (1993) Poliomyelitis. In: Jamison et al., eds. Disease control priorities in developing countries. New York, Oxford Medical Publications, pp. 117–129. John TJ (1976) Antibody response of infants in tropics to five doses of oral polio vaccine. British Medical Journal, 1(6013):812. 146 Global Epidemiology of Infectious Diseases Joseph B et al. (1983) Comparison of techniques for the estimation of the prevalence of poliomyelitis in developing countries. Bulletin of the World Health Organization, 61:833–837. Kadi Z et al. (1988) Polio outbreak in Algeria: epidemiological, virological and vaccinational aspects. Transactions of the Royal Society of Tropical Medicine and Hygiene, 82:305–310. Khajuria R et al. (1989) Impact of annual immunisation programme with oral polio vaccine on the prevalence of paralytic poliomyelitis. Indian Journal of Pediatrics, 56:343–347. Khare S et al. (1991) Sero-epidemiological study of severe forms of paralytic poliomyelitis. Indian Pediatrics, 28:619–624. Kim-Farley RJ et al. (1984) Outbreak of paralytic poliomyelitis, Taiwan. Lancet, 2:1322–1324. Kumar R, Singh A, Kumar V (1991) Survey of village informants: an alternate method to estimate paralytic poliomyelitis prevalence in rural area. Indian Journal of Pediatrics, 58:239–243. LaForce FM et al. (1980) Clinical survey techniques to estimate prevalence and annual incidence of poliomyelitis in developing countries. Bulletin of the World Health Organization, 58:609–620. Lasch EE et al. (1984) Successful results of a program combining live and inactivated poliovirus vaccines to control poliomyelitis in Gaza. Reviews of Infectious Diseases, 6(Suppl 2):S467–470. Magnus HV, Peterson I (1984) Vaccination with inactivated poliovirus vaccine and oral poliovirus vaccine in Denmark. Reviews of Infectious Diseases, 6(Suppl 2):S471–474. Mahadevan S et al. (1989) Poliomyelitis: 20 years: the Pondicherry experience. Journal of Tropical Medicine and Hygiene, 92:416–421. Mandke VB, Pawar RM, Naik DD (1991) Epidemiology of poliomyelitis in a slum of Bombay. Indian Pediatrics, 28:615–618. Maru M, Getahun A, Hoshna S (1988) Prevalence of paralytic poliomyelitis in rural and urban populations in Ethiopia: report of a house-to-house survey. Journal of Tropical Medicine and Hygiene, 38:633–635. Melnick JL (1990) Poliomyelitis. In: Tropical and geographical medicine, 2nd ed. AAF Mahmoud, KS Warren, eds. New York, McGraw-Hill, pp. 559–576. Melnick JL (1978) Advantages and disadvantages of killed and live poliomyelitis vaccines. Bulletin of the World Health Organization, 56:21–38. Melnick JL (1988) Vaccination against poliomyelitis: present possibilities and future prospects. American Journal of Public Health, 78:304–305. Melnick JL, Ledinko N (1951) Social serology: antibody levels in a normal young population during an epidemic of poliomyelitis. American Journal of Hygiene, 54:354–382. Montefiore D et al. (1963) Trial of type 1 oral poliomyelitis vaccine (Sabin) in Nigerian children. British Medical Journal, 1:1569–1572. Poliomyelitis 147 Moore M et al. (1982) Poliomyelitis in the United States, 1969–1981. Journal of Infectious Diseases, 146:558–563. Murray CJL, Lopez AD (1996) The global burden of disease. Cambridge, Harvard University Press. Musgrove P (1988) Is polio eradication in the Americas economically justified? Bulletin of the Pan American Health Organization, 22:1–16. Nathanson N, Martin JR (1979) The epidemiology of poliomyelitis: enigmas surrounding its appearance, epidemicity, and disappearance. American Journal of Epidemiology, 110:672–692. Nicholas DD et al. (1977) Is poliomyelitis a serious problem in developing countries? – the Danfa experience. British Medical Journal, 1:1009–1012. Ofosu-Amaah S, Kratzer JH, Nicholas DD (1977) Is poliomyelitis a serious problem in developing countries? – Lameness in Ghanaian schools. British Medical Journal, 1:1012–1014. Onadeko MO, Familusi JB (1990) Observation on the age and spatial distribution of paralytic poliomyelitis in Ibadan, Nigeria. Annals of Tropical Paediatrics, 10:133–138. Otten MW, et al. (1992) Epidemic poliomyelitis in The Gambia following the control of poliomyelitis as an endemic disease. I. Descriptive findings. American Journal of Epidemiology, 135:381–392. Pan American Health Organization (1988) Practical guide for the eradication of poliomyelitis. Washington, D.C, Pan American Health Organization. Technical Paper No. 6. Pan American Health Organization (1993) Expanded Programme on Immunization. – Strategies for the certification of the eradication of wild poliovirus transmission in the Americas. Bulletin of the Pan American Health Organization, 27:287–296. Pan American Health Organization (1994) Polio eradication field guide. 2nd ed. Washington, DC, Pan American Health Organization. Technical Paper No. 40. Patriarca PA (1993) Poliomyelitis in selected African and Asian countries. Public Health Review, 21:91–98. Patriarca PA, Sutter RW, Oostovogel PM (1997) Outbreaks of paralytic poliomyelitis, 1976–1995. Journal of Infectious Diseases, 175(Suppl 1):S165–172. Patriarca PA, Wright PF, John TJ (1991) Factors affecting the imuunogenicity of oral poliovirus vaccine in developing countries: review. Reviews of Infectious Diseases, 13:926–939. Paul JR (1955) Epidemiology of poliomyelitis. In Debre R., ed. Geneva, World Health Organization, pp. 9–30. WHO Monograph Series No. 26. Paul JR et al. (1952) A survey of neutralizing antibodies to poliomyelitis virus in Cairo, Egypt. American Journal of Hygiene, 55:402–413. Prabakar N et al. (1981) The epidemiology and prevention of poliomyelitis in a rural community in south India. Indian Pediatrics, 18:527–537. Reichler MR et al. (1997) Increased immunogenicity of oral poliovirus vaccine administered in mass vaccination campaigns compared with the routine vac- 148 Global Epidemiology of Infectious Diseases cination program in Jordan. Journal of Infectious Diseases, 175(Suppl 1): S198–204. Richardson G et al. (1995) Immunogenicity of oral poliovirus vaccine administered in mass campaigns versus routine immunization programmes. Bulletin of the World Health Organization, 73:769–777. Robbins FC, de Quadros CA (1997) Certification of the eradication of indigenous transmission of wild poliovirus in the Americas. Journal of Infectious Diseases, 175(Suppl 1):S281–285. Robertson SE et al. (1990) Worldwide status of poliomyelitis in 1986, 1987 and 1988, and plans for its global eradication by the year 2000. World Health Statistics Quarterly, 43:80–90. Sabin AB, Silva E (1983) Residual paralytic poliomyelitis in a tropical region of Brazil, 1969–1977: prevalence surveys in different age groups as indicators of changing incidence. American Journal of Epidemiology, 117:193–200. Samuel R, Balraj V, John TJ (1993) Persisting poliomyelitis after high coverage with oral poliovaccine. Lancet, 341:903. Satcher D (1999) From the Surgeon General. Polio eradication by the year 2000. Journal of the American Medical Association, 281:221 Sen S, Sharma D, Santhanam S. (1985) Clinical and serological study of acute paralytic poliomyelitis in immunized and nonimmunized children. Indian Pediatrics, 22:891–897. Shepard DS, et al. (1989) Cost-effectiveness of routine and campaign vaccination strategies in Ecuador. Bulletin of the World Health Organization, 67:649–662. Snyder JD et al. (1981) Prevalence of residual paralysis from paralytic poliomyelitis in a rural population of Bangladesh. American Journal of Tropical Medicine and Hygiene, 30:426–430. Sutter RW et al. (1991) Outbreak of paralytic poliomyelitis in Oman: evidence for widespread transmission among fully vaccinated children. Lancet, 338:715–720. Tangermann RH, et al. (1997) Current status of the global eradication of poliomyelitis. World Health Statistics Quarterly, 50:188–194. Tangermann RH, et al. (2000) Eradication of poliomyelitis in countries affected by conflict. Bulletin of the World Health Organization, 78:330–338. Thuriaux MC (1982) A prevalence survey of lower limb motor disorders in schoolage children in Niger and an estimation of poliomyelitis incidence. Tropical and Geographical Medicine, 34:163–168. Tulchinsky T et al. (1989) A ten-year experience in control of poliomyelitis through a combination of live and killed vaccines in two developing areas. American Journal of Public Health, 79:1648–1652. Ward N, et al. (1993) A global overview and hope for the eradication of poliomyelitis by the year 2000. Tropical and Geographical Medicine, 45:198–202. Wise J (1999) Unicef warns of polio risk in conflict regions. British Medical Journal, 319:214. World Bank (1993) World development report 1993:investing in health. New York, Oxford University Press. Poliomyelitis 149 World Health Organization (1982) Poliomyelitis surveillance. Weekly Epidemiological Record, 57:179, 195, 218. World Health Organization (1994) Progress toward the global eradication of poliomyelitis. Geneva, World Health Organization. WHO/GPV/POLIO/94.1 World Health Organization (1998). Acute flaccid paralysis (AFP) surveillance: the surveillance strategy for poliomyelitis eradication. Weekly Epidemiological Record, 73:113–120. World Health Organization (1999) Progress towards global poliomyelitis eradication, as of May 1999. Weekly Epidemiological Record, 74:165–170. World Health Organization (2000) Certification of poliomyelitis eradication: WHO Western Pacific Region, October 2000. Weekly Epidemiological Record, 75: 399–400. Wright PF, et al. (1991) Strategies for the global eradication of poliomyelitis by the year 2000. New England Medical Journal, 325:1774–1779. Chapter 6 Tetanus A Galazka, M Birmingham, M Kurian, FL Gasse Introduction Tetanus, although vaccine-preventable, remains a common disease in many developing countries. In these countries, tetanus affects mainly younger persons, especially newborns. Neonatal tetanus can be considered as a paradigm of underdevelopment: each case demonstrates multiple failures in the health system. Each is a result of failure to protect the mother with tetanus toxoid, coupled with unhygienic practices which bring tetanus spores into contact with the umbilical stump. For a long time, the persistence of this disease was attributable, at least in part, to a lack of awareness by health authorities about the true magnitude of this problem. In industrialized countries, the incidence of tetanus decreased considerably with rising standards of living and effective primary health care services that include routine immunization services for children, adolescents and adults, and modern midwifery. However, elderly persons not vaccinated during their youth are still victims of tetanus. In this chapter we outline the extent of tetanus in developing and industrialized countries, and discuss relevant clinical and epidemiological aspects of this disease. Such knowledge is essential for adjusting and strengthening control strategies. Causative agent and pathogenesis of tetanus The tetanus bacillus, Clostridium tetani, is the causative agent of tetanus. C. tetani is an obligate anaerobic bacillus which has the ability to form spores resistant to various disinfectants and to heat. Boiling for 20 minutes does not destroy it. For practical purposes, autoclaving at 120ºC for 15 minutes is the best method of sterilizing contaminated materials. Spores are ubiquitous; they are present in soil samples, house and operating room dust, fresh and salt water, and the gastrointestinal tract of humans and other animals. 152 Global Epidemiology of Infectious Diseases C. tetani produces disease through a potent neurotoxin, tetanospasmin. The clinical syndrome is caused entirely by tetanospasmin produced by the vegetative forms of C. tetani. The bacillus is usually introduced into an area of injury in the form of spores. Spores convert to the vegetative forms, multiply and produce toxin only when they find anaerobic conditions—i.e. when they are placed in an area of low oxygen availability—which are produced by necrosis, or bacterial or chemical changes in a wound. This explains the association of non-neonatal tetanus with puncture wounds, lacerations and burns contaminated with soil, street dust, or animal or human faeces. Tetanus can also follow induced abortions, chronic otitis (common in India), administration of non-sterile injections (particularly with necrotizing materials such as intramuscular quinine), body piercing, scarification rituals, insect bites (bees, scorpions, etc.) and other unnoticed, minor and trivial injuries. It can also occur as a postoperative complication. Neonatal tetanus may occur during delivery, through the introduction of tetanus spores by cutting the umbilical cord with an unclean instrument, or after delivery, by dressing the umbilical stump with substances heavily contaminated with tetanus spores. Contaminated by dirt or other foreign substances, the necrotized and infected umbilical stump is an ideal place for the growth of C. tetani and the production of tetanus toxin. Toxin produced in the place of injury travels along nerves towards the central nervous system (CNS) by retrograde intra-axon transport. The severity of the disease is determined by the amount of tetanospasmin produced. In the CNS, tetanus toxin becomes fixed in the ganglion cells of the anterior horn of the spinal and cranial nerves. Tetanus toxin causes increased irritability of the CNS and exaggerated motor activity. It disinhibits spinal cord reflex arcs, presumably by interfering with release of inhibitory neurotransmitters. Freed from inhibition, excitatory reflexes lead to multiple spasms. Clinical manifestations, diagnosis and treatment There are considerable differences in the epidemiology, clinical picture, morbidity and mortality rates of the two main age groups of tetanus (neonatal and non-neonatal), as well as their predisposing causes and prevention. Clinical course of neonatal tetanus The clinical course of neonatal tetanus is well known to people in areas where the disease is prevalent. As early as 1764, the Reverend Kenneth MacAuley of St Kilda, a remote island off the Atlantic coast of Scotland, whose population was progressively decimated by neonatal tetanus in the 18th and 19th centuries, gave an excellent description of the disease (Collacot 1981, Woody & Ross 1989). In some developing countries, the disease is so common it has descriptive names such as “sickness with Tetanus 153 inability to breast feed,” “convulsion sickness,” or “stiff body sickness” (Marshall 1968). Numerous reports have described the clinical course of the disease (Jellife 1950, Ogbeide 1966, Athavale et al. 1975, Handalage & Wickramasinghe 1976). The first symptom noticed by mothers on the 3rd to 10th day of life is the child’s excessive crying and inability to suck at the breast after developing normally for the few first days. The baby wants to feed but cannot because of trismus (or “lockjaw”), a spasm of the jaw muscles (masseters) that prevents the proper positioning of the baby’s lips and the rhythmic contraction of the appropriate muscles necessary for effective suckling. Within hours after the first symptoms the baby develops generalized stiffness of the body and begins to have spasms. Trismus and risus sardonicus are the symptoms most often noted upon hospital admission of neonatal tetanus cases. These two symptoms are characterized by a clenched jaw, raised eyebrows and lips drawn laterally and upwards—the whole appearance producing the risus sardonicus facial expression. In neonates, sometimes the lips are pursed in a whistling expression (Ogbeide 1966). Tetanus spasms are infrequent at first, but increase in frequency and often are precipitated by a stimulus such as touch, light or noise. Spasms may last from a few seconds to over a minute. During spasms breathing is affected and the child may become cyanotic or pale. Children may die from severe apnoea during spasms, or 2 to 4 days later as a result of aspiration pneumonia or acute gastroenteritis. During spasms, both arms are flexed at the elbows and held in the front of chest. The legs are drawn up and flexed at the knees. The feet are dorsiflexed and the toes acutely flexed. This hyperflexion of toes, a very marked feature, is indicative of severe generalized rigidity and of hypertonia in the small muscles of the soles of the feet (Athavale & Pai 1965). The neck is somewhat retracted and there is a marked stiffness of abdominal and back muscles. In severe cases, spasms of the erector muscles of the spine cause the baby’s back to be arched backward (opisthotonos). Clinical course of non-neonatal tetanus The incubation period of non-neonatal tetanus varies from 3 to 21 days with extremes of 1 day to several months and a median of seven days (Wassilak, Orenstein & Sutter 1997). Tetanus may appear in one of three clinical forms: localized, cephalic, or generalized. In the localized form, there is persistent, painful rigidity of the group of muscles in close proximity to the site at which C. tetani was introduced. Symptoms may persist for several weeks or months and finally disappear, without leaving any sequelae, or progress to generalized tetanus. Local tetanus is usually mild, and has a case-fatality ratio of about 1 per cent. Cephalic tetanus is rare and is usually associated with injury of the head, eye, face, ear (e.g. chronic otitis media), or neck. Clinical features are palsies of cranial nerves III, IV, VII, IX, X, and XI. The disease has a short incubation period and may progress to generalised tetanus. 154 Global Epidemiology of Infectious Diseases Generalized tetanus is the most common form of the disease. It is characterized by painful muscular contractions, primarily of the masseters (i.e. trismus) and neck muscles. A common first sign is abdominal rigidity, although other groups of muscles may be involved, usually accompanied by severe pain. Generalized spasms occur, frequently induced by external stimuli such as touch, light or noise. The case fatality ratios are highest in infants and the elderly, and vary inversely with the length of the incubation period and the quality of medical care received (Einterz & Bates 1991). Diagnosis The diagnosis of tetanus is established primarily on clinical grounds and secondarily on epidemiological grounds. A history of a wound contaminated by soil or other material, or a history of unhygienic delivery are helpful in the diagnosis. Usually diagnosis is made clinically by excluding other possibilities. Attempts at laboratory confirmation are of little help; the organism is rarely recovered from the site of infection and usually there is no detectable antibody response. Treatment Management of generalized tetanus aims to: Neutralize any toxin still present in the blood before it comes into contact with the nervous system. This is accomplished by the prompt administration of tetanus antitoxin. Prevent further toxin production by eliminating the organism at the infected site. This can be done by the thorough debridement of wounds, excising all devitalised tissue and removing all foreign bodies. Antibacterial therapy with antibiotics is routinely used to reduce the number of vegetative forms of the organism. Provide constant and meticulous nursing and medical care. Along with assisted ventilation, many tetanus patients are managed with heavy sedation, muscle relaxants and curare-like drugs, and drugs that reduce the hyperactivity of the autonomic nervous system. Closely monitor fluid, electrolyte, and caloric balance, especially in neonates and other patients with repeated seizures or inability to take food or liquids because of severe trismus, dysphagia or hydrophobia. Prevent recurrence through vaccination. Since the disease does not stimulate the development of immunity against further attacks, every patient should be given an injection of tetanus toxoid followed by a second dose after at least 4 weeks and a third dose 6 to 12 months after the second dose. Tetanus 155 Mortality related to tetanus Case-fatality ratios The prognosis of tetanus is influenced by various factors, including duration of incubation or onset (a function of the amount and rapidity of tetanus toxin production), age, immunization status, underlying illness, delay-time from onset to treatment and quality of supportive care (Marshal 1968, Sutter, Overstein & Wassilak 1994). The highest case-fatality ratios (CFRs) occur at extreme ages: neonates and elderly persons (Table 6.1). Three main causes of death from tetanus include: deaths that occur from asphyxiation attributable to a muscle spasm resulting in prolonged hypoventilation (57 per cent); deaths attributable to metabolic troubles such as dehydration, hypoglycaemia, hyponatriaemia and hypocalcaemia (20 per cent); and deaths related to superinfection (23 per cent) (Sow, Coll & Diop-Mar 1985). Case fatality ratios for tetanus reported through national surveillance systems in industrialized countries are varied. In the past, CFRs for general tetanus exceeded 40 per cent, but have declined over time to as low as 10 per cent as a result of improved care (Edmonston & Flowers 1979). In Denmark, CFRs for non-neonatal tetanus decreased from 80 per cent in the early 1900s to 38 per cent in the 1930s to 13–27 per cent in the 1960s to 9–11 per cent during the 1970s and early 1980s (Simonsen, Block & Heron 1987) (Figure 6.1). Data from other national tetanus surveillance systems indicate CFRs of 11 per cent in Finland (Luisto 1989), 17–30 per cent in Spain, Switzerland (Zuber et al. 1993) and the United States of America (Prevots et al. 1992); and 39–43 per cent in France (Pelletier & Roure 1991) and Poland (Kuszewski & Dutkiewicz 1994). The large majority of tetanus cases reported from these surveillance systems were non-neonatal. Table 6.1 Case fatality ratios in tetanus patients by age, India 1954– 1968 and 1975, Nigeria 1974–1984 Age < 1 month 1 month–4 years 5–9 years 10–19 years 20–29 years 30–39 years 40–49 years 50–59 years > 60 years India 1954–68a India 1975b Nigeria 1974–84c 86.4 32.8 — 33.7 61.5 45.9 46.8 49.3 53.4 74.4 50 37.8 50 52.9 56.7 46.2 70.6 — — — — 46.7 35.8 30.9 43.6 66.7 85.7 a. Patel and Mehta (1975) b. Ghosh (1984) c. Bandele, Akinyan & Bojowoye et al. (1991) 156 Global Epidemiology of Infectious Diseases Neonatal tetanus incidence and mortality rate (per 10 000 live births) Figure 6.1 Incidence and mortality rates and case fatality ratios for neonatal tetanus, Denmark, 1920–1960 8 97%* 97%* Mortality rate Incidence rate 7 6 87%* 5 87%* 67%* 63%* 4 40%* 3 2 40%* 1 0 42%* 25%* 1920– 1925– 1930– 1935– 1940– 1945– 1950– 1955– 1960– 1965– 1924 1929 1934 1939 1944 1949 1954 1959 1964 1969 Year * case fatality ratio Source: Simonsen, Block & Heron, 1987 Reported CFRs associated with neonatal tetanus from commmunity-based studies in developing countries range from 25–100 per cent (Table 6.2). Neonatal tetanus CFRs reported from health facilities ranged from 11–100 per cent. Facility-based data are, however, difficult to interpret since they are affected by health care seeking behaviour, and hospitalized patients, particularly those who are moribund, may be discharged or removed from the hospital by family members before their outcome is recorded. Modern therapeutic methods such as artificial respirators and neuromuscular blocking agents are often unavailable in health facilities in developing countries; nevertheless, implementation of simple therapeutic schemes and good basic nursing care have been reported to reduce neonatal tetanus CFRs to between 24 per cent and 63 per cent (Daud, Mohammad Ahmad 1981, Einterz & Bates 1991). Maternal tetanus is associated with unhygienic conditions primarily during home deliveries and abortion procedures, and is estimated to account for 5 per cent of all maternal mortality (Faveau et al. 1993). Maternal tetanus is mentioned as the cause in 1–15 per cent of all non-neonatal tetanus illness or death (La Force, Young & Bennet 1969, Patel & Mehta 1999, Yadav, Kala & Yadav 1990, World Health Organization 1991). In an urban hospital in South Africa, 35 per cent of all tetanus patients among women between the ages of 15 and 50 years were admitted after abortions; the incidence of post-abortion tetanus was 1 in 800 cases of abortion (Bennett 1976). In Lagos Teaching Hospital in Nigeria, 82 per cent of 27 maternal tetanus cases were post-abortal (Adadevoh & Akinla 1970). In a series of 503 adult tetanus cases seen at a hospital in Ibaden, Niger, Tetanus Table 6.2 157 Case fatality ratios associated with tetanus, 1920–1995 Country Year(s) to which data pertain Type of tetanusa Case fatality ratio (%) Reference Community-based studies Colombia Mexico Nigeria Nigeria Nigeria Pakistan Sudan Viet Nam 1961 1988–1989 1991–1992 1988 1990 1986–1987 1981 1989 NT NT NT NT NT NT NT NT 95 25 50 80 40–90 1000 74 80 Newell et al. 1966 Ayala et al. 1995 Abuwa et al. 1997 Babaniyi & Parakoyi 1991 Eregie 1993 Khan & Hardjotanajo 1989 Olive, Gadir El Sid & Abbas 1982 Thanh et al.1990 India Nigeria, Lagos South Africa, Cape Town India Nigeria, Lagos Senegal, Dakar 1984–1987 1963–1967 1965–1972 1984–1987 1963–1967 1982–1983 MT, PA MT, PA MT, PA MT, PP MT, PP MT, PP 70 50 16 47 20 50 Yadav, Kala & Yadav 1991 Adadevoh & Akinla 1970 Bennett, 1976 Yadav,Yadav & Kala 1991 Adadevoh & Akinla 1970 Diop-Mar et al. 1985 Denmark Finland France Greece 1978–1982 1969–1985 1990 1966–1977 NNT NNT NNT NNT 9 11 43 32 Haiti India India, Bombay 1959–1966 1985–1987 1954–1962 NNT NNT NNT 30 31 33 India, Bombay Italy Italy Italy Italy Italy Switzerland Tanzania, Dar es Salam Turkey United Kingdom, Leeds United States United States United States United States 1954–1968 1956–1960 1961–1965 1966–1970 1971–1975 1976–1980 1980–1989 1970 1977–1991 1961–1977 NNT NNT NNT NNT NNT NNT NNT NNT NNT NNT 86 72 52 47 64 70 30 20 28 10 Simonsen, Block & Heron 1987 Luisto 1989 Pelletier & Roure 1991 Teknetzi, Manios S & Katsouyanopoulos V 1983 Marshall 1968 Yadav YR, Kala & Yadav 1990 Vakil BJ, Tulpule TH & Rananvare MM 1965 Patel & Mehta 1975 Rosmini et al. 1987 Rosmini et al. 1987 Rosmini et al. 1987 Rosmini et al. 1987 Rosmini et al. 1987 Zuber et al. 1993 Ayim 1972 Tutuncuoglu et al. 1994 Edmonsen & Flowers 1979 1965–1966 1947 1976 1989–1990 NNT NNT NNT NNT 68 91 44 24 La Force,Young & Bennett 1969 Prevots et al. 1992 Prevots et al. 1992 Prevots et al. 1992 Bangladesh Benin 1994–1995 1977–1986 NT NT 55 29 Kalter et al.1999 Ayivi et al. 1992 Facility-based studies continued 158 Table 6.2 Global Epidemiology of Infectious Diseases Case fatality ratios associated with tetanus, 1920–1995 (continued) Country Year(s) to which data pertain Type of tetanusa Brazil Denmark Denmark Denmark Denmark Denmark Denmark Denmark Denmark Denmark Egypt Egypt Germany, Dresden Ghana Haiti, rural Haiti, rural 1964 1930–1950 1920–1930 1930–1940 1940–1945 1945–1950 1950–1955 1955–1960 1960–1965 1965–1970 1988 1980 1957 1971–1974 1953 1967 NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT 77 85 97 87 67 63 46 40 42 25 66 78 60 64 53 1000 India 1987–1988 NT 82 India 1989 NT 63 1985–1986 1973–1974 1970 1964–1970 1967–1970 1945–1965 1956–1963 1959–1972 1954–1968 1961–1964 1960–1973 1973–1982 1984–1985 1940–1982 1974–1977 1968–1972 1970–1979 1980–1990 1970–1990 1976–1982 1989 1983–1986 NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT 97 72 57 88 57 91 73 77 86 59 39 43 51 80 11 21 47 13 25 65 33 51 India India India India India India, Bikaner India, Bombay India, Bombay India, Bombay India, Hyderabad Indonesia Iran Iraq Japan Malaysia, Kuala Lumpur Malaysia, west Mexico Mexico Mexico Mozambique, Maputo Nigeria Nigeria Case fatality ratio (%) Reference Pinheiro 1964 Christensen 1970 Simonsen, Block & Heron 1987 Simon, Block & Heron 1987 Simonsen, Block & Heron 1987 Simonsen, Block & Heron 1987 Simonsen, Block & Heron 1987 Simonsen, Block & Heron 1987 Simonsen, Block & Heron, 1987 Simonsen, Block & Heron 1987 El-Sherbini 1991 Riyad & El Gereidy 1983 Marshall 1968 Blankson 1977 Marshall 1968 Berggren, Ewbank & Berggren 1981 Deivanayagam, Nedunchelian & Kamula 1991 Deivanayagam, Nedunchelian & Kamula 1991 Verma, Dhanwade & Singh 1989 Bhat, Joshi & Kandoth 1979 Mazumdar & Chakraborthy 1974 Phatak & Shah 1973 Majumdar & Kaur S 1971 Saxena & Saxena 1965 Athavale & Pai 1965 Athavale et al. 1975 Patel & Mehta 1975 Jaffari, Pandit & Ismail 1966 Barten 1973 Hasanabadi 1987 Al Mukhtar 1987 Ebisawa & Homma 1986 Khoo, Lee & Lam 1978 Chen 1974 Simental et al. 1993 Simental et al. 1993 Simental et al. 1993 Cliff 1985 Antia-Obong & Ikpatt 1991 Antia-Obong & Ikpatt 1991 continued Tetanus Table 6.2 159 Case fatality ratios associated with tetanus, 1920–1995 (continued) Country Nigeria Nigeria Nigeria Nigeria Nigeria Nigeria Nigeria Nigeria, Benue Pakistan Pakistan Pakistan Senegal Senegal Senegal Sierra Leone Singapore Singapore Singapore South Africa South Africa South Africa Sri Lanka Sudan, Juba Turkey Turkey Uganda United States United States United States, New Orleans United States, Southern Region United States, Texas Venezuela Venezuela Venezuela Year(s) to which data pertain Type of tetanusa Case fatality ratio (%) Reference 1986–1988 1975–1977 1978–1980 1962–1964 1953–1956 1953–1956 1950 1984–1989 1979–1980 1979–1980 1982 1980 1984 1992–1993 1955–1960 1946–1949 1960–1970 1946–1950 1953–1955 1956–1959 1967–1972 1971–1974 1983 1977–1991 1978–1988 1985–1989 1953–1961 1965–1966 1946–1951 NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT 43 63 34 68 89 90 96 65 48 44 38 69 62 53 73 91 77 90 97 83 21 60 84 53 42 78 62 77 77 Owa & Makinde 1990 Okoro & Okeahialam 1987 Okaro & Okeahialam 1987 Ogbeide 1966 Tompkins 1958 Miller 1972 Jelliffe 1950 Einterz & Bates 1991 Daud, Mohammad & Ahmad 1981 Akbar 1981 Waheed et al. 1981 Sow et al.1985 Sow, Coll & Diop-Mar 1985 Sow, Coll & Diop-Mar 1995 Wilkinson 1961 Miller 1972 Chen 1973 Gek 1951 Klenerman & Scragg 1955 Wright 1960 Smythe, Bowie & Voss 1974 World Health Organization 1977 Woodruff et al. 1984 Tutuncuoglu et al. 1994 Gurses & Aydin 1993 Bwire & Kawuma 1992 Bytchenko 1966 La Force,Young & Bennett 1969 Spivey, Grulee & Hickman 1953 1957 NT 77 Axnick & Alexander 1957 1954 1971–1975 1971–1975 1970–1992 NT NT NT NT 42 42 42 61 Box 1964 World Health Organization 1978 World Health Organization 1978 World Health Organization 1993a a. NT = neonatal tetanus; NNT = non-neonatal tetanus; MT = maternal tetanus, PA = post-abortal, PP = post-partum the uterus was the portal of entry in 72 (14 per cent) cases of which 20 (28 per cent) were post-abortal (Adeuja & Osuntokun 1971). High CFRs have been reported for maternal tetanus (Bennett 1976, Adadevoh & Akinla 1970, La Force, Young & Bennett 1969, Patel & 160 Global Epidemiology of Infectious Diseases Mehta 1999, Yadav, Yadav & Kala 1991). In a review of 1 101 cases of maternal tetanus from 60 studies in developing countries between 1958 and 1990, a median CFR of 52 per cent (range: 16–80 per cent) was found in Africa and 54 per cent (range: 42–64 per cent) in Asia (Faveau et al. 1993). An earlier review of seven studies of postpartum and post-abortal tetanus reported CFRs ranging from 65–72 per cent (Bytchenko 1966). CFRs were generally higher for post-abortal tetanus than for postpartum tetanus in developing countries. Reviews of maternal tetanus (both postpartum and post-abortal) in Europe and the USA before 1960 showed CFRs between 57 per cent and 100 per cent (Adams & Morton 1955, Weinstein & Beacham 1941). Complications and sequelae after tetanus It is generally thought that patients who recover from tetanus do not have sequelae and that neurological complications are rare (Marshall 1968). However, a careful analysis of the literature shows that residua after tetanus may have an important practical significance when follow-up periods cover several months to several years after the acute phase of the disease. The complications of tetanus may be divided into two categories—those induced by the toxin itself and those resulting from debility (Sutter, Orenstein & Wassilak 1994). Toxin-related complications Spasm-related complications include fractures of the long bones and vertebrae, asphyxia from obstruction of the glottis, spontaneous rupture of muscles, intramuscular haematomas, and traumatic glossitis. Vertebral fractures should be considered among the most common complications of tetanus. They are recognized rarely because of their poor clinical expression; local signs and symptoms are uncommon even when the vertebral injury is extensive (Chaubey 1965, Januszkiewicz 1975, Veronesi, Vitule Filho & Ferrarceto 1975). In some cases there may be some mobility limitations (Yurchuk & Luchko 1989) but even in the presence of some pain, movements of the spine (flexion, extension and rotation) are often impaired very little. The diagnosis is based mainly on X-ray examination. Vertebral fractures are not seen after neonatal tetanus but they are frequent in children and adults. The extent of vertebral injury is directly related to the severity of tetanus, and especially to the degree of rigidity of dorsal muscles (Chaubey 1965, Januszkiewicz 1975). The fourth, fifth and sixth thoracic vertebrae are most commonly affected; multiple fractures are common (Chaubey 1965, Patel, Mehta & Goodluck 1965, Januszkiewicz 1975, Veronesi, Vitule Filho & Ferraceto 1975). The vulnerability of these three vertebrae may be attributable to the powerful muscles operating in an area of minimal flexibility, causing excessive outward curvature of the spine (i.e. kyphosis). The reported frequency of post-tetanus vertebral fractures varies: from 48 Tetanus 161 per cent (Januszkiewicz 1975), to 51 per cent (Chaubey 1965) to 57 per cent (Patel, Mehta & Goodluck 1965) to 78 per cent (Veronesi, Vitule Filho & Ferraceto 1975). Vertebral fractures may lead to kyphosis, “pectum carinatum,” and gibbous dorsal deformity (i.e. humpback) (Chaubey 1965, Veronesi, Vitule Folho & Ferraceto 1975). In one study, 16 patients (13 per cent) with degeneration-dystrophic changes in the spine (osteochondrosis, spondylitis) experienced difficulties in their jobs and had to be transferred to easier work; four of them were officially declared as invalids (Yurchuk & Luchko 1989). Bilateral anterior dislocation of the temporomandibular joint was reported as a direct effect of toxin-induced spasms (Thachil, Philip & Sridhar 1993). Tetanus toxin may also induce urinary retention, cardiac arrhythmia, hypotension, and hypertension (Sutter, Orenstein & Wassilak 1994). Cardiovascular, gastric, and renal complications usually worsen the prognosis. All these complications can be minimized by paying strict attention to supportive care and by instituting appropriate therapy. The effects of tetanus toxin on the autonomic nervous system are probably the main causes of different conditions seen in post-tetanus follow-up. Among 25 adult tetanus survivors who were followed up for 3 months to 11 years, various proportions of them had symptoms of irritability, sleep disturbances, seizures, clonic spasms, decreased libido, postural hypotension, and electroencephalographic abnormalities (Illis & Taylor 1971), but these symptoms were generally self-limiting. Complications attributable to debility Pneumonia, probably as a result of aspiration of the contents of the stomach during a spasm, occurs often in neonates. Pneumonia may also occur in the third or fourth week of the illness, possibly as a result of general debility. Other complications may include pulmonary embolism, decubitus ulcers, faecal impaction, catheter-associated infections, and muscle contracture (Sutter, Orenstein & Wassilak 1994). Neurological defects observed in survivors of neonatal tetanus, especially in those who suffered frequent and prolonged bouts of spasms and apnoea, have been attributed to brain damage caused by anoxia at a vulnerable period of brain growth and development (Anlar et al. 1989, Khoo et al. 1978, Teknetzi et al. 1983, Tutuncuoglu et al. 1994). In many of these children, mental or growth retardation and behavioural disturbances were observed from 2 to 15 years after recovery. Age distribution of tetanus There are considerable differences in the epidemiology, clinical picture, and the morbidity and mortality rates of neonatal and non-neonatal tetanus, as well as in their predisposing causes and prevention. The overall age distribution of tetanus has changed considerably, 162 Global Epidemiology of Infectious Diseases Figure 6.2 Reported age-specific tetanus incidence rates, by 5-year intervals, United States, 1965–1994 0.70 1965–1969 1970–1974 1975–1979 1980–1984 1985–1989 1990–1994 Average annual rate per 100 000 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0 <1 5–19 20–39 40–49 50–59 60+ Age group (years) Source: Wassilak, Orenstein & Sutter 1997 particularly in age groups covered by routine immunization. Figure 6.2 shows the progressive impact of tetanus immunization in the United States according to age. After 30–40 years of routine immunization, tetanus virtually disappeared in infants, children and teenagers and its incidence greatly decreased in adults, with the highest incidence rate among adults of 60 years of age and older. In other industrialized countries, tetanus has also become a rare disease in children and young adults, the age groups that were most frequently affected before the introduction of immunisation services. In the industrialized world, persons under 20 years of age constitute only 2–9 per cent of all tetanus cases, and the majority of tetanus cases occur in persons over 50 years of age (Table 6.3). In France, of the 30 cases reported in 1990, 22 cases (73 per cent) were over 70 years of age; the median age of patients was 77 years (Pelletier & Roure 1991). In the United States, by 1985, 70 per cent of reported tetanus cases occurred in persons aged 50 years and older (Centers for Disease Control and Prevention 1985). In England and Wales, 60 per cent of 123 reported tetanus cases between 1984 and 1992 were patients over the age of 65 years (Anonymous 1993). The age-specific shift in tetanus incidence parallels changes in tetanus immunity; the level of tetanus antibody is highest in persons under 30 years of age. Then immunity decreases with increasing age and falls to the lowest levels in the over-50 age group (Christenson and Bottiger 1987, Galazka and Kardymowicz, 1989, Bourleaud and Huet 1985, Misra and Rao 1988, Tetanus Table 6.3 163 Age distribution of tetanus cases (percentage of cases < 20 and > 50 years of age), 1954–1992 Country Year Type of Study < 20 > 50 years years Reference Developing countries Cameroon, Yaounde Chile 1971–72 Hospital study 23 18 Ancelle et al. 1974 1975–76 National surveillance 56 19 Chile 1982–83 National surveillance 28 27 India, Bombay Jamaica Nigeria, Ibaden Nigeria, Lagos Senegal Senegal Upper Volta Upper Volta 1954–68 1980–86 1963–69 1967–70 1960–67 1985–86 1968–72 1972–75 71 49 28 48 68 65 62 69 4 32 12a 8a 6 9 4 3a de la Fuente, Gonzalez & Guevara 1986 de la Fuente, Gonzalez & Guevara 1986 Patel & Mehta 1975 Figueroa & Clarke 1988 Adeuja & Osuntokun 1971 Afonja 1973 Rey & Tikhomorov 1989 Rey & Tikhomorov 1989 Galazka & Gasse 1995 Cosnard, Joullie & Davin 1977 0 85 Hospital study Hospital study Hospital study Hospital study Hospital study Hospital study Hospital study Hospital study Industrialized countries Denmark 1978–82 National surveillance England, Wales 1970 National surveillance 27b 50c England, Wales 1979 National surveillance 0b 55c 9d 1 54 69e National surveillance National surveillance National surveillance 46 National surveillance 15 National surveillance 3 National surveillance 2 Hospital study 0 National surveillance 3 National surveillance 4 93f 93f 26 53 81 86 64 81 71a Finland France 1969–85 Hospital study 1969–79 Hospital study France France Poland Poland Poland Poland Switzerland Switzerland United States 1990 1991–92 1960 1970 1980 1990 1968–89 1980–89 1982–84 United States 1985–86 National surveillance 5 71 United States 1987–88 National surveillance 6 68 United States 1989–90 National surveillance 6 64 a. includes 50-year-olds b. c. d. e. f. < 14 years > 45 years includes 20-year-olds < 15 years > 60 years Simonsen, Block & Heron 1987 Galbraith, Forbes & Tillett 1981 Galbraith, Forbes & Tillett 1981 Luisto 1989 Lamisse, Sonneville & Choutet 1981 Pelletier & Roure 1991 Lombard & Lepoutre 1993 Galazka & Abgarowicz 1973 Galazka & Abgarowicz 1973 Galazka & Gasse 1995 Galazka & Gasse 1995 Jolliet et al. 1990 Zuber et al. 1993 Centers for Disease Control and Prevention CDC 1985 Centers for Disease Control and Prevention CDC 1985 Centers for Disease Control and Prevention CDC 1985 Prevots et al. 1992 164 Global Epidemiology of Infectious Diseases Crossley et al. 1979, Lau 1987). Serosurveys from industrialized countries since 1977 indicate that 49–68 per cent of elderly persons lack protective immunity against tetanus (Beytout et al. 1988, CDC 1985, Crossley et al. 1979, Simonsen et al. 1987b, Kjeldsen et al. 1988). In developing countries, tetanus affects the younger segment of the population. A major portion of the burden is among neonates, and children or young adults up to the age 20 years (Table 6.3). Before implementation of routine immunization of pregnant women against tetanus and improvement of delivery conditions, neonatal tetanus was responsible for about half of all neonatal deaths (ranging from 23 per cent to 72 per cent) and for about 25 per cent of infant mortality (Stanfield & Galazka 1984, Galazka, Kardymowicz 1989). Four decades ago, a similar situation existed in what are now industrialized countries; in Denmark, neonatal tetanus accounted for more than 50 per cent of all tetanus deaths until the mid1950s (Simonsen, Block & Heron 1987) and in the 1930s and 1940s the incidence of neonatal tetanus was about five cases per 10 000 live births with an average case-fatality ratio of 80 per cent (Figure 6.1). As tetanus immunity levels increase in target groups covered by routine immunization—infants or young children of both sexes and women of child-bearing age—one may expect a considerable drop in incidence and a shift in age distribution of tetanus cases (Misra & Rao 1988, Morgan et al. 1981, Mya et al. 1985, Vernacchio et al. 1993). Data from Sri Lanka show a 36-fold and 4-fold decline in neonatal tetanus and non-neonatal tetanus incidence, respectively, over a 10-year period beginning in 1978 when national immunization services were introduced for both pregnant women and infants (Galazka & Kardymovicz 1989). Sex distribution of tetanus During the pre-immunisation era, reports from all regions of the world indicated a male predominance of tetanus cases, with the exception of the child-bearing age group (Ebisawa 1971, Wassilak, Orenstein, K Sutter 1997). A review of several studies indicated a male:female ratio of approximately 2:1 (Bychentko 1966). A male predominance is also reported for neonatal tetanus from many community-based studies (Table 6.4), suggesting that the male predominance of reported neonatal tetanus is not just a function of health care seeking practices (i.e. male babies are more likely to be brought to hospital than female babies). Although the male predominance of adult tetanus is considered to be largely related to occupation (e.g. agricultural practices), some authors have suggested that there may be biological factors; males may be more sensitive to tetanus toxin than are females (Bychentko 1966, Ebisawa 1971). Animal studies offer further evidence of a male predominance for tetanus (Bychentko 1966). Systematic immunzation of children and military recruits has been shown in many countries to reduce the incidence among children and adult males, Tetanus Table 6.4 165 Gender ratios of tetanus cases and deaths, 1944–1990 Country Year Male: Female Ratio Outcome Reference Community-based studies Bangladesh Ethiopia Pakistan 1990 1988–89 1988 1.5 2.5 1.4 cases cases cases Hlady et al. 1992 Alemu 1993 Traverso et al. 1989 Columbia Egypt India Côte d’Ivoire Pakistan Sudan 1961–66 1981 1957–59 1981–82 1981 1981 0.9 2.6 3.2 1.1 1.6 1.4 deaths deaths deaths deaths deaths deaths Newell et al. 1966 World Health Organization 1982b Gordon, Singh & Wyon 1961 World Health Organization 1983 Suleman 1982 Olive, Gadir El Sid & Abbas 1982 Egypt Egypt Ghana Greece 1980 1988 1971–74 1966–77 2.5 2.3 1.1 3.3 cases cases cases cases India India India India India, Bikaner Iran Mexico Mozambique Nepal Nepal Nigeria Singapore Sri Lanka Syrian Arab Republic United States United States, New Orleans 1956–63 1967–71 1970–71 1975 1945–65 1967–76 1970–90 1976–82 1975–77 1975–77 1953–56 1960–70 1972–74 1982–84 1965–66 1946–51 1.4 2.0 2.5 3.7 2.7 2.5 2.1 1.3 2.2 3.0 1.4 2.2 1.5 2.5 1.0 1.0 cases cases cases cases cases cases cases cases cases cases cases cases cases cases cases cases Riyad & El Geneidy 1983 El-Sherbini 1991 Blankson 1977 Teknetzi, Manios & Katsouyanopoulos 1983 Athavale & Pai 1965 Mazumder & Chakraborthy 1974 Parija & Mohanta 1973 Seghal, Gupta & Sidhu 1980 Saxena & Saxena 1965 Salimpour 1978 Simmental et al. 1993 Cliff 1985 Shreshta 1978 Manandhar 1978 Tompkins 1958 Chen 1973 Handalage & Wickramasinghe 1976 World Health Organization 1988 La Force,Young & Bennett 1969 Spivey, Grulee & Hickman 1953 India Japan Nigeria 1966–77 1967 1953–56 4.0 2.0 0.9 deaths deaths deaths Bildhaiya 1983 Ebisawa 1971 Tompkins 1958 Facility-based studies respectively, resulting in a relative increase in the proportional morbidity among women (Bychentko 1966, Ebisawa 1971). 166 Global Epidemiology of Infectious Diseases Weakness of tetanus surveillance Reporting of neonatal tetanus separately Until recently, many countries did not provide data to WHO on neonatal tetanus separately from tetanus in other age groups. However, following WHO recommendations to do so, separate reporting of neonatal tetanus by countries increased from 18 per cent in 1974 to 89 per cent in 1995. Notification efficiencies Both forms of tetanus, neonatal and non-neonatal, are grossly underreported diseases. Neonatal tetanus especially, with its “peculiar quietness” is often overlooked by public health authorities and is hidden within the community (Stanfield & Galazka 1984). The neonate often dies at home or fails to reach a hospital. Table 6.5 presents the estimated notification efficiency of tetanus reporting from the published literature. The WHO estimated that at the beginning of the 1980s routine reporting systems identified no more than 5 per cent of tetanus cases occurring worldwide (Stanfield & Galazka 1984). When notifications of tetanus cases to health authorities were compared with results of a survey of tetanus cases admitted to nine hospitals in Jamaica between 1980 and 1986, it was found that only 32 per cent of tetanus admissions were officially reported to health authorities (Figueroa & Clarke 1988). The level of underreporting was underestimated by this study as it excluded most of the small rural and private hospitals. In Switzerland, through a capture-recapture method, only 6–13 per cent of tetanus cases were reported to the Federal Office of Public Health; by contrast, mortality data reported by physicians on death certificates to the Swiss Federal Statistical Office were more complete at 81–100 per cent (Zuber et al. 1993). In a review of all tetanus cases from 1969–1978 in England and Wales, it was estimated that only 25 per cent of all tetanus cases were notified (Galbraith Forbes & Tillett 1981). A review of surveillance data from 1979–1984 in the United States showed that only 40 per cent of all tetanus deaths were reported to the US Centers for Disease Control and 60 per cent to the US National Center for Health Statistics, with a combined notification efficiency of 76 per cent by a capture-recapture method (Sutter et al. 1990). In Denmark, only 66 per cent of hospitalized tetanus cases were reported (Simonsen, Block & Heron 1987). Trends in tetanus morbidity reported to WHO Despite the poor notification efficiency of tetanus in most countries, the case reports are still useful to monitor trends. The global number of reported cases of tetanus decreased from 114 000 in 1980 to 64 000 in 1990, representing a 44 per cent decline in a 10-year period during which national immunization services were expanding rapidly in most developing countries (Figure 6.3). When neonatal tetanus and non-neonatal tetanus are presented separately, inverse patterns can be seen (Figure 6.3). The number of cases of reported non-neonatal tetanus was more than halved, from 101 Burundi Cameroon Cote d’Ivoire Egypt, Alexandria Egypt Egypt Ethiopia Gambia India India India India India Iran Kenya Malawi Mexico Somalia Sudan Thailand Togo Uganda 1981–1986 1981–1986 1981–1982 1980 1981–1986 1988 1981–1986 1981–1986 1989–1992 1981 1989 1992 1992 1986 1981–1986 1983 1988–89 1981–1986 1981–1986 1980 1981–86 1981–86 Year(s) to which data pertain Community-based survey Community-based survey Community-based survey Combined active/passive case finding Community-based survey Hospital record review Community-based survey Community-based survey Community-based survey Community-based survey Community-based survey Community-based survey Community-based survey Community-based survey Community-based survey Community-based survey Community-based survey Community-based survey Community-based survey Community-based survey Community-based surveys Community-based surveys Study type Notification efficiencies of tetanus cases, 1969–1992 Neonatal tetanus cases Country Table 6.5 <100 10–22 2 94 45 10 <100 <100 10 8 11 13 15 5 <100 6 0 <100 <100 14 10–22 <100 Notification efficiency (%) continued World Health Organization 1987a World Health Organization 1987a Sokal et al. 1988 Riyad & El Geneidy 1983 World Health Organization 1987a El-Sherbini et al. 1991 World Health Organization 1987a World Health Organization 1987a Singh, Datta & Foster 1997 Singh, Datta& Foster 1997 Singh, Datta & Foster 1997 Singh, Datta & Foster 1997 Singh,Datta & Foster 1997 Hasanabadi 1987 World Health Organization 1987a World Health Organization 1987a Ayala et al. 1995 World Health Organization 1987a World Health Organization 1987a World Health Organization 1993c World Health Organization 1987a World Health Organization 1987a Reference Tetanus 167 1982–1987 1969–1978 1988 1980–1989 1980–1989 1979–1984 1981–86 1981–86 Year(s) to which data pertain Hospital-based data Surveillance data Hospital-based data Surveillance data Surveillance data Surveillance data Community-based surveys Community-based surveys Study type 66 25 32 81–100a 6–13 76 <100 10–22 Notification efficiency (%) Notification efficiencies of tetanus cases, 1969–1992 (continued) a. notification efficiency is for tetanus deaths (not cases) Denmark England and Wales Jamaica Switzerland Switzerland U.S Total tetanus Zaire Zimbabwe Country Table 6.5 Simonsen, Block & Heron 1987a Galbraith, Forbes & Tillett 1981 Figueroa & Clarke 1988 Zuber et al. 1993 Zuber et al. 1993 Sutter et al. 1990 World Health Organization 1987a World Health Organization 1987a Reference 168 Global Epidemiology of Infectious Diseases Tetanus 169 Figure 6.3 Non-neonatal tetanus Neonatal tetanus Number of reported cases 140 000 100 90 120 000 80 100 000 70 60 80 000 TT2plus 60 000 50 b 40 30 40 000 20 20 000 10 95 94 19 93 19 92 19 91 19 90 19 89 19 88 19 87 19 86 19 85 19 84 19 38 19 82 19 81 19 80 19 79 19 78 19 77 19 19 19 19 76 0 75 0 Reported immunization coverage (%) Number of tetanus cases and immunization coverage reported to WHO by national authorities, 1975–1995 Year Source: WHO 2003 a Third dose of diptheria, tetanus and pertussis vaccine in infants b Two or more doses of tetanus toxoid in pregnant women in developing countries or in countries with economies in transition. 000 in 1980 to 39 000 in 1990. In contrast, the number of neonatal tetanus cases increased until 1988, probably reflecting a combination of improved notification efficiency within countries, more use of health facilities to treat cases, and better reporting by national authorities to WHO. The surge in reported cases between 1987–1988 is an artifact; Bangladesh and India did not provide data on tetanus cases to WHO in 1987, but did so in 1988. In 1996, neonatal tetanus cases were reported for the first time from China. Despite these artifacts, the reported data suggest that neonatal tetanus contributes an increasing proportion to the overall tetanus burden. In 1980 and 1990, reported neonatal tetanus cases accounted for 11 per cent and 39 per cent, respectively, of the total reported tetanus cases, compared to the 5–6 per cent contribution noted in the 1970s. Reported morbidity by regions Global figures hide variations among regions and countries. Some developing countries have demonstrated a spectacular impact of immunization on tetanus incidence (Galazka & Gasse 1995). The reported numbers of non-neonatal and neonatal tetanus cases from 1974 to 1996 in six WHO regions are presented in Figure 4. A decreasing trend in non-neonatal tetanus cases has been seen in all WHO regions. The trend for neonatal tetanus cases varies by Region. In 1994, 58 per cent of all tetanus cases and 38 per cent of neonatal tetanus cases were reported from India and other Asian countries (data from China are unavailable). 170 Global Epidemiology of Infectious Diseases Figure 6.4 Reported number of neonatal and non-neonatal tetanus cases by WHO region, 1974–1992 African region 25 000 Number of cases 20 000 15 000 10 000 5 000 92 91 19 90 19 89 19 88 19 87 19 86 19 85 19 84 19 38 19 82 19 81 19 80 19 79 19 78 19 77 19 76 19 75 19 19 19 74 0 Year Region of the Americas Number of cases 8 000 6 000 4 000 2 000 19 74 19 75 19 76 19 77 19 78 19 79 19 80 19 81 19 82 19 38 19 84 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 0 Year Eastern Mediterranean region 25 000 Number of cases 20 000 15 000 10 000 5 000 92 19 91 19 90 19 89 19 88 87 19 19 86 19 85 19 84 19 38 19 82 19 81 19 80 19 79 19 78 19 77 19 76 19 75 19 19 74 0 Year Non-neonatal tetanus Source: World Health Organization 1996 Neonatal tetanus continued Tetanus 171 Figure 6.4 Reported number of neonatal and non-neonatal tetanus cases by WHO region, 1974–1992 (continued) European region Number of cases 3 000 2 000 1 000 92 91 19 90 19 89 19 88 19 87 19 86 19 85 19 84 19 38 19 82 19 81 19 80 19 79 19 78 19 77 19 76 19 75 19 19 19 74 0 Year South-east Asia region 100 000 Number of cases 80 000 60 000 40 000 20 000 19 74 19 75 19 76 19 77 19 78 19 79 19 80 19 81 19 82 19 38 19 84 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 0 Year Western Pacific region Number of cases 8 000 6 000 4 000 2 000 Year Non-neonatal tetanus Source: World Health Organization 1996 Neonatal tetanus 92 19 91 19 90 19 89 19 88 87 19 19 86 19 85 19 84 19 38 19 82 19 81 19 80 19 79 19 78 19 77 19 76 19 75 19 19 74 0 Burma Afghanistan Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh Bhutan Bhutan Burkina Faso Community-based surveys Country Table 6.6 Bam, Sanmatenga, Namentenga districts EPI area Dhaka Chittagong Rajshahi division MCH area non-MCH area Kabul Matlab Matlab Teknaf Teknaf, Chittagong Area 1985 1989 1975–1977 1975–1977 1976–1977 1976–1977 1978 1986 1986 1986 1989 1990 1990 1993 1993–1994 1994 1995 1978–1983 1982 1989 Year(s) to which data pertain 15 19 19 30 3 13 18 3 6 25 22 89 48 45 54 82 12 37 37 27 27.4 27 3 11 41 10 13 18 2 Neonatal tetanus 19 Neonatal Mortality rates per 1000 live births 21 12 15 23 15 67 67 11 31 31 56 7 21 50 56 Percentage of neonatal deaths attributable to tetanus Stroh et al.1987 World Health Organization 1995 Chen, Rahman & Sarder 1980 World Health Organization 1995 World Health Organization 1995 Islam 1982 Galazka, Gasse & Henderson1989 Fauveau et al. 1990 Fauveau et al. 1990 Galazka, Gasse & Henderson1989 World Health Organization 1995 World Health Organization 1995 World Health Organization 1995 World Health Organization 1995 Baqui et al. 1998 World Health Organization 1995 Perry et al. 1998 Stanfield & Galazka 1984 World Health Organization 1982a Schwoebel et al. 1992 Reference Estimated neonatal mortality rates and neonatal tetanus mortality rates, and proportion of all neonatal deaths due to tetanus, 1953–1995 172 Global Epidemiology of Infectious Diseases Burma Burma Burundi Cameroon Cameroon China China China China Colombia Congo Côte d’Ivoire D Yemen Egypt Egypt Egypt Egypt Egypt Ethiopia Ethiopia Ethiopia Gambia Ghana Ghana India India India India India Rajasthan (rural) Meerut district Uttar Pradesh, rural Madhya Pradesh, rural Uttar Pradesh, urban Alexandria all Egypt urban rural (urban) North & South Omo Gondar Boundiali, Tengrela 300 counties 9 Provinces rural urban non-EPI areas 1985 1985 1984 1982 1984 1989 1989 1989 1989 1961–1966 1988 1981–1982 1978–1983 1980 1986 1986 1986 1981 1988–1989 1983 1988 1980 1989 1992 1993 1991 1981 1981 1981–1982 19 26 16 8 15 34 19 8 12 5 18 45 22 23 18 9 5 8 7 8 6 4 5 2 1100 2 18 4 4 7 2 10 3 7 5 5 11 7 2 17 5 67 20 15 33 21 72 36 59 29 40 63 15 58 20 63 54 53 54 13 19 41 30 Stroh et al. 1987 Stroh et al. 1987 Galazka & Kardymowicz 1989 Galazka & Kardymowics 1989 Galazka & Kardymowics 1989 World Health Organization 1993b World Health Organization 1993b World Health Organization 1993b World Health Organization 1993b Newell et al. 1996 Steinglass, Brenzel & Percy 1993 Sokal et al. 1988 Stanfield & Galazka 1984 Riyad & El Geneidy 1983 World Health Organization 1987b World Health Organization 1987b World Health Organization 1987b Galazka 1985 Alemu 1993 Galazka, Gasse & Henderson1989 Maru, Getahun & Hosana 1988 Galazka, Gasse & Henderson1989 World Health Organization 1995 World Health Organization 1995 Gupta & Keyl 1998 Garg et al. 1993 Basu & Sokhey 1984 Basu & Sokhey 1984 Galazka 1985 continued Tetanus 173 1981–1982 1981 1981–1982 1981–1982 1981–1982 1981–1982 1981–1982 1981 1981–1982 1981–1982 1981–1982 1981–1982 1981–1982 1981 1981–1982 1981–1982 1981–1982 1981–1982 Area Madhya Pradesh,urban West Bengal, rural Bihar, rural Orissa, rural Haryana, Punjab & Chandigarh, rural Andhra Pradesh, rural Gujarat & Dadra& Nagar Haveli, rural Bihar, rural Bihar, urban Karnataka & Goa, rural Tamil Nadu & Pondicherry, rural Maharashtra, urban Maharashtra, rural Rajasthan, rural Rajasthan, urban Haryana, Punjab & Chandigrarh, urban Andhra Pradesh, urban Orissa, urban India India India India India India India India India India India India India India India India India India Year(s) to which data pertain Neonatal 3 2 5 5 14 3 3 11 5 5 5 7 6 1 12 11 9 8 Neonatal tetanus Mortality rates per 1000 live births 18 13 37 16 42 22 26 38 65 24 31 28 17 30 40 38 23 30 Percentage of neonatal deaths attributable to tetanus Galazka 1985 Galazka 1985 Galazka 1985 Galazka 1985 Basu & Sokhey 1984 Galazka 1985 Galazka 1985 Basu & Sokhey 1984 Galazka 1985 Galazka 1985 Galazka 1985 Galazka 1985 Galazka 1985 Galazka 1985 Basu & Sokhey 1984 Galazka 1985 Galazka 1985 Galazka 1985 Reference Estimated neonatal mortality rates and neonatal tetanus mortality rates, and proportion of all neonatal deaths due to tetanus, 1953–1995 (continued) Country Table 6.6 174 Global Epidemiology of Infectious Diseases India India India India India India India India India Indonesia Indonesia Indonesia Indonesia Indonesia Indonesia Indonesia Indonesia Indonesia Indonesia Indonesia India India India India India India India Aceh Province Pidie District Jawa Barat central Java rural Jakarta 19 of 27 provinces urban Goa, urban Delhi, urban West Bengal, urban Tamilnadu & Pondicherry, urban rural urban Bombay rural Haryana Aligarh Guna JJ colonies Resettlement colonies total Gujerat, Dadra, Hagar Haveli, urban Kerala, rural Kerala, urban Karnakaka & 1978–1983 1978–1983 1973–1974 1986 1989–1990 1986 1990 1990 1982 1972 1979 1980 1982 1982 1983 1983 1984 1984 1984 1986 1981–1982 1981–1982 1981–1982 1981–1982 1981–1982 1981–1982 1981–1982 39 60 17 21 17 49 51 37 19–93 5–26 1320 4 12 5 4 0 8 11 23 12 7 11 17 17 21 32 15 3 5–67 0–15 1 1 0 2 2 2 2 54 54 51 40 46 29 14 16–72 0–59 30 10 6 0 36 22 14 10 Stanfield & Galazka 1984 Stanfield & Galazka 1984 Bhat, Joshi & Kandoth 1979 Kumar et al.1988 Khalique, Sinha & Yunus 1993 World Health Organization 1995 World Health Organization 1995 World Health Organization 1995 World Health Organization 1995 World Health Organization 1982a World Health Organization 1982a World Health Organization 1982a Arnold, Soewarso & Karyadi 1986 Arnold, Soewarso & Karyadi 1986 Galazka, Gasse & Henderson1989 World Health Organization 1995 Yusuf, Solter & Bertsch 1991 Yusuf Solter & Bertsch 1991 Arnold & Soewarso 1986 World Health Organization 1995 Galazka 1985 Galazka 1985 Galazka 1985 Galazka 1985 Galazka 1985 Galazka 1985 Galazka 1985 continued Tetanus 175 Kenya Lao PDR Lesotho Lesotho Liberia Pidie District rural urban total Fars Province Sepidan district Eqlid district Firouzabad district Kazeroun district Neireez district total Indonesia Indonesia Indonesia Iran (Islamic Republic of) Iran (Islamic Republic of) Iran (Islamic Republic of) Iran (Islamic Republic of) Iran (Islamic Republic of) Iran (Islamic Republic of) Iran (Islamic Republic of) Iran (Islamic Republic of) Jordan Kenya Kenya Kenya Kenya Kenya Meru Kisii, Meru, Tana Districts, Rural Kilifi Kisii Meru Area 1989 1985 1986 1989 1984–1985 1987 1978–1983 1978–1983 1985 1986 1986 1986 1986 1986 1986 1986 1983 1983 1984 1984–1985 1986 1988 Year(s) to which data pertain 4 80 21 16 29 21 17 21 18 24 6 14 29 16 18 7 10 16 16 14 13 Neonatal 3 4 4 0 15 5 11 7 5 6 13 0 6 6 6 6 2 6 11 11 4 8 Neonatal tetanus Mortality rates per 1000 live births 0 19 20 25 17 51 40 24 34 54 0 43 19 38 34 13 59 71 69 25 70 Percentage of neonatal deaths attributable to tetanus Bjerregard, Steringlass & Mutie 1993 Galazka & Kardymowiscz 1989 Galazka & Kardymowiscz 1989 World Health Organization 1995 Becker, Diop & Thornton 1993 Yusuf, Solter & Bertsch 1991 Stanfield & Galazka 1984 Stanfield & Galazka 1984 Galazka, Gasse & Henderson1989 Hasanabadi 1987 Hasanabadi 1987 Hasanabadi 1987 Hasanabadi 1987 Hasanabadi 1987 Hasanabadi 1987 Hasanabadi 1987 Galazka, Gasse & Henderson1989 World Health Organization 1995 World Health Organization 1995 Melgaard, Mutie & Kimani 1988 World Health Organization 1995 Melgaard, Mutie & Kimani 1988 Reference Estimated neonatal mortality rates and neonatal tetanus mortality rates, and proportion of all neonatal deaths due to tetanus, 1953–1995 (continued) Country Table 6.6 176 Global Epidemiology of Infectious Diseases Pakistan Nepal Niger Nigeria Nigeria Nigeria Nigeria Pakistan Pakistan Pakistan Pakistan Pakistan Pakistan Liberia Madagascar Madagascar Malawi Malawi Malawi Mexico Mexico Mexico Mexico Nepal Nepal Nepal Nepal Nepal Punjab, North West Frontier Province Punjab rural: agricultural urban slum Kwara State Kano Kano Ile Ife Morang District Dhanusa 4 districts (Dang, Kapilaspu, Panchthar, Sindhuli) Dang district Lilongwe District, rural Jalisco, rural Veracruz, rural Veracruz, urban Veracruz Terai rural urban 1987 1991 1989 1988 1989–1990 1990 1991–1995 1978–1983 1981 1981 1981 1984 1986–1987 1986–1988 1989 1989 1978–1983 1990 1991 1988 1988–1989 1988–1989 1988–1989 1980 1982 1988 1990 1983 14 11 4 31 32 34 18 28 4 52 52 57 41 30 30 9 9 12 21 21 4 4 2 3 15 24 4 10 22 6 8 1 12 5 22 26 19 8 4 7 37 44 19 33 88 20 7 29 29 36 68 68 24 60 60 60 45 39 33 28 20 50 50 43 39 55 23 27 7 38 18 41 Galazka & Kardymowicz 1989 World Health Organization 1995 World Health Organization 1995 Babaniyi & Parakoyi 1991 Eregie 1993 Eregie 1993 Davies-Adetugbo, Tomimiro & Nai 1998 Stanfield & Galazka 1984 World Health Organization 1982b Stanfield & Galazka 1984 Stanfield and Galazka 1984 Galazka 1985 Khan & Hardjotanajo 1989 Becker, Diop & Thornton 1993 WHO 1995 WHO 1995 Stanfield and Galazka 1984 Ministry of Health, Malawi 1991 Chiomba 1991 Fidler et al 1994 Ayala et al 1995 Ayala et al 1995 Ayala et al 1995 Galazka, Gasse & Henderson1989 Galazka, Gasse & Henderson1989 World Health Organization 1995 World Health Organization 1995 World Health Organization 1995 continued Tetanus 177 Sudan Sudan Sudan Sudan Sudan Sudan Sudan Sudan Sudan Non-Punjab areas Punjab rural: cattle/horse raising Pakistan Pakistan Pakistan Pakistan Philippines Somalia Somalia Somalia Somalia Sri Lanka Sri Lanka Sri Lanka Sudan Nationwide, north and south Nationwide, rural Nationwide, urban North, rural North, semi-urban North, urban North, urban+rural South, rural South, urban South, urban+rural 2 villages Kalutara Anuradhapura Balad, Mogadishu Area 1981 1981 1981 1981 1981 1981 1981 1981 1981 1990 1990 1981 1988 1978–1983 1978–1983 1981 1981 1987–1989 1982 1984 1984 1981 Year(s) to which data pertain 31 14 21 20 14 19 38 14 35 29 91 48 8 15 58 21 13 91 Neonatal 10 5 5 10 5 5 13 5 12 21 38 <1 1 1 9 42 12 6 21 Neonatal tetanus Mortality rates per 1000 live births 31 35 28 50 38 31 33 32 33 32 73 58 48 23 49 23 80 6 7 Percentage of neonatal deaths attributable to tetanus Olive, Gadir El Sid & Abbas 1982 Olive, Gadir El Sid & Abbas 1982 Olive, Gadir El Sid & Abbas 1982 Olive, Gadir El Sid & Abbas 1982 Olive, Gadir El Sid & Abbas 1982 Olive, Gadir El Sid & Abbas 1982 Olive, Gadir El Sid & Abbas 1982 Olive, Gadir El Sid & Abbas 1982 Olive, Gadir El Sid & Abbas 1982 Dietz et al. 1996 Dietz et al. 1996 Stanfield & Galazka 1984 World Health Organization 1995 Stanfield & Galazka 1984 Stanfield & Galazka 1984 Aden & Birk 1981 Galazka & Kardymowicz 1989 Ibrahim 1996 World Health Organization 1995 Galazka & Kardymowicz 1989 De Silva 1984 Olive, Gadir El Sid & Abbas 1982 Reference Estimated neonatal mortality rates and neonatal tetanus mortality rates, and proportion of all neonatal deaths due to tetanus, 1953–1995 (continued) Country Table 6.6 178 Global Epidemiology of Infectious Diseases Nigeria Nigeria Senegal Singapore Singapore South Africa Uganda Hospital-based data Sudan Syrian Arab Republic Syrian Arab Republic Tanzania Tanzania Thailand Thailand Thailand Thailand Thailand Togo Tunisia Tunisia Tunisia Uganda Uganda VietNam VietNam VietNam VietNam VietNam Yemen Yemen Zaire Zaire Zambia Zimbabwe Calabar Calabar Niakhar rural Pai-Kongila Binh Tri Thien Han Giang Vinh Phu Mbale District 3 regions rural urban rural rural Zanzibar Central Sudan 1983–1986 1984–1987 1983–1986 1960 1970 1953–1955 1985–1989 1989–1990 1981 1990 1988 1989 1953–1957 1978–1983 1993 1980 1991 1984 1988 1988 1988 1984 1990 1985 1989 1989 1989 1989 1984 1978–1983 1983 1984 1986 1983 51 12 14 10 21 11 11 21 8 12 38 16 12 13 3 7 8 19 31 9 11 16 21 20–36 15 16 5 1 6 3 0 1 15 6 2 5 1 3 3 4 3 9 1 4 4 6 5 1 2 1 6 5 31 49 31 45 74 14 40 7 30 39 25 9 38 23 6 23 22 52 14 4 10 40 42 16 39 50 38 40 20 8 29 Antia-Obong & Ikpatt 1991 Asindi, Ibia & Udo 1994 Leroy & Garenne 1991 Chen 1973 Chen 1973 Miller 1972 Bwire & Kawuma 1992 Taha, Gray & Abdelwahab 1993 Khan & Hardjotanajo 1989 World Health Organization 1995 World Health Organization 1995 World Health Organization 1995 Stahlie 1960 Stanfield & Galazka 1984 Chongsuvivatwong, Impat & Tayakkanonta 1993 World Health Organization 1982a World Health Organization 1995 Galazka & Kardymowiscz 1989 World Health Organization 1995 World Health Organization 1995 World Health Organization 1995 Galazka & Kardymowiscz 1989 World Health Organization 1995 Galazka & Kardymowiscz 1989 Thanh et al. 1990 Thanh et al.1990 Thanh et al. 1990 Thanh et al. 1990 Galazka & Gasse 1995 Stanfield & Galazka 1984 World Health Organization 1987a World Health Organization 1995 World Health Organization 1995 Galazka & Kardymowiscz 1989 Tetanus 179 180 Global Epidemiology of Infectious Diseases Estimated morbidity and mortality due to tetanus Estimated morbidity and mortality attributable to neonatal tetanus There are no valid reported data on neonatal tetanus morbidity and mortality because of the variable (and generally low) notification efficiency associated with tetanus cases (Table 6.5). Knowledge of the magnitude of neonatal tetanus mortality increased greatly with the conduct of community surveys in the 1980s. More than 40 developing countries, including 18 of the 25 most populous countries, undertook special community-based surveys to measure the neonatal tetanus mortality (Basu & Sokhey 1984, Galazka & Stroh 1986) (Table 6.6). Neonatal tetanus mortality rates ranged from 1–2 per 1000 live births in Jordan, Sri Lanka and Tunisia to 67 per 1000 live births in some areas in India. Based on these studies, it was estimated in 1987 that, globally, approximately 800 000 newborns die every year from tetanus (Galazka & Kardymowicz 1989). By exposing the magnitude of neonatal tetanus mortality, the community-based surveys were critically important in providing baseline information, in changing the perception of health policy makers and in generating political support for the control of neonatal tetanus. Surveys of mortality were repeated in some countries in the late 1980s or early 1990s, and the results generally suggest decreasing neonatal tetanus mortality rates and a lower proportion of neonatal tetanus deaths among total neonatal mortality (Table 6.6). Using the results of these mortality surveys, a methodology was developed by the WHO to estimate the number of neonatal tetanus deaths and cases still occurring in the presence of tetanus toxoid immunization and clean delivery practices. Each country is assigned a pre-immunzation era mortality rate for neonatal tetanus, using data from the first communitybased neonatal tetanus mortality survey in the country. In the absence of a survey, countries are assigned rates based on the surveys done in neighbouring countries at similar risk. This rate is multiplied by the country’s annual number of live births to calculate a baseline number of neonatal tetanus deaths that would occur without tetanus toxoid immunzation activities. Using a country’s most recent and reliable coverage estimates for two or more doses of tetanus toxoid in pregnant women (TT2+) and assuming a vaccine efficacy of 95 per cent, the annual number of neonatal tetanus deaths which would occur is estimated. Specifically, TT2+ coverage data and assumed vaccine efficacy are used to estimate the number of births that would still be susceptible to neonatal tetanus NT in the presence of tetanus toxoid immunization activities. The number of susceptible births is then multiplied by the baseline neonatal tetanus mortality rate assigned to that country to obtain an estimate of the number of neonatal tetanus deaths. This result is subtracted from an estimate of neonatal tetanus deaths using the same method but assuming no tetanus toxoid immunization to determine the annual number of neonatal tetanus deaths prevented by immunization services. The same method is not used for estimates in countries achieving at Tetanus 181 least 70 per cent coverage with clean delivery practices. It is assumed that a health system able to achieve such a high level of clean delivery coverage would also have a surveillance system with data reliable enough to be adjusted and used to derive an estimate of neonatal tetanus cases. For most of the countries in this category, the notification efficiency of the surveillance system is assumed to be about 33 per cent and the number of neonatal tetanus cases reported is therefore multiplied by three. Assumed case fatality ratios are applied to derive neonatal tetanus death estimates from the neonatal tetanus case estimates, or vice versa (depending on which method was used). A CFR of 90 per cent is used for least developed countries, 80 per cent for developing countries, 60 per cent for countries with economies in transition, and 40 per cent for countries with developed market economies. This method was used to produce an estimate of 446 000 neonatal tetanus deaths in 1990, excluding China. A rough estimate of 18 000 tetanus deaths was made for China assuming a neonatal tetanus incidence rate of 1 per 1000 live births and a case fatality ratio of 80 per cent. Including China, an estimated total of 464 000 deaths occurred in 1990 that were attributable to neonatal tetanus. Estimated morbidity and mortality attributable to non-neonatal tetanus There are no reliable data showing the real magnitude of non-neonatal tetanus globally. It was estimated that in the 1980s, 120 000 to 300 000 deaths attributable to non-neonatal tetanus occurred each year in the world, excluding China (Rey & Tikhomirov 1989). Estimates of nonneonatal tetanus cases occurring in developing countries are derived from estimates of neonatal tetanus cases and a general assumption that in 1990 the latter accounted for approximately 80–85 per cent of all cases under five years of age, although in reality this proportion varies according to the level of childhood immunization against tetanus. With this assumption, an estimate was derived of tetanus cases among children under five years of age. Tetanus cases were then assigned to older age groups by extrapolation, according to an estimated age distribution based on reported age-specific distributions and age-specific population immunity (Rey & Tikhomirov 1989, Galazka & Gasse 1995). For countries with developed market economies and economies in transition (i.e. former Soviet Union countries), reported tetanus cases and a 75 per cent notification efficiency were used to derive the estimated numbers of tetanus cases. Once estimated cases were calculated, case fatality ratios were applied by age group and income level to estimate deaths (Table 6.7). Based on these methods, we estimated that 155 000 cases and 78 000 deaths occurred in 1990 as result of non-neonatal tetanus. The total estimated burden for tetanus (both neonatal and non-neonatal) in 1990 was 542 000 deaths. 163 402 Total CFR = case fatality rate. 90 631 24 213 17 796 14 435 8 025 3 760 2 325 2 217 Est. no. of cases 51 50 50 50 50 55 60 65 65 Est. CFR (%) 83 159 45 315 12 107 8 898 7 218 4 414 2 256 1 511 1 441 Est. no. of deaths Developing and least developed countries 1 167 12 0 35 58 58 350 350 303 Est. no. of cases 44 25 25 30 40 50 50 25 Est. CFR (%) 510 3 0 9 15 18 140 175 152 Est. no. of deaths Countries with established market economies and economies in transition 164 568 90 642 24 213 17 831 14 494 8 083 4 110 2 675 2 521 Total estimated number of cases 83 670 45 318 12 107 8 907 7 232 4 431 2 396 1 686 1 593 Total estimated number of deaths Estimated number of non-neonatal tetanus cases and deaths by age in developing and developed countries, 1990 <5 5–14 15–29 30–44 45–59 60–69 70–79 80+ Age in years Table 6.7 182 Global Epidemiology of Infectious Diseases Tetanus 183 Updated Estimates Based on more complete data available since the early 1990s, the estimates for burden of tetanus in 1990 have since been updated. The estimate for China was updated based on a review of community surveys conducted in 1993 of tetanus incidence in 300 high risk counties (World Health Organization 1993b). The review team concluded that, based on a neonatal tetanus mortality rate of 4.2 per 1000 live births, an estimated 90 000 neonatal tetanus deaths were occurring annually in China. We considered this estimate to be high, since it was based on surveys conducted in high risk counties only. Using the methods described in this chapter for estimating the burden of neonatal tetanus and assuming a pre-immunization neonatal tetanus mortality rate of 3 per 1000 live births, we revised the estimate for neonatal tetanus for China to 71 000 neonatal tetanus deaths. With these and other minor updates to some pre-immunization neonatal tetanus mortality rates assigned to countries, we estimated that 427 000 neonatal tetanus deaths and 514 000 cases occurred in 1990. Neonatal tetanus was the second leading cause of deaths among childhood diseases prevented by routine immunization services. An estimated 533 000 infant deaths were prevented through tetanus toxoid immunization and clean delivery practices in 1990. The updated estimate would suggest a 6 per cent notification efficiency when compared to the neonatal tetanus cases reported to WHO for 1990. The Regions with the highest burden in terms of absolute numbers are Asia and sub-Saharan Africa with 67 per cent and 28 per cent, respectively, of the total estimated deaths (Figure 6.5). Several small countries of Africa have some of the highest estimated neonatal tetanus mortality rates (Figure 6.6). The updates made to the neonatal tetanus model also caused changes Figure 6.5 Estimated number of tetanus deaths occurring in the world by WHO region, 1990 Non-neonatal tetanus WPR 19 000 Neonatal tetanus WPR 82 000 AFR 180 000 AFR 113 000 AMR 3 000 AMR 9 000 EMR 15 000 SEAR 28 000 EUR 300 SEAR 146 000 EUR 200 EMR 78 000 Estimated neonatal tetanus mortality rate per 1000 live births, 1990 <1 1– 4 >4 Neonatal tetanus mortality rate per 1000 live births Figure 6.6 184 Global Epidemiology of Infectious Diseases Tetanus 185 in the non-neonatal tetanus estimates (since the model for the latter is based on the former). The updated estimates for non-neonatal tetanus in 1990 are 165 000 cases and 84 000 deaths. For China, we assumed that only about one-third of all tetanus cases were non-neonatal because of the high infant immunization coverage levels already being achieved in 1990 relative to the low level of tetanus protection at birth. In all, 70 per cent of all non-neonatal deaths were estimated to occur in India and other Asian countries, and 24 per cent in sub-Saharan Africa. In summary, based on updated estimates, an estimated 679 000 cases and 511 000 deaths occurred worldwide as a result of tetanus (both neonatal and non-neonatal) in 1990. Estimated morbidity and mortality by age The age distribution of tetanus cases and age-specific case-fatality ratios are different for the developing and industrialized world. In industrialized countries, a neonatal tetanus case is a rare event and non-neonatal tetanus is limited to older persons. In developing countries, neonatal tetanus continues to be an important cause of preventable morbidity and mortality. Non-neonatal tetanus also takes a terrible toll, especially in younger segments of the population. It is estimated that in 1990 about 70 per cent of all non-neonatal tetanus cases and deaths occurred among persons under 15 years of age. The large majority of these cases and deaths occurred in developing countries (Table 6.7). Conclusions Tetanus is a preventable disease. Maternal and neonatal tetanus could be eliminated through the immunization of women to produce and maintain immunity against tetanus during their childbearing years and through clean delivery and umbilical stump care practices. A review of empirical studies and simulations concluded that tetanus immunization is highly cost-effective and an effective means of reducing infant mortality in the developing world (Steinglass, Brenzel & Percy 1993). Providing tetanus toxoid during pregnancy is the most cost-effective tetanus immunization strategy in developing countries, particularly when it is targeted at high-risk populations (Cvjetanovic et al. 1972). To protect the rest of the population from wound tetanus, the strategy of first choice includes immunization in infancy and early childhood, reinforced by booster doses as part of a school health programme. In 1989, recognizing the magnitude of neonatal tetanus incidence and the fact that the disease is preventable through simple interventions, the World Health Assembly resolved to eliminate neonatal tetanus from the globe by 1995. This goal was reaffirmed at the World Summit for Children in 1991. The definition of neonatal tetanus elimination is the achievement of less than one case of neonatal tetanus per 1000 live births in every district of every country. 186 Global Epidemiology of Infectious Diseases Immunisation with tetanus toxoid For previously non-immunized women, the WHO recommends an immunization schedule consisting of five doses of adsorbed tetanus toxoid (TT) be given to women of childbearing age, and especially to pregnant women. In developing countries, TT coverage among pregnant women has progressed slowly (Figure 6.3). Between 1990 and 1995, coverage with at least two doses of TT rose from 43 per cent to 49 per cent in developing countries and of the former Soviet Union, a pace too gradual to eliminate neonatal tetanus by 1995. By 1995, China, the world’s most populous country, had only begun widespread TT immunization by targeting all women of child-bearing age in 320 counties considered at high risk. High TT coverage levels are attainable. By 1995, a total of 23 (17 per cent) of the 135 developing countries which are WHO Member States reported TT2+ coverage of at least 80 per cent (Figure 6.7). Coverage with TT immunization in developing countries may rise substantially in the next few years as China introduces routine TT immunization for pregnant women. All countries are providing TT-containing-vaccines (DiphtheriaPertussis-Tetanus and Diptheria-Tetanus) in their immunization services for children. Globally, according to WHO/UNICEF estimates, 72 per cent of infants received a primary series of three doses of DPT vaccine in 1995. There are, however, considerable differences in immunization coverage between regions, countries and areas within individual countries (Table 6.8). Clean delivery practices Over the past decade, many developing countries have concentrated on training traditional birth attendants, providing home delivery kits and Table 6.8 Percentage of children immunized in the first year of life with 3 doses of DTP vaccine and percentage of pregnant women immunized with at least 2 doses of tetanus toxoid, by WHO Region, 1993 WHO region African Region Eastern Mediterranean Region European Region Region of the Americas South East Asia Region Western Pacific Region Global No. of countries providing data % of global population represented by reporting countriesa 3 doses of DTP vaccine At least 2 doses of tetanus toxoid 51 65 84 81 64 91 72 1730 35 50 77 43 77 21 49 107 98 88 a. For tetanus toxoid, this figure represents the percentage of the population from developing countries and countries of the Former Soviet Union only. b. Based on WHO/UNICEF estimates based on official country reports. Reported coverage with at least two doses of tetanus toxoid among pregnant women, 1995 Source: World Health Organization 1996 No data <50% 50%–79% ≥80% TT2+ coverage Figure 6.7 Tetanus 187 188 Global Epidemiology of Infectious Diseases educating pregnant women about the “three cleans” (clean hands, clean delivery surface and clean cord care). Operational research has shown that training traditional birth attendants alone can reduce both the total neonatal death rate and neonatal tetanus death rate (Mangay-Angara 1981, Rahman 1982). It was found that total neonatal mortality was reduced more by training traditional birth attendants than by immunizing against tetanus, although immunzation alone was more effective in reducing the incidence of neonatal tetanus (Rahman 1982). The results of a small number of nationwide surveys are available to estimate global and regional coverage rates with clean delivery, defined as those with a trained attendant present (World Health Organization 1996). The latest regional and subregional estimates show that countries in Latin America have the highest rates of maternity care coverage in the developing world, 75 per cent of women having a skilled person attending at delivery, followed by 53 per cent in Asia, 52 per cent in Oceania, and 42 per cent in Africa (World Health Organization 1996). Priority activities to eliminate neonatal tetanus The first priority for all countries where neonatal tetanus is endemic is to identify foci of high neonatal tetanus risk so that women of child-bearing age in these foci can be immunized with tetanus toxoid. Implementing intensified immunization activities in such high risk areas will rapidly reduce the incidence of neonatal tetanus. Meanwhile, longer term strategies should be put in place to achieve and sustain high levels of routine tetanus immunization as well as clean delivery services and umbilical stump care, particularly among those most at risk. References Abuwa PNC et al. (1997) Epidemiology of neonatal tetanus in Rivers state of Nigeria. Journal of Epidemiology and Community Health, 51:336. Adadevoh BK, Akinla O (1970) Postabortal and postpartum tetanus. Journal of Obstetrics and Gynaecology of the British Commonwealth, 77:1019–1023. Adams JQ, Morton RF (1955) Puerperal tetanus. American Journal of Obstetrics and Gynecology, 69(1):169–173. Aden A, Birk S (1981) A study of child mortality in Mogadishu, Somalia. Journal of Tropical Pediatrics, 27:279–284. Adeuja AOG, Osuntokun BO (1971) Tetanus in the adult Nigerian. East African Medical Journal, 48(12):683–691. Afonja AO (1973) Tetanus in Lagos: a review of 228 adult Nigerian patients. Journal of Tropical Medicine and Hygiene, 76:171–174. Akbar P (1981) Neonatal tetanus. Pakistan Paediatric Journal, 5(2):57–64. Al Mukhtar MY (1987) Birth care practice and tetanus neonatorum: a hospitalbased study in Mosul. Public Health, 101:453–456. Tetanus 189 Alemu W (1993) Neonatal tetanus mortality survey, North and South Omo Administrative Regions, Ethiopia. Ethiopian Medical Journal, 31:99–107. Ancelle JP et al. (1974) Le tetanos à Yaoundé [Tetanus in Yaoundé]. Medecine d’Afrique Noire, 21(12):953–957. Anlar B, Yalaz K, Dizmen R (1989). Long-term prognosis after neonatal tetanus. Developmenatal Medicine and Child Neurology, 31:76–80. [Anonymous] (1993) Tetanus: the elderly are at greatest risk. Communicable Disease Report, 17(3):73. Antia-Obong OE, Ikpatt NW (1991) Neonatal tetanus in Calabar: a 4-year retrospective study. Nigerian Journal of Pediatrics, 18(3):94–99. Arnold RB, Soewarso TI, Karyadi A (1986) Mortality from neonatal tetanus in Indonesia: results of two surveys. Bulletin of the World Health Organization, 64(2):259–262. Asindi AA, Ibia EO, Udo JJ (1994) Mortality pattern among Nigerian children in the 1980s. Journal of Tropical Medicine and Hygiene, 94:152–155. Athavale VB, Pai PN (1965) Tetanus neonatorum: clinical manifestations. Journal of Pediatrics, 67:649–657. Athavale VB et al. (1975) Tetanus neonatorum. Progress in Drug Research, 19: 189–208. Axnick NW, Alexander ER (1957) Tetanus in the United States: a review of the problem. American Journal of Public Health, 47:1493–1501. Ayala CVM et al. (1995) Neonatal tetanus Mortality in Veracruz Mexico, 1989. Bulletin of Pan American Sanitary Bureau, 29(29):116–128. Ayim EN (1972) The management of tetanus: a study of 51 cases treated at Muhimbili hospital, Dar es Salaam. East African Medical Journal, 49(2):171–176. Ayivi B et al. (1992) Le tetanos chez l’enfant aspects epidemiologique clinique et thérapeutique: à propos de 784 cas [Tetanus in children: epidemiologic, clinical and therapeutic aspects: in connection with 784 cases]. Publications Médicales Africaines, 121:54–64. Babaniyi O, Parakoyi B (1991) Cluster Survey for polio and neonatal tetanus in Ilorin, Nigeria. International Journal of Epidemiology, 20(2):515–520. Bandele EO, Akinyanju OO, Bojowoye BJ (1991) An analysis of tetanus deaths in Lagos. Journal of the National Medical Association, 83:55–58. Baqui AH et al. (1998) Causes of childhood deaths in Bangladesh: results of a nationwide verbal autopsy study. Bulletin of the World Health Organization, 76(2):161–171. Basu RN, Sokhey J (1984) Special combined survey for poliomyelitis and neonatal tetanus as supplement to routine surveillance system in India. Journal of Communicable Diseases, 16:148–153. Barten J (1973) Neonatal tetanus in Indonesia. Tropical Doctor, 3:107–109. Becker SR, Diop F, Thornton JN(1993) Infant and child mortality in two countries of Liberia. International Journal of Epidemiology, 22(l):S56–S63. 190 Global Epidemiology of Infectious Diseases Bennett MJ (1976) Postabortal and postpartum tetanus: a review of 19 cases. South African Medical Journal, 50:513–516. Berggren WL, Ewbank DC, Berggren GG (1981) Reduction of mortality in rural Haiti through a primary health care program. New England Journal of Medicine, 304:1324–1330. Beytout D et al. (1988) Controle de l’état d’immunité antitétanique dans la population du Puy de Dòme [Control of anti-tetanus immunity in the population of Puy de Dòme]. Medecine Maladies Infectieuses, 12:897–899. Bhat GJ, Joshi MK, Kandoth PW (1979) Neonatal tetanus: clinical study of 100 cases. Indian Pediatrics, 16(2):159–166. Bildhaiya GS (1983) A tetanus study in a hospital at Ahmedabad. Journal of the Indian Medical Association, 80(2):21–23. Bjerregard P, Steinglass R, Mutie DM (1993) Neonatal tetanus mortality in coastal Kenya. International Journal of Epidemiology, 22(1):163–169. Blankson JM (1977) Problems of neonatal tetanus as seen in Ghana. African Journal of Medical Sciences, 6:1–6. Bourleaud J, Huet M (1985) Contribution à l’étude du tetanus en France [Contribution to the study of tetanus in France]. Medecine Maladies Infectieuses, 15(1):7–16. Box QT (1964) The treatment of tetanus. Pediatrics, 33:872. Bwire R, Kawuma JJ (1992) Hospital study on neonatal tetanus, Buluba Hospital 1985–1989. Journal of Tropical Medicine and Hygiene, 95:62–66. Bytchenko B (1966) Geographical distribution of tetanus in the world, 1951–60: a review of the problem. Bulletin of the World Health Organization, 34:71–104. Centers for Disease Control and Prevention(1985) Tetanus – United States, 1982–1984. Morbidity and Mortality Weekly Report, 34(39):602–611. Centers for Disease Control and Prevention (1987) Tetanus – United States, 1985–1986. Morbidity and Mortality Weekly Report, 36(29):477–481. Centers for Disease Control and Prevention (1990) Tetanus – United States, 1987–1988. Morbidity and Mortality Weekly Report, 39(3):37–41. Chaubey BS (1965) Vertebral compression in tetanus. In: Proceedings of the First International Conference on Tetanus. 8–10 November 1963, JC Patel, ed. pp. 226–234, Study Group of Tetanus, King Edward Memorial Hospital, Bombay. Chen AJ (1973) Neonatal tetanus in Singapore 1960–1970. Singapore Medical Journal, 14(2):111–116. Chen LC, Rahman M, Sarder AM (1980) Epidemiology and causes of death among children in a rural area of Bangladesh. International Journal of Epidemiology, 9(1):25–33. Chen ST (1974) Tetanus neonatorum in West Malaysia. Journal of Tropical Medicine and Hygiene, 77(9):204–207. Chiomba WM (1991) Risk factors for neonatal mortality in Malawi: a community-based case control study in rural Lilongwe district. International Journal of Health Sciences, 2(1):37–43. Tetanus 191 Chongsuvivatwong V, Impat A, Tayakkanonta K (1993), A survey of neonatal tetanus and perinatal mortality in southern Muslim communities of Thailand. Southeast Asian Journal of Tropical Medicine and Public Health, 24:654–658. Christenson B, Bottiger M (1987) Epidemiology and immunity to tetanus in Sweden. Scandinavian Journal of Infectious Diseases, 19:429–435. Christensen PE (1970) Comments on epidemiology and immunology of tetanus in Denmark. Proceedings of the Third International Conference on Tetanus, Sao Paulo, Brazil, 17–22 August 1970, Scientific Publication, No. 253, pp. 27–30. Pan American Health Organization, Washington DC. Cliff J (1985) Neonatal tetanus in Maputo, Mozambique Part II: preventive measures. Central African Journal of Medicine, 31(2):27–32. Collacott RA (1981) Neonatal tetanus in St Kilda. Scottish Medical Journal, 26: 224–227. Cosnard G, Joullie M, Davin R (1977) Le Tetanos à l’Hôpital de Quahigouya (Haute-Volta): a propos de 183 cas [Tetanus at the Quahigouya hospital (Upper Volta): study of 183 cases]. Medicine Tropicale, 37(1):45–49. Crossley K et al. (1979) Tetanus and diphtheria immunity in urban Minnesota adults. Journal of the American Medical Association, 242:2298–2300. Cvjetanovic B et al. (1972) Epidemiologic model of tetanus and its use in the planning of immunization programmes. International Journal of Epidemiology, 1(2):125–137. Daud S, Mohammad T, Ahmad A (1981) Tetanus neonatorum: a preliminary report of assessment of different therapeutic regimens. Journal of Tropical Pediatrics, 27:308–311. Davies-Adetugbo AA, Torimiro SEA, Ako Nai NA (1998) Prognostic factors in neonatal tetanus. Tropical Medicine and International Health, 3(1)9–13. Deivanayagam N, Nedunchelian K, Kamula KG (1991) Neonatal tetanus: observations on antenatal immunisation, natal and immediate postnatal factors. Indian Journal of Pediatrics, 58:119–122. de la Fuente M, González S, Guevara G (1986) Epidemiología del tétanos en Chile en los últimos 25 años [Epidemiology of tetanus in Chile in the last 25 years]. Revista Chilena de Pediatría, 57(3):273–277. De Silva (1984) Neonatal tetanus in Sri Lanka. Geneva, World Health Organization. Dietz V, Milstien JB, van Loon F et al. (1996) Performance and potency of tetanus toxoid: implications for eliminating neonatal tetanus. Bulletin of the World Health Organization, 74(6):619–628. Diop-Mar I et al. (1985) Clinical aspects of tetanus in Africa with 1019 cases in Dakar (1982–83). In: Proceedings of the Seventh International Conference on Tetanus, Copanello, 10–15 September 1984, P Mastroeni, G. Nistico, M Pitzutta, eds. pp. 499–507. Rome, Gangemi Publishing Company. Ebisawa I (1971) Mortality and sex difference of tetanus in Japan. Japanese Journal of Experimental Medicine, 41(1):21–29. 192 Global Epidemiology of Infectious Diseases Ebisawa I, Homma R (1986) Tetanus in Japan: trends of mortality, case fatality, and causes of death. Japanese Journal of Experimental Medicine, 56(4):155–161. Edmondson RS and Flowers MW (1979) Intensive care in tetanus: management, complications, and mortality in 100 cases. British Medical Journal, 1:1401–1404. Einterz EM and Bates ME (1991) Caring for neonatal tetanus patients in a rural primary health setting in Nigeria: a review of 237 cases. Journal of Tropical Pediatrics, 37:179–181. El-Sherbini A (1991) Study of Tetanus Neonatorum in Tanta fever hospital 1988–89. Journal of Tropical Pediatrics, 37:262–263. Eregie CO (1993) Epidemiological factors associated with neonatal tetanus mortality: observations from a cluster survey in Nigeria. East African Medical Journal, 70(7):434–436. Fauveau V et al. (1990) The effect of maternal and child health and family planning services on mortality: is prevention enough? British Medical Journal, 301: 103–107. Fauveau V et al. (1993) Maternal tetanus: magnitude, epidemiology and potential control measures. International Journal of Gynecology and Obstetrics, 40:3–12. Fidler A, Hartog R, Lezana MA (1994) Field test of a rapid assessment technique against a probabilistic community survey: operational implications for neonatal tetanus elimination. International Journal of Epidemiology, 23(2):386–392. Figueroa JP, Clarke R (1988) Under-reporting of tetanus and diphtheria cases in Jamaica (1980–1986). West Indian Medical Journal, 37:143–147. Galazka A, Abgarowicz A (1973) Tetanus. In: Kostrzewski, Lekarskich, Panstowowy Zaklad and Wydawnictw, eds. Infectious diseases in Poland and their control in 1961–1970, Warszawa pp. 252–265. Galazka AM (1985). Control of tetanus. Indian Journal of Pediatrics, 52(417): 329–341. Galazka A, Gasse F, Henderson RH (1989) Neonatal tetanus in the world and the global Expanded Programme on Immunization. In: Proceedings of the Eighth International Conference on Tetanus, Leningrad 25–28 August 1987. Nistico G, et al. eds. pp. 470–487. Rome-Milan, Pythagora Press. Galazka A, Kardymowicz B (1989) Tetanus incidence and immunity in Poland. European Journal of Epidemiology, 5(4):474–480. Galazka A, Gasse F (1995) The present status of tetanus and tetanus vaccination. In: Clostridial Neurotoxins, C. Montecuccio, eds. Current Topics in Microbiology and Immunology, 115:31–52. Galazka A, Stroh G (1986) Neonatal tetanus: Guidelines on the community-based survey on neonatal tetanus mortality. Geneva, World Health Organization, WHO/EPI/GEN/86/8. Galbraith NS, Forbes P, Tillett H (1981) National surveillance of tetanus in England and Wales 1930–79. Journal of Infection, 3:181–191. Garg SK et al. (1993) Neonatal mortality in Meerut District. Indian Journal of Medical Science, 47(9):222–229. Tetanus 193 Gek LH (1951) A review of tetanus neonatorum in Singapore during the years 1946–50. The Medical Journal of Malaya, 5(3):3:181–194. Ghosh TK (1984) Surveillance of tetanus in Delhi hospitals: a report of the national course on planning and management of EPI. National Institute of Communicable Diseases, pp. 32–35. Gordon JE, Singh S, Wyon JB (1961) Tetanus in villages of the Punjab: an epidemiologic study. Journal of the Indian Medical Association, 37(4):157–161. Gupta SD, Keyl PM (1998) Effectiveness of prenatal tetanus toxoid immunization against neonatal tetanus in a rural area of India. Pediatric Infectious Disease Journal, 17(4):316–321. Gurses N, Aydin M (1993) Factors affecting prognosis of neonatal tetanus. Scandinavian Journal of Infectious Diseases, 25:353–355. Handalage DCMA, Wickramasinghe SYDC (1976) Neonatal tetanus: some observations from Sri Lanka. Environmental Child Health, 8:167–171. Hasanabadi AS (1987) Neonatal tetanus in Fars Province, Islamic Republic Of Iran. Eastern Mediterranean Region Health Services Journal, 3:14–21. Hlady WG et al. (1992) Neonatal tetanus in rural Bangladesh: risk factors and toxoid efficacy. American Journal of Public Health, 82(10):1365–1369. Ibrahim MM (1996) Child mortality in a collapsing African society. Bulletin of the World Health Organization, 74(5):547–552. Illis LS, Taylor FM (1971) Neurological and electroencephalographic sequelae of tetanus. Lancet, 1:826–830. Islam MS (1982) Birth care practice and neonatal tetanus in a rural area of Bangladesh. Journal of Tropical Pediatrics, 28:299–302. Jaffari SMH, Pandit MM, Ismail M (1966) Neonatal tetanus in Hyderabad. Indian Pediatrics, 3(5):177–182. Januszkiewicz J (1975) Compression fractures of vertebrae in tetanus: prospective study of 50 cases. In: Proceedings of the Fourth International Conference on Tetanus, Dakar, Senegal, 6–12 April 1975, Fondation Merieux, Lyon pp. 285–290. Jelliffe DB (1950) Tetanus neonatorum. Archives of Disease in Childhood, 25: 190–192. Jolliet P et al. (1990) Aggressive intensive care treatment of very elderly patients with tetanus is justified. Chest, 197(3):702–705. Kalter HD, et al. (1999) Validation of caregiver interviews to diagnose common causes of severe neonatal illness. Pediatric and Perinatal Epidemiology, 13: 99–113. Khalique N, Sinha SN, Yunus M (1993) Early childhood mortality: a rural study. Journal of the Royal Society of Health, 113(5):247–249. Khan MA, Hardjotanajo W (1989) Neonatal tetanus profile in Pakistan. In: Proceedings of the Eighth International Conference on Tetanus. Leningrad 25–28 August 1987. G. Nistico, et al., eds. Rome-Milan Pathogora Press, pp. 488–505. Khoo BH, Lee EL, Lam KL (1978) Neonatal tetanus treated with high dosage of diazepam. Archives of Disease in Children, 53:737–739. 194 Global Epidemiology of Infectious Diseases Kjeldsen K, Simonsen O, Heron I (1988) Immunity against diphtheria and tetanus in the age group 30–70 years. Scandinavian Journal of Infectious Diseases, 20:177–85. Klenerman P, Scragg J (1955) Preliminary report on a modified treatment of tetanus neonatorum. South African Medical Journal, 29(37):853–854. Kumar V, et al. (1988) Neonatal tetanus mortality in a rural community of Haryana. Indian Pediatrics, 25:167–169. Kuszewski K, Dutkiewicz G (1994) Tezec w 1992 roku [Tetanus in 1992]. Przeglad Epidemiologiczny, 48:129–131. La Force PM, Young LS, Bennett JV (1969) Tetanus in the United States (1965– 1966): epidemiologic and clinical features. New England Journal of Medicine, 280(11):569–574. Lau RCH (1987) Detection of tetanus toxoid antibodies in human sera in New Zealand by ELISA. Epidemiology and Infection, 98:199–202. Lamisse F, Sonneville A, Choutet P (1981) Le tetanos: à propos de cinq cent trois observations [Tetanus: 503 observations]. La Semaine des hôpitaux de Paris, 57: 1076–1080. Leroy O, Garenne M (1991) Risk factors of neonatal tetanus in Senegal. International Journal of Epidemiology, 20(2):521–526. Lombard L, Lepoutre A (1993) Le Tetanos en France en 1991 et 1992. Bulletin Épidémiologique Hebdomaire, 28:125–126. Luisto M (1989) Epidemiology of tetanus in Finland from 1969–1985. Scandinavian Journal of Infectious Diseases, 21:655–663. Majumdar H, Kaur S (1971) Neonatal tetanus. Indian Pediatrics, 8(4):145–151. Malawi Ministry of Health, UNICEF, WHO (1991) Malawi national neonatal tetanus mortality survey: 1990 Malawi. African Index Medicus. Manandhar DS (1978) Management of tetanus neonatorum: a report of 100 cases treated in Bheri Zonal Hospital. Journal of the Nepal Medical Association, 16(2):13–21. Mangaya-Angara A (1981) Philippines: the development and use of the registry of traditional birth attendants. In: The traditional birth attendants in seven countries. A Mangaya-Maglacas, H Pizurki, eds. pp. 37–70. World Health Organization (Public Health Papers, No. 75). Marshall FN (1968) Tetanus of the newborn, with special reference to experiences in Haiti. In: SZ Levine, ed., Year Book Medical Publishers, Chicago, Adv Pediatr, volume 15. pp. 65–110. Maru M, Getahun A, Hosana S (1988) A house-to-house survey of neonatal tetanus in urban and rural areas in the Gondar region, Ethiopia. Tropical & Geographical Medicine, 40(3):233–6. Mazumder GDN, Chakraborthy AK (1974) Problem of tetanus in Calcutta. Indian Journal of Public Health, 18:68–78. Melgaard B, Mutie DM, Kimani G (1988) Cluster survey of mortality due to neonatal tetanus in Kenya. International Journal of Epidemiology, 17(1):174–177. Tetanus 195 Miller JK (1972) The prevention of neonatal tetanus by maternal immunization. Environmental Child Health, 6:160–167. Misra SR, Rao RS (1988) Serological survey of tetanus antibodies in the Pondicherry region. Indian Journal of Medical Research, 87:161–165. Morgan P et al. (1981) A serological evaluation of an immunization programme in Salt Cay, Turks, and Caicos Islands, West Indian Medical Journal, 30:68–71. Mya MM, To M, Aye KK et al. (1985) Tetanus immunity: Trends and/or status in some selected adult Burmese populations. Burma Medical Journal, 31:149–164. Newell KW et al. (1966) The use of toxoid for the prevention of tetanus neonatorum: final report of a double-blind controlled field trial. Bulletin of the World Health Organization, 35:863–871. Ogbeide MI (1966) Problems of neonatal tetanus in Lagos. Journal of Tropical Pediatrics, 12:71–79. Okoro BA, Okeahialam TC (1987) Prognostic determinants in neonatal tetanus. East African Medical Journal, 64(1):61–64. Olive JM, Gadir El Sid A, Abbas M (1982) Neonatal tetanus retrospective study, Sudan, Oct–Nov 1981. Pakistan Pediatric Journal, 6:260–275. Owa JA, Makinde OO (1990) Neonatal tetanus in babies of women immunized with tetanus toxoid during pregnancy. Tropical Doctor, 10:156–157. Parija AC, Mohanta KD (1973) Neonatal tetanus. Indian Journal of Pediatrics, 40:353–360. Patel JC, Mehta MM, Goodluck PL (1965) Vertebral fractures in tetanus. In: Proceedings of the First International Conference on Tetanus, Bombay 8–10 November 1963, JC Patel, ed. Study Group of Tetanus, King Edward Memorial Hospital, Bombay, pp. 239–249. Patel JC, Mehta BC (1999) Tetanus: study of 8697 cases. Indian Journal of Medical Sciences, 53:393–401. Pelletier A, Roure C (1991) Le tetanos en France en 1990 (à partir des déclarations obligatoires) [Tetanus in France in 1990 (obligatory declarations)]. Bulletin Épidémiologique Hebdomaire, 31:127–128. Perry H et al. (1998) Tetanus toxoid immunization coverage among women in Zone 3 of Dhaka city: the challenge of reaching all women of reproductive age in urban Bangladesh. Bulletin of the World Health Organization, 76(5): 449–457. Pinheiro D (1964) Tetanus of the newborn infant: a review of 238 cases. Pediatrics, 7:32–37. Phatak AT, Shah SB (1973) Indices of severity of neonatal tetanus. Indian Pediatrics, 10(2):87–91. Prevots R et al. (1992) Tetanus surveillance: United States, 1989–1990. Morbidity and Mortality Weekly Report, 41(Suppl. 8):1–9. Rahman S (1982) The effect of traditional birth attendants and tetanus toxoid in reduction of neo-natal mortality. Journal of Tropical Pediatrics, 28:163–165. 196 Global Epidemiology of Infectious Diseases Rey M, Tikhomirov E (1989) Non neonatal tetanus over the world. In: Proceedings of the Eighth International Conference on Tetanus. Leningrad, 25–28 August 1987. G Nistico, et al. eds. Pythagora Press, Rome-Milan, pp. 509–518. Riyad SM, El Geneidy MM (1983) Tetanus neonatorum in Alexandria. Journal of the Egyptian Public Health Association, 58(5,6):389–404. Rosmini F et al. (1987) Year of birth, sex and residence as “determinants” of tetanus incidence and immunity in Italy. European Journal of Epidemiology, 3(4):377–380. Salimpour R (1978) Tetanus of the newborn in Tehran: a ten-year study of 880 cases. Tropical Pediatrics and Environmental Child Health, 6:140–142. Saxena O, Saxena S (1965) Tetanus neonatorum: a review of cases admitted to the P.B.M. Group of Hospitals, Bikaner, during a 201⁄2-year-period (1945–1965). Indian Journal of Pediatrics, 32:395–401. Schwoebel V et al. (1992) Community-based evaluation survey of immunizations in Burkino Faso. Bulletin of the World Health Organization, 70(5):583–590. Sehgal H, Gupta SP, Sidhu KK (1980) Tetanus neonatorum as seen in Kalawati Saran Children’s Hospital, New Delhi in the year 1965 and a decade later. Indian Journal of Public Health, 14(2):88–91. Shrestha SM (1978) Incidence of tetanus in Nepalganj Hospital and its prevention. Journal of the Nepal Medical Association, 16(2):41–47. Simental PS et al. (1993) Neonatal tetanus experience at the National Institute of Pediatrics at the Mexico City. Pediatric Infectious Disease Journal, 12:722–725. Simonsen O, Block AV, Heron I (1987) Epidemiology of tetanus in Denmark 1920–1982. Scandinavian Journal of Infectious Diseases, 19:437–444. Simonsen O et al. (1987) Immunity against tetanus and response to revaccination in surgical patients more than 50 years of age. Surgery, Gynecology and Obstetrics, 164:329–34. Singh J, Datta KK, Foster SO (1997) Sensitivity of NT surveillance system in India. Indian Pediatrics, 34:398–401. Smythe PM, Bowie MD, Voss TJV (1974) Treatment of tetanus neonatorum with muscle relaxant and intermittent positive pressure ventilation. British Medical Journal, 1:223–226. Sokal DC et al. (1988) Mortality from neonatal tetanus in rural Cote d’Ivoire. Bulletin of the World Health Organization, 66(1):69–76. Sow A, Coll AM, Diop-Mar I (1985) Prognosis and immediate causes of death in tetanus neonatorum in Africa. In: Proceedings of the Seventh International Conference on Tetanus, Copanello 10–15 September 1984. G Nistico, P Mastroeni and M Pitzurra eds., pp. 517–539. Rome, Gangemi Publishing Company. Sow PS et al. (1995) Les aspects épidémiologiques du tétanos néonatal à Dakar [Epidemiological aspects of neonatal tetanus in Dakar]. Médecine et Maladies Infectieuses, 25:1178–1182. Spivey OS, Grulee CG, Hickman BT (1953) Tetanus neonatorum: review of 26 cases. The Journal of Pediatrics, 345–351. Tetanus 197 Stahlie TD (1960) The role of tetanus neonatorum in infant mortality in Thailand. Journal of Tropical Pediatrics, 6:15–18. Stanfield JO and Galazka A (1984) Neonatal tetanus in the world today. Bulletin of the World Health Organization, 62(4):647–669. Steinglass R, Brenzel L, Percy A (1993) Tetanus. In: Disease control priorities in developing countries. Jamison DT et al. eds., pp. 189–220. New York, Oxford Medical Publication. Stroh G et al. (1987) Measurement of mortality from neonatal tetanus in Burma. Bulletin of the World Health Organization, 65(3):309–316. Suleman O (1982) Mortality from tetanus neonaturm in Punjab (Pakistan). Pakistan Pediatric Journal, 6(2-3)152–183. Sutter RW et al. (1990) Assessment of vital statistics and surveillance data for monitoring tetanus mortality, US, 1979–1984. American Journal of Epidemiology, 131(1):132–142. Sutter RW, Orenstein WA, Wassilak SG (1994). Tetanus. In: Infectious diseases. 5th ed. Hoeprich PD, Jordan MC, Ronald AR eds., pp. 1175–1185. Lippincott Comp, Philadelphia, eds. Taha TE, Gray RH, Abdelwahab MM (1993) Determinants of neonatal mortality in central Sudan. Annals of Tropical Pediatrics, 13:359–364. Teknetzi P, Manios S, Katsouyanopoulos V (1983) Neonatal tetanus: long term residual handicaps. Archives of Disease in Children, 58:68–69. Thachil RT, Philip B, Sridhar CB (1993) Temporo-mandibular dislocation: a complication of tetanus. Journal of Tropical Medicine and Hygiene, 96:60–61. Thanh NT et al. (1990) Preliminary survey of neonatal tetanus mortality rate in Vietnam. Asia-Pacific Journal of Public Health, 4(4):271–273. Tompkins AB (1958) Neonatal tetanus in Nigeria. British Medical Journal, 1: 1382–1385. Traverso HP et al. (1989) Ghee applications to the umbilical cord: a risk factor for neonatal tetanus. Lancet, 8636:486–488. Tutuncuoglu S et al. (1994) The evaluation of late sequelae of tetanus infection. Indian Journal of Pediatrics, 61:263–267. Vakil BJ, Tulpule TH, Rananvare MM (1965) The cause of death in tetanus. In: Proceedings of the First International Conference on Tetanus, Bombay, 8–10 November 1963. Patel JC, ed., pp. 282–294. Study Group of Tetanus, King Edward Memorial Hospital, Bombay. Verma M, Dhanwade SM, Singh T (1989) Clinical study of tetanus neonatorum. Indian Journal of Medical Sciences, 43(9):235–238. Vernacchio L et al. (1993) Neonatal tetanus in Peru: Risk assessment with modified enzyme-linked-immunosorbent assay and toxoid skin test. American Journal of Public Health, 83(12):1754–1756. Veronesi R, Vitule Filho A, Ferrareto I (1975) Vertebral fractures in tetanus: a clinical and roentgenographic study of 161 cases. In: Proceedings of the Fourth International Conference on Tetanus, Dakar, Senegal 6–12 April 1975. Lyon, Fondation Merieux, pp. 291–304. 198 Global Epidemiology of Infectious Diseases Waheed A et al. (1981) Neonatal tetanus. Pakistan Pediatric Journal, 8(1):28–32. Wassilak SG, Orenstein WA, Sutter RW (1997) Tetanus toxoid. In: Vaccines, 3rd ed. Plotkin SA, Orenstein WA, Philadelphia, Saunders, eds, pp. 1–12. Weinstein BB, Beacham WD (1941) Postabortal tetanus. review of the literature, and a report of 14 additional cases. American Journal of Obstetrics and Gynecology, 42:1031–1040. Wilkinson JL (1961) Neonatal tetanus in Sierra Leone. British Medical Journal, 6:1721–1723. Woody RC, Ross EM (1989) Neonatal tetanus (St Kilda, 19th century). Lancet, 1:1339. Woodruff AW et al. (1984) Neonatal tetanus: mode of infection, prevalence, and prevention in Southern Sudan. Lancet, 18:378–379. World Health Organization (1977) Surveillance of tetanus. Weekly Epidemiological Record, 52:286. World Health Organization (1978) Surveillance of tetanus. Weekly Epidemiological Record, 53:361–368. World Health Organization (1982) Prevention of neonatal tetanus. Weekly Epidemiological Record, 57(18):137–144. World Health Organization (1982b) Prevention of neonatal tetanus. Report of a meeting. Lahore, 22–25 February 1982. (WHO EMRO/ SEARO Technical) Publication No. 3 and No. 7. Sep 1982 EM-SEA/MTG.PREV.NNL.TTN/8. World Health Organization (1983) Neonatal tetanus mortality survey. Weekly Epidemiological Record, 58(10):71–72. World Health Organization (1987a) Expanded Programme on Immunization. Neonatal tetanus mortality surveys. Weekly Epidemiological Record, 62(41): 305–308. World Health Organization (1987b) Expanded Programme on Immunization. Neonatal tetanus mortality surveys. Weekly Epidemiological Record, 62(44): 332–335. World Health Organization (1988) Report on a visit to the Syrian Arab Republic, 14–26 March 1988 (unpublished). World Health Organization (1991) Maternal mortality: a global factbook. World Health Organization, Geneva. (WHO/MCH/MSM/91.3). World Health Organization (1993a) Expanded Programme on Immunization. Neonatal tetanus. Weekly Epidemiological Record, 68(16):109–112. World Health Organization (1993b) Expanded Programme on Immunization: reassessment of the neonatal tetanus problem. Weekly Epidemiological Record, 68(28):201–204. World Health Organization (1993c) Expanded Programme on Immunization: elimination of neonatal tetanus. Weekly Epidemiological Record, 68(46): 337–341. World Health Organization (1995) Computerized EPI information system. Geneva, World Health Organization (WHO/EPI/CEIS 95.1). Tetanus 199 World Health Organization (1996) Coverage of maternity care: a listing of available information. 4th edition. Geneva, World Health Organization (WHO/ RHT/MSM/96.28). World Health Organization (2003) WHO vaccine-preventable diseases: monitoring system 2003 global summary. Geneva, World Health Organization (WHO/VAB/03.20). Wright R (1960) Tetanus neonatorum. South African Medical Journal, 34:111–113. Yadav YR, Kala PC, Yadav S (1990) Study of site and nature of injury as prognostic factors in tetanus. Tropical Doctor, 20:136–137. Yadav YR, Yadav S, Kala PC (1991) Puerperal tetanus. Journal of the Indian Medical Association, 89:336–337. Yurchuk IS, Luchko AS (1989) Residual effects after tetanus. In: Proceedings of the Eighth International Conference on Tetanus, Leningrad 25–28 August 1987. B Bizzini, Bytchenko B, G Nistico, eds. Pythagora Press, Rome-Milan pp. 647–650. Yusuf B, Solter S, Bertsch D (1991) Impact of a tetanus toxoid immunization mass campaign on neonatal tetanus mortality in Aceh Province, Indonesia. Southeast Asian Journal of Tropical Medicine & Public Health, 22(3):351–356. Zuber PL et al. (1993) Tetanus in Switzerland 1980–1989. European Journal of Epidemiology, 9:617–624. 200 Global Epidemiology of Infectious Diseases Chapter 7 Leprosy D Daumerie Introduction Leprosy is an infectious disease caused by Mycobacterium leprae. Humans are considered to be the main reservoir, and the main mode of transmission is droplet dispersion from multibacillary cases. The clinical spectrum of leprosy varies from a single skin patch that may heal spontaneously to widespread damage to nerves, bones, eyes and vital organs. The disease may produce severe mutilation of the face and extremities, making the psychological trauma of leprosy patients as important as the physical suffering. The onset of leprosy is usually gradual, and the first signs may appear 2 to 10 years after infection (Godal & Negassi 1973, Prasad & Mohamed Ali 1967). According to the cellular immune response to the infectious agent, the number of bacilli that patients harbour may vary from less than 1 million (paucibacillary leprosy or PB) to more than 100 billion (multibacillary leprosy or MB). Before the days of chemotherapy, the duration of illness was almost lifelong. With the introduction of multidrug therapy (MDT), the duration of the disease has been drastically reduced to less than 1 year for PB patients and less than 3 years for MB patients (World Health Organization 1988). Although the fatality rate is negligible in leprosy (less than 1 per 1000, Noordeen 1972), the occurrence of severe disabilities has important consequences. The disability and economic loss, together with the social and psychological problems the patient suffers, are a burden for many communities and societies. A first attempt to estimate the global leprosy burden was made by the World Health Organization in 1966 (Bechelli & Martinez-Dominguez 1966): the estimated total number of cases in the world was 10 786 000. This figure was updated in 1972 and 1976 and again in 1983 and has varied from 10 to 12 million. Estimated prevalence was based on cross-sectional random sample surveys in the most endemic countries, complemented by questionnaires for less endemic countries. For each country, a correction 202 Global Epidemiology of Infectious Diseases factor was applied. Correction factors were derived from reliable information available from a limited number of areas, information on prevalence among children, prevalence of deformities and rapid village surveys. Definition and measurement Leprosy is defined as a chronic human disease resulting from infection with Mycobacterium leprae, which affects mainly nerves and skin. This definition has not changed over time but is of very little use for epidemiological studies (Pannikar 1992). The definition of a case of leprosy is not universally accepted (Fine 1982, 1992). This is because there is no “gold standard” to identify leprosy infection in the large majority of patients (Sirumban et al. 1988). The diagnosis of leprosy is mainly based on clinical grounds and therefore lacks specificity, notwithstanding intra-observer and inter-observer variations (Gupte et al. 1990, Ponnighaus & Fine 1988). Newell (1966) said “there is no definite, finite or absolute test, sign or finding which can be said to divide a person with leprosy infection or leprosy illness from the rest of the population”. Clinical, bacteriological, histopathological and immunological tools are all unsatisfactory with regard to reaching a high positive predictive value for screening leprosy in the community. In view of the above, the WHO has proposed an operational definition of a case of leprosy as “a person showing clinical signs of leprosy, with or without bacteriological confirmation of the diagnosis, and requiring chemotherapy”. (World Health Organization 1988). This definition excludes individuals cured of the infection but having residual disabilities attributable to leprosy. The classification of various forms of leprosy is also very controversial. For operational purposes, the WHO has proposed classifying patients Box 7.1 ICD 9 030.9 030.0 030.1 Categories relating to leprosy in the International Classification of Diseases Leprosy Lepromatous leprosy Tuberculoid leprosy ICD 10 A30 A30.0 A30.1 A30.2 A30.3 A30.4 A30.5 A30.8 A30.9 B92 Leprosy Indeterminate Tuberculoid Borderline tuberculoid Borderline Borderline lepromatous Lepromatous Other forms Unspecified Sequelae of leprosy Leprosy 203 as either paucibacillary or multibacillary leprosy cases (World Health Organization Study Group 1985). The paucibacillary group will include smear-negative, indeterminate, tuberculoid, and borderline tuberculoid cases. The multibacillary group will include all smear-positive cases. Most of the endemic countries use the operational definitions proposed by WHO for the preparation of annual reports (Lechat, Mission & Walter, 1981). This has greatly contributed to standardizing information, but can make the comparison of time series difficult, particularly with information collected before the 1980s. With regard to disabilities caused by leprosy, the situation is more complex and several grading systems have been developed. In 1969, the WHO proposed a simple classification, which was modified in 1988. The WHO’s 3-grade classification (see Box 7.3) is a simple tool allowing rapid assessment of the problem in the field. Data sources In many parts of the world, sample surveys and even total population surveys have been conducted to assess the magnitude of the leprosy problem at national level (Burkina Faso, India, Indonesia, Myanmar). These surveys were very useful in enabling a better understanding of the distribution of the disease, and in standardizing procedures. They were acceptably costeffective (manageable sample size) when prevalence of the disease was about 1 per cent. Prevalence and incidence of leprosy are, however, decreasing in most endemic countries, and surveys have become too expensive, time-consuming and often inadequate (Jesudasan et al. 1984). Statistical methods to assess small rates, especially for very unevenly distributed events, are less robust. For these reasons, sample surveys are conducted only in special situations and in limited places, mainly for research purposes. But the results of such surveys do not help in estimating the leprosy problem at national or global levels (Fine 1981). Since 1985, the leprosy situation has undergone major changes following the widespread introduction of multidrug therapy (MDT) (World Health Organization Study Group 1982). With the implementation of the global strategy for elimination of leprosy as a public health problem, it was decided to use existing information and to improve existing information systems in order to update estimates annually (Noordeen, Lopez Bravo & Sundaresan, 1992). Recent information on registered cases shows that 25 countries contribute nearly 95 per cent of all registered cases in the world (International Federation of Anti-Leprosy Associations 1993, Noordeen, Lopez Bravo & Daumerie 1991). Therefore, in order to make a reasonably reliable global estimate of the prevalence of the disease, greater attention has been given to making estimates for those 25 countries by a complete review of information on registered cases. In general, estimates are obtained by applying correction factors to registered cases, in addition to extrapolating from 204 Global Epidemiology of Infectious Diseases other related information provided by the WHO and using results from the most recent total or random-sample population surveys. Most of the information available on the leprosy burden in the world is based on registration. Annual reports from most endemic countries provide point prevalence, annual detection, treatment coverage and number of patients released from registers. Some countries provide more details, such as age-specific detection (under 15 years of age or adults), the proportion of multibacillary patients among new cases and the proportion of disabled patients (WHO classification, grade 2, see Box 7.3) among new cases. Information generated by national information systems is supplemented by: surveys (total population surveys, selected population surveys, random sample surveys); WHO questionnaires; regular evaluations of national programmes; reports from various consultants; prospective studies for research purpose (such as the vaccine trials in India, Venezuela and Malawi). It is clear that data collected by control programmes are mainly actionoriented and provide limited information on the epidemiological pattern of the disease (Fine 1992). More detailed information is collected through special studies. In addition, epidemiometric modelling, as developed by Lechat et al. (1986) is used for a better understanding of transmission of the disease and the impact of control measures. A new leprosy simulation model is being developed in India to enable the best use to be made of existing information, to validate estimates, and to predict future leprosy trends according to various control strategies. Estimation Methods Leprosy estimates at the global level are coordinated by the WHO. Although these estimates are crude, they have nevertheless been found to be useful for planning purposes and for setting priorities. Estimates of prevalence are based on the operational case definition of leprosy, i.e. on the number of cases requiring chemotherapy. Greater attention is given to the most endemic countries, using the following methods: systematic review of all available information, including re-examination of registered cases in Brazil, India, Madagascar, Nigeria, Sudan (Rao & Sirumban 1990); review of latest epidemiological surveys; discussion with national programme managers; Leprosy 205 making estimates based on registered figures at country level, follow- ing the guidelines outlined in Box 7.2 (World Health Organization 1994b). Information on incidence of leprosy is very limited. Some prospective studies are carried out in highly endemic areas for research purposes; however, the results cannot be used for extrapolation to country or regional levels. Considering the major difficulties in assessing the incidence of leprosy (case definition, incubation period, age at onset, delay between onset and detection), incidence is not estimated and it is assumed that reported detection reflects incidence. Age-specific incidence and prevalence Detailed information is not available on the age-specific incidence and prevalence of leprosy. Some programmes report the proportion of children (0–14 years) among newly detected cases. For the purpose of making the estimates given here, the following assumptions were made, based on a number of studies (Brubaker, Meyers & Bourland 1985, Irgens & Skjaerven, Becx-Bleumink 1992, Ponnighaus et al. 1994) and on other expert opinion. In endemic countries (crude prevalence above 1 per 10 000 population) 1 per cent of incident and prevalent cases are found in the 0–4 year age group, 20 per cent in the 5–14 year age group, 60 per cent in the 15-44 year age group, 15 per cent in 45–59 year age group and 4 per cent in over-60 year age group, as reported by Noordeen during the dapsone era (before 1981). For the MDT era (after 1981), the distribution is as follows: 0 per cent, 18 per cent, 64 per cent, 20 per cent, 9 per cent (Ponnighaus et al. 1994). In non-endemic countries, most of the incident and prevalent cases are from age groups over 45 years. This assumption is based on observations in China, Norway and the United States of America. Age at onset for each age group Most of the available information refers to age at detection for two age groups: 0–14 years; and 15 years and above. In this report, age at onset is estimated by using the median age of the age group. Sex-specific incidence and prevalence In the past, in many parts of the world, males were reported to be more frequently affected by leprosy than females. Sex difference in leprosy is often attributed to ascertainment bias, and it seems reasonable to assume that the risk is equal for males and females. For this reason, tables for each region do not show sex distribution, and specific rates will vary only according to the size of population by sex. 206 Global Epidemiology of Infectious Diseases Box 7.2 Method for estimating leprosy prevalence based on registered figures at country level Step 1 List all areas of the country. An area is defined using administrative, geographical or operational criteria. Step 2 For each area, indicate the registered prevalence, detection, the year of multidrug therapy (MDT) implementation, and the year of 100 per cent MDT coverage. Calculate the prevalence-to-detection ratio. Step 3 For each area, assess the situation and apply the appropriate score, as follows. Score 1 No leprosy control and no information available in the area. Score 2 MDT has not been implemented (MDT coverage below 20 per cent), and the prevalence-to-detection ratio is equal to or above 5. Score 3 MDT has not been implemented (MDT coverage below 20 per cent), and the prevalence-to-detection ratio is below 5. Score 4 MDT coverage has been 20 to 100 per cent for less than 5 years. Score 5 MDT coverage has been about 100 per cent for 5 years or more. Step 4 Calculate the correction factor. Extrapolation factors are based on experience in implementing MDT in various countries (Sundaresan 1990). Score 1 Population of the area × prevalence rate of the whole country. This situation is the most difficult to assess and may reflect: the absence of leprosy endemicity; the absence of health services; a low leprosy control coverage; a combination of the above. The easiest way to estimate the number of potential cases is to apply the existing national prevalence rate to the population of the area. Of course, this is not very satisfactory and urgent action should be taken to implement leprosy control activities in such areas. Score 2 Registered cases in the area × 1. Very often in this situation the number of registered cases is cumulative. Reviewing registers and re-examining patients would lead to a 50 per cent reduction in the prevalence, but it could be estimated that an equal number of existing patients would not have been detected. Score 3 Registered cases in the area × 2. Registers have been updated, and the MDT is being implemented. In this situation we generally observed a dramatic increase in detection. Score 4 Registered cases in the area × 1.5. MDT is being implemented and the backlog has been reduced. Score 5 Registered cases in the area multiplication sign × 1.2. The registered figures are very close to reality, and the correction factor reflects mainly the delay in detection. Leprosy 207 Duration of the disease The duration of the disease is computed using the following assumptions: the fatality rate is negligible; the self-healing rate is 10 per cent a year for paucibacillary patients (Sirumban 1988); monotherapy treatment cures paucibacillary patients (PB) in 5 years and multibacillary patients (MB) in 15 years; assuming that 75 per cent of the patients are PB, the average duration with monotherapy is 7.5 years; multidrug therapy (MDT) cures PB patients in 1 year and MB patients in 3 years so the average duration with MDT is 1.5 years; the relapse rate is negligible. For each region, the duration was computed according to the MDT coverage (proportion of estimated cases treated with MDT). Biases and Limitations of Estimates Crude prevalence and detection rates derived from registered figures are reasonably accurate. Age-specific and sex-specific indicators are, however, based only on expert opinion and need to be refined through special studies. Duration of the disease is based on assumptions made according to the most recent results of research projects on MDT treatment. Assumptions on the natural history of the disease are based on research studies conducted before 1981 and on epidemiological analysis of information collected by special programmes. Prevalence and incidence Estimates of prevalence and incidence are derived mainly from registered figures and the primary data are subject to biases arising from various factors: standardization; coverage of information systems; and sensitivity and specificity of the diagnosis. Efforts were made to standardize definitions and reporting systems as far as possible, but there is no simple way to validate the quality of the information. Figures for age-specific prevalence and incidence were estimated using well-documented studies. With the introduction of MDT, leprosy trends are changing rapidly and most likely the risk of infection is also changing. In the context of declining prevalence, it becomes more difficult to use stable assumptions (constant risk of infection and exposure). Duration of the disease During the past decade, the extensive use of MDT and the improvement of many leprosy control programmes have led to a drastic revision of the concepts of the disease. The duration of the disease is still controversial. 208 Global Epidemiology of Infectious Diseases The main issue is that the definition of cure is not universally accepted. For operational purposes, the WHO recommends that a patient who has completed an adequate course of MDT should be considered cured. The question of considering as leprosy patients those who have completed treatment but who are disabled as a result of the disease, is very much debated. Considering that a large majority of incident cases are cured without disability, it was considered practical to estimate an average duration of the disease taking into account self-healing rates, cure rates with monotherapy, and cure rates with MDT. The main limitation of this method is that drop-out rates and survival statistics are not included. Complications and disabilities after cure While the estimates relate to cases requiring or receiving chemotherapy, there are a significant number of patients who have been cured and discharged from registers but still have residual deformities (WHO classification grade 2, see Box 7.3). This problem is discussed and tentatively estimated below. Leprosy Estimates By World Bank Region The global number of leprosy cases in 1993 was estimated to be about 2.5 million, of which 1.7 million were registered. The same year, about 600 000 new cases were detected. Sub-Saharan Africa Leprosy is endemic in most of the sub-Saharan African countries: Cote d’Ivoire, Chad, Ethiopia, Guinea, Madagascar, Mali, Mozambique, Nigeria, Niger, Sudan, and Zaire. The national authorities in these countries have reviewed the registered figures and published estimates for 1993. A correction factor of 1.9 was applied to registered figures. Tables 7.1 and 7.2 summarize the results of the major studies in the region. India India contributes 60 per cent of the global leprosy burden. The correction factor applied to registered figures was 1.2. The disease is not equally Table 7.1 Detection and prevalence of leprosy in five African countries, 1986–1989 Sample size 12 683 Prevalence per 100 000 720 Detection per 100 000 360 Central African Rep. (1986) 6 468 1 370 840 Gabon (1988) 6 227 940 360 Congo (1989) 6 542 940 150 Equatorial Guinea (1989) 6 428 580 110 Country (year) Cameroon (1986) Source: Louis et al. (1991). Leprosy 209 distributed in the country and wide variations in prevalence and incidence are observed. The first epidemiometric model was based on data collected in Polambakkam, southern India, and gives an indication of age- and sexspecific incidence as well as mean age at onset. Table 7.3 summarizes the results of epidemiological studies conducted in southern India. Table 7.4 gives specific incidence rates from Polambakkam in 1967–1971. China Leprosy in China is endemic in three south-western provinces. Epidemiological surveillance and specialized services provide accurate information, and the correction factor applied to registered figures was 1.25. Other Asia and Islands The most endemic countries in this region are Bangladesh, Indonesia, Myanmar, Nepal, Philippines, and Viet Nam. Latin America and the Caribbean Brazil is the most endemic country in this region. Motta & Zuniga (1990) predicted the trend of incidence in Brazil based on actual detection rates from 1950 to 1983. The results are summarized in Table 5. Table 7.2 Detection rates per 1000 by age and sex in five African countries Age (Years)a 15–19 20–24 25–29 30–39 40–49 >50 2.0 2.0 5.0 1.3 5.7 2.8 5.5 3.0 6.0 7.5 6.5 5.5 Male Female a. People 0–14 years of age not included in the sample. Source: Louis et al. (1991). Table 7.3 Detection by age group and type of leprosy in four districts, India, 1993 Under 15 years of age Age 15 years and over, (number detected) (number detected) District Chenglepet PB MB Detection rate per 1 000 PB MB PB MB Total — — — — — — 1.64 Chittoor (total population) 1 738 381 1 044 27 — — 1.05 North Arcot (total population) 2 416 10 3 582 651 1.18 0.13 1.31 Visakhapatnam (total population) (0.8)a (0.0)a (0.8)a (0.1)a 0.80 0.10 0.90 a. Detection rate per 1000. 210 Table 7.4 Global Epidemiology of Infectious Diseases Incidence rates by age, sex and type of leprosy, Polambakkam, India, 1967–1971 Male incidence per 10 000 Female incidence per 10 000 Age group (years) PB MB PB MB 0–4 5–9 10–14 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65+ 5 39 53 36 35 41 42 39 33 32 30 30 21 21 0 3 5 8 9 14 16 18 15 11 10 7 8 3 5 30 32 18 20 28 31 32 30 31 28 28 15 42 0 2 3 3 3 3.5 5 4 5 4 3.5 3 2.5 0 Source: Lechat et al. (1986). PB = paucibacillary leprosy; MB = multibacillary leprosy. Table 7.5 Detection rates of leprosy, Brazil: actual rates, 1950 to 1983; predicted rate, 2000 Year Detection per 100 000 1950 1973 1978 1983 2000 5.55 7.40 9.87 13.17 35.03 Source: Molta & Zuniga (1990). Middle Eastern Crescent Egypt, the Islamic Republic of Iran and Pakistan are the most endemic countries in this region. Based on a cluster sample survey conducted in Egypt in 1986, Sundaresan (1986) estimated the age-specific prevalence (using a case-definition differing from the World Health Organization definition); the results are shown in Table 7.6. Estimation of Disabilities Attributable to Leprosy While disability is an important measure in evaluating the leprosy burden and the impact of control programmes, data on disabilities are scarce and not easy to interpret. Most of the epidemiological studies on disabilities attributable leprosy are based on cross-sectional surveys and give an indication of the prevalence of visible disabilities. Since the implementation of multidrug therapy, some experts have been collecting information on cohorts of patients, making it possible to estimate the incidence Leprosy 211 Table 7.6 Age-specific and type-specific prevalence of leprosy in four Governorates, Egypt, 1986 Prevalance per 1000 Age group (years) Sample size PB MB 4 955 4 431 3 980 3 221 4 502 3 542 2 851 1 880 1 925 — 1.35 2.01 1.24 3.55 1.97 3.50 10.630 8.31 — — — 0.62 0.22 2.54 2.80 4.20 8.83 31 2870 2.78 1.43 0–4 5–9 10–14 15–19 20–29 30–39 40–49 50–59 60+ Total Source: Sundaresan (1986). of disability. Based on information provided by countries, Bechelli & Martinez-Dominguez (1966) estimated that globally 3.87 million leprosy patients had disabilities of all grades, and 2 million of these had severe disabilities. In 1975, it was estimated that 3.5 million individuals had been disabled by leprosy (World Health Organization 1976). In 1994, the World Health Organization (1994b) estimated this number at between 2 and 3 million. Definition and measurement of disabilities related to leprosy Recognizing that no single system of grading would meet all requirements, in 1988 the WHO proposed a simplified 3-grade system of classification for collecting general data on disabilities or impairments, as shown in Box 7.3 (World Health Organization 1988). The International classification of impairments, disabilities and handicaps (World Health Organization 1980) is related more to rehabilitation Box 7.3 Classification of grades of disability attributable to leprosy Grade Hands and feet Eyes 0 No anaesthesia, no visible deformity or damage No eye problem attributable to leprosy, no evidence of visual loss 1 Anaesthesia present, but no visible deformity or damage Eye problems attributable to leprosy but vision not severely affected as a result of these (vision 6/60 or better, can count fingers at 6 metres) 2 Visible deformity or damage present Severe visual impairment (vision worse than 6/60, inability to count fingers at 6 metres) 212 Global Epidemiology of Infectious Diseases needs than to the collection of information for epidemiological purposes. It describes three stages: impairment: loss or abnormality of psychological, physiological, or anatomical structure or function; disability: restriction or lack of ability to perform an activity in the manner within the range considered as normal for a human being; handicap: a disadvantage for a given individual, resulting from an impairment or disability, that limits or prevents fulfillment of a role that is normal, depending on age, sex, and social and cultural factors for the individual. Disability in Leprosy: Magnitude of the Problem Before the implementation of multidrug therapy Before the widespread application of MDT, leprosy was generally considered as a lifelong disease, and cross-sectional studies were carried out to estimate prevalence and risk factors for disabilities. The disability rate, defined as the proportion of patients disabled at a given point of time, ranged from 13 to 57 per cent. The most generally observed risk factors were age, duration of the disease, sex, type of leprosy, occupation, and treatment. Results of the most important studies on prevalence of disabilities are summarized in Table 7.7. Almost all the studies show that the disability risk is higher for males and for multibacillary patients, and that this risk Table 7.7 Prevalence of disability among leprosy patients before the implementation of multidrug therapy: results of various studies Type of study Case definition Results Country Reference Patient survey WHO 41.5% Cameroon 37.6% Nigeria 41.5% Thailand Martinez-Dominguez, Bechelli & Patwary (1966) Patient survey Social & physical 165/465 (35.5%) India Noordeen & Srinivasan (1966) Retrospective cohort WHO 15 869 patients from 28% in 1956 to 13% in 1978 India Christian, Jesudasan & Pannikar (1980) Patient survey Deformity 120/1334 (9%) India Srinivasan (1982) Retrospective cohort WHO 48/145 (33.1%) Myanmar Htoon & Win (1994) Retrospective cohort WHO 229/514 (44.6%) India Girdhar et al. (1989) Retrospective cohort WHO 282/1264 (22.3%) India Saha & Das (1993) Patient survey WHO 8122/14 257 (57%) China Zhang et al. (1993) Leprosy 213 increases with age and duration of the disease. Based on these studies, we estimate that, before the implementation of MDT, the number of individuals suffering from severe disabilities because of leprosy varied from 1 million to 4 million at any given time. There is no published literature on the incidence of disabilities attributable to leprosy before 1981. Most leprosy control programmes use the disability rate at detection as an indirect indicator of the effectiveness of case-finding activities. This indicator could be used to approximate annual incidence of disability without interventions. Considering the results of surveys and the proportion of patients who are already disabled at the time of diagnosis, we estimate that the crude disability attack rate ranges from 1 to 3 per 100 person-years. For modelling purposes, Htoon, Bertolli & Kosasih (1993) estimated that 16 per cent of individuals affected by leprosy would be permanently disabled and that all patients will be temporarily disabled, for an average period of 30 days. With multidrug therapy interventions Many experts considered that chemotherapy might increase the incidence of disabilities, especially while using powerful bactericidal drugs. With the introduction of MDT and its relative short duration, the occurrence of reactions—and of subsequent disabilities during and after treatment—was expected to be high. Nevertheless, data accumulated in the field do not confirm this expectation. On the contrary, by reducing the duration of the disease, the incidence and seriousness of leprosy reaction, as well as the incidence of the disease, MDT has also reduced the overall incidence of disabilities. This reduction is not entirely attributable to the efficacy of drugs. Other factors, such as regular monthly contacts with patients, and improved monitoring and treatment of leprosy reactions in the field, have also played an important role in this process. Ponnighaus et al. (1990) have reported a disability incidence of 5 per 1000 person-years in Karonga district (Malawi) and Rao (1994) has observed a disability incidence of 0.68 per 1000 person-years among patients treated with MDT. Table 7.8 summarizes the most recent information on the risk of disabilities for patients treated with MDT. Available information shows clearly that MDT interventions have significantly reduced the risk of disability. This risk ranges from 1 to 5 per 1000 person–years. It seems that application of MDT prevents the occurrence of 80–90 per cent of disability. Situation in 1994 In collaboration with the WHO, the major endemic countries estimated the disability problem defined according to the WHO grading scheme as follows: number of individuals already disabled at the time of case detection: 214 Global Epidemiology of Infectious Diseases 47 100 (8.4 per cent) of 558 066 cases diagnosed in 34 countries, accounting for 88 per cent of the leprosy detected globally; number of individuals estimated to be disabled because of leprosy: 1 255 700 in 11 countries, representing 72 per cent of the global leprosy burden. Based on this information, the WHO estimated that in 1994 the annual incidence of disability attributable to leprosy was about 100 000, and that the prevalence of individuals disabled because of leprosy was between 2 and 3 million (World Health Organization 1995). Burden and intervention Only severe and permanent physical disabilities have been considered here. Temporary physical disabilities before and during treatment, reactions and complications, and the inconvenience of sustaining long-duration treatment, should be considered in order to estimate the global burden. Moreover, among all infectious diseases, leprosy has a specific cultural connotation, and it is difficult to assess the individual social and psychological impact it causes. With or without disabilities, any individual in any part of the world classified as a leprosy patient will be exposed to social problems and stigma. The benefits of leprosy control programmes are very often underestimated. The elimination strategy based on MDT has contributed to an impressive reduction in the global prevalence. Estimated prevalence was reduced from 10 million in 1966 to 5 million in 1991, and to 2.5 million Table 7.8 Impact of multidrug therapy on disability related to leprosy— prevalence of disability: results of various studies Type of study Case definition Results Country Reference Retrospective 5-year cohort Modified WHO 1654 patients: 2.9 to 8 per 1000 person years (average 5 per 1000) Malawi Ponnighaus et al. (1990) Retrospective 4-year cohort WHO 12/482 (2.5%) Malawi Boerrigter et al. (1991) Retrospective cohort Nerve damage 87/2161 (4%) India Smith (1992) Retrospective cohort Deformity 5655 urban patients: 1.8% India Thomas (1993) India Rao (1994) 6104 rural patients: 2% Retrospective WHO 7- year cohort 2285 patients: 0.68 per 1000 person years (MB 6.14, PB 0.19) MB = multibacillary leprosy; PB = paucibacillary leprosy. Leprosy 215 in 1993. Many factors have contributed to the reduced leprosy burden, but it is clear that MDT interventions have had a tremendous impact on prevalence and thus on the caseload and disabilities. The efficacy and acceptability of MDT are now very well known to health workers and the patients themselves. This contributes to increased opportunities for diagnosis and to reduced stigma. The incidence of the disease is decreasing in many parts of the world. Even if it is difficult to measure to what extent MDT contributes to this reduction, it is clear that secondary and primary resistance to the drugs used previously would have hampered leprosy control. It is estimated that MDT has averted 0.5 to 1 million relapses compared to previous interventions. In India, it is estimated that extensive implementation of MDT will avert 1.2 million cases. Since the implementation of MDT in 1982, about 6.5 million patients have been cured with MDT, including about 2.5 million old cases (backlog). About 4 million new cases have been detected during the past 10 years. Assuming that 30 per cent of those patients were already disabled before starting treatment, we estimate that MDT intervention has prevented 1 to 2 million people from becoming disabled. The cost of implementing the elimination strategy varies from US$30 to US$400 per patient (global average US$100) according to prevalence, the PB/MB ratio and the existence of a health infrastructure (Htoon, Bestolliz & Kosasik 1993, World Health Organization 1994). A cost-effectiveness analysis indicates that the cost per year of healthy life gained (YHLG) from implementing MDT is very low, from US$10 to US$38 per YHLG (Htoon, Bertolliz & Kosasik 1993). Conclusions Leprosy is a public health problem in terms of widespread occurrence in the world, its potential to cause permanent progressive disability, and the social consequences unique to the disease. The significant progress made in leprosy control enabled most of the endemic countries to reach the goal of eliminating leprosy as a public health problem (prevalence below one per 10 000 population) by the year 2000. Update The elimination of leprosy as a public health problem (prevalence below one per 10 000 population) was attained at the global level by the end of 2000. At the beginning of 2003, the number of leprosy patients in the world was around 530 000, as reported by 106 countries. About 615 000 new cases were detected during 2002. Among newly detected cases 17 per cent were children (below the age of 15 years), 39 per cent had MB based on the clinical classification (more than 5 skin lesions), and 4 per cent presented grade 2 disability at the time of diagnosis. With a cumulative number of about 13 million patients cured with MDT, the numbers of relapses remain 216 Global Epidemiology of Infectious Diseases remarkably low, at less than one case per 1000 patients per year. No drug resistance following MDT has been reported. The number of countries showing prevalence rates above 1 per 10 000 population has been reduced from 122 in 1985 to 12 at the end of 2002. Today, Brazil, India, Nepal, Madagascar, and Mozambique are the most endemic countries. References Bechelli LM, Martinez-Dominguez V (1966) The leprosy problem in the world. Bulletin of the World Health Organization, 34:811–826. Becx-Bleumink M (1992) Multidrug therapy in the control of leprosy, A field study. Amsterdam University. Boerrigter G et al. (1991) Four-year follow-up results of a WHO-recommended multiple-drug regimen in paucibacillary leprosy patients in Malawi. International Journal of Leprosy and Other Mycobacterial Diseases, 59:255–261. Brubaker ML, Meyers WM, Bourland J(1985) Leprosy in children one year of age and under. International Journal of Leprosy and Other Mycobacterial Diseases, 53:517–523. Christian M, Jesudasan K, Pannikar VK (1980) Deformity profile in Gudiyattam Taluk: 1956–1978 (unpublished data). Fine PEM (1981) Problems in the collection and analysis of data in leprosy studies. Leprosy Review, 52:197s–206s. Fine PEM (1982) Leprosy: the epidemiology of a slow bacterium. Epidemiologic Reviews, 4:161–188. Fine PEM (1992) Reflections on the elimination of leprosy. International Journal of Leprosy and Other Mycobacterial Diseases, 60:71–80. Girdhar M et al. (1989) Pattern of leprosy disabilities in Gorakhpur (Uttar Pradesh) Indian Journal of Leprosy, 61:503–513. Godal T, Negassi K (1973) Subclinical infection in leprosy. British Medical Journal, 3:557–559. Gupte MD et al. (1990) Inter-observer agreement and clinical diagnosis of leprosy for prophylaxis studies. Indian Journal of Leprosy, 62:281–295. Htoon MT, Bertolli J, Kosasih LD (1993) Leprosy. In: Disease control priorities in developing countries. Jamison DT, et al., eds. New York, Oxford Medical Publication, pp. 261–280. Htoon MT, Win Z (1994) Disabilities among rural leprosy patients in Myanmar. International Journal of Leprosy and Other Mycobacterial Diseases, 62: 126–130. Indian Association of Leprologists (1993) Report on the workshop on impact of MDT on trend of leprosy. Madras, 20–21 July 1993. International Federation of Anti-Leprosy Associations (1993) ILEP Statistics. London, ILEP. Leprosy 217 Irgens LM, Skjaerven R (1985) Secular trends in age at onset, sex ratio, and type index in leprosy observed during declining incidence rates. American Journal of Epidemiology, 122:695–705. Jesudasan K et al. (1984) Time trends in the analysis of incidence rate of leprosy among household contacts. Indian Journal of Leprosy, 56:792–806. Lechat MF, Misson CB, Walter J (1981) OMSLEP: recording and reporting system for leprosy patients. Geneva, World Health Organization. Lechat MF et al. (1986) Analysis of trends in the occurrence of leprosy. World Health Statistics Quarterly, 39:129–137. Louis JP et al. (1991) Evaluation of endemic leprosy in 5 OCEAC member countries. Bulletin de Liaison et de Documentation de l’OCEAC, 97:5–10. Martinez-Dominguez V, Bechelli LM, Patwary KM (1966) WHO surveys of disabilities in leprosy in northern Nigeria (Katsina), Cameroon and Thailand (Khon Kaen). International Journal of Leprosy, 34:244–254. Motta CP, Zuniga M (1990) Time trends of Hansen’s disease in Brazil. International Journal of Leprosy and Other Mycobacterial Diseases, 58:453–461. Newell KW (1966) An epidemiologist's view of leprosy. Bulletin of the World Health Organization, 34:827–857. Noordeen SK (1972) Mortality in leprosy. Indian Journal of Medical Research, 60:439–445. Noordeen SK, Srinivasan H (1966) Epidemiology of disability in leprosy: I. a general study of disability among male leprosy patients above fifteen years of age. International Journal of Leprosy and Other Mycobacterial Diseases, 34: 159–69. Noordeen SK, Lopez Bravo L and Daumerie D (1991) Global review of multidrug therapy (MDT) in leprosy. World Health Statistics Quarterly, 44:2–15. Noordeen SK, Lopez Bravo L, Sundaresan TK (1992) Estimated number of leprosy cases in the world. Bulletin of the World Health Organization, 70:7–10. Pannikar VK (1992) Defining a case of leprosy. Leprosy Review, 63:61s–65s. Ponnighaus JM, Fine PEM (1988) Leprosy in Malawi: 1. — sensitivity and specificity of the diagnosis and the search for risk factors for leprosy. Transactions of the Royal Society of Tropical Medicine and Hygiene, 82:803–809. Ponnighaus IM et al. (1990) Disabilities in leprosy patients ascertained in a total population survey in Karonga District, Northern Malawi. Leprosy Review, 61:366–374. Ponnighaus JM et al. (1994) Incidence rates of leprosy in Karonga District, northern Malawi: patterns by age, sex, BCG status and classification. International Journal of Leprosy, 62:10–23. Prasad KV, Ali PM (1967) Incubation period of leprosy. Indian Journal of Medical Research, 55:29–42. Rao PS (1994) Deformity incidence in leprosy patients treated with multidrug treatment. Indian Journal of Leprosy, 66:449–454. Rao PS, Sirumban P (1990) Screening of registered leprosy cases and its effects on prevalence rate. Indian Journal of Leprosy, 62:180–185. 218 Global Epidemiology of Infectious Diseases Saha SP, Das KK (1993) Disability pattern amongst leprosy cases in an urban area Sirumban (1975) (Calcutta). Indian Journal of Leprosy, 65:305–314. Sirumban PS (1988) Healing time in untreated paucibacillary leprosy: a crosssectional study. International Journal of Leprosy and Other Mycobacterial Diseases, 56:223–227. Sirumban P et al. (1988) Diagnostic value of cardinal signs/symptoms in paucibacillary leprosy. Indian Journal of Leprosy, 60:207–214. Smith WC (1992) The epidemiology of disability in leprosy including risk factors. Leprosy Review, 63:23s–30s. Srinivasan H (1982) Changes in epidemiology of deformity in leprosy in a rural area in south India. Indian Journal of Medical Research, 76:795–803. Sundaresan TK (1986) Analysis of sample survey data on leprosy in Egypt and suggestions for future plans. Geneva, World Health Organization. Sundaresan TK (1990) Estimation of leprosy prevalence in India and validation of district-wise prevalence data. Geneva, World Health Organization. Thomas A (1993) Impact of MDT on disability profile. Paper presented during the International Leprosy Congress, Orlando, 1993. World Health Organization (1976) Disability prevention and rehabilitation. World Health Organization Chronicle, 30:324–328. World Health Organization (1980) International classification of impairments, disabilities, and handicaps. Geneva, World Health Organization World Health Organization (1982) Chemotherapy of leprosy for control programmes. Geneva, World Health Organization. (Technical Report Series, No. 675). World Health Organization (1985) Epidemiology of leprosy in relation to control. Geneva, World Health Organization. (Technical Report Series, No. 716). World Health Organization (1988) Expert Committee on Leprosy. Sixth Report. Geneva, World Health Organization (Technical Report Series, No. 768). World Health Organization (1994a) Global strategy for the elimination of leprosy as a public health problem. Geneva, World Health Organization (WHO/CTD/ LEP/94.2). World Health Organization (1994b) Guidelines for conducting a leprosy elimination programme review. Geneva, World Health Organization. World Health Organization (1994c) Report from the major endemic countries. International Conference on Elimination of Leprosy, 4–7 July 1994, Hanoi, Viet Nam. World Health Organization (1994d) Progress towards eliminating leprosy as a public health problem. Weekly Epidemiological Record, 69:145–151, 153–157. World Health Organization (1995). Leprosy disabilities: magnitude of the problem. Weekly Epidemiological Record, 70: 269–276. Zhang G et al. (1993) An epidemiological survey of deformities and disabilities among 14 257 cases of leprosy in 11 counties. Leprosy Review, 64:143–149. Chapter 8 Dengue and dengue haemorrhagic fever James W LeDuc, Karin Esteves, Norman G. Gratz Introduction Dengue, a mosquito-borne viral disease, is becoming an increasing public health problem in developing countries. This was recognized by the WHO, which confirmed in a resolution at the forty-sixth World Health Assembly in 1993 that dengue prevention and control should be among the priorities of the Organization. Dengue is caused by any of four antigenically and genetically distinct viruses of the family Flaviviridae, named dengue-1, -2, -3 and -4 (Craven 1991). Humans are the primary vertebrate hosts of these viruses, although there is evidence of a silent zoonotic cycle involving mosquitoes and monkeys in parts of Asia and Africa (Rudnick 1965, Fagbami, Monath & Fabiyi 1977, Rudnick 1984, Cordellier et al. 1983). The Aedes aegypti mosquito is the principal urban vector (Clark & Casals 1965). Dengue fever is characterized by an abrupt onset of fever, headache, myalgia, loss of appetite, and varying gastrointestinal symptoms, often accompanied by rash and bone or joint pains (Rush 1789, Hammon, Rudnick & Sather 1960, Ehrenkranz et al. 1971, San Juan Laboratories 1979, Kuberski et al. 1977, Lopez-Correa et al. 1979). Symptoms persist for 3–7 days, and while convalescence may be prolonged for several weeks, mortality is rare. Homologous immunity is lifelong, but cross-protection to other dengue viruses is not elicited and, indeed, there is evidence to suggest that heterologous antibodies may form infectious immune complexes which may increase the severity of subsequent infections (Halstead, Nimmannitya & Margiotta 1969, Halstead 1970, 1980, Kliks et al. 1989). Infected people are infective for feeding mosquitoes during the febrile phase of their illness, and mosquitoes may then transmit the virus to others at subsequent blood-meals after an incubation period of approximately one week, depending upon ambient temperatures (Watts et al. 1987). Infected female mosquitoes may also pass the virus on to progeny by transovarial transmission (Rosen et al. 1983, Khin & Than 1983, Hull et al. 1984, Freier & Rosen 1987, 1988, Rosen 1987, 1988, Fauran, 220 Global Epidemiology of Infectious Diseases Laille & Moreau 1990, Shroyer 1990, Mitchel & Miller 1990, de Souza & Freier 1991, Bosio et al. 1992, Serufo et al. 1993). Mosquitoes infected transovarially may transmit dengue virus to susceptible humans at their first blood-meal. Once infected, mosquitoes remain infected for life and may transmit dengue virus either during probing, interrupted blood feeding, or when feeding to repletion. Thus, a single infected mosquito may transmit the virus to several susceptible humans over its lifetime. Dengue haemorrhagic fever (DHF) and dengue shock syndrome (DSS) are more severe manifestations of dengue infection, and are most often associated with infection of people with pre-existing antibodies to dengue, either actively or passively acquired (Quintos et al. 1954, Hammon et al. 1960, Halstead, Nimmannitya & Margiotta 1969, Ramirez-Ronda et al. 1979, Halstead 1970, 1980). The clinical course follows that of classic dengue fever initially; however, as the febrile phase subsides, the patient’s condition rapidly deteriorates with signs of circulatory failure, neurological manifestations, and hypovolemic shock with potential fatal outcome if prompt, proper, clinical management is not implemented. Haemorrhage may be evident by petechiae, easy bruising, bleeding at injection sites, and occasionally gastrointestinal bleeding, especially where gastric ulcers already exist (Tsai et al. 1991). There is no specific treatment for dengue fever. Although convalescence may last several weeks following dengue infection, especially among adults, disabling sequelae are not thought to follow infection. Dengue haemorrhagic fever and DSS require symptomatic treatment, coupled with careful assessment of fluid and electrolyte balance for management of shock (World Health Organization 1986, Craven 1991). Areas where clinicians are not familiar with the proper treatment of DHF/DSS may have case fatality rates substantially higher than those where physicians are experienced in the proper treatment of the disease. Diagnosis is made on clinical grounds and confirmed by laboratory demonstration of specific immune response, virus isolation, or identification of dengue virus nucleic acid sequences in clinical material (Russell & McCown 1972, Russell, Brandt & Dalrymple 1980, Pan American Health Organization 1982, Repik et al. 1983, Henchal et al. 1986, Kerschner et al. 1986, World Health Organization 1986). Dengue is the most widespread vector-borne virus disease in the world, and DHF/DSS are rapidly increasing in incidence in many tropical areas (Halstead 1992, 1993a, 1993b). This dramatic increase is a result of changes in human lifestyle, increased international travel, and urban crowding. Dengue has been known for at least 200 years, but DHF/DSS has only been routinely recognized since the 1950s, although there is evidence to suggest that DHF occurred as early as 1897 in north-eastern Australia (Hare 1898, Craven 1991, Halstead 1993a). Following the end of the Second World War, Asian cities experienced rapid growth and urban crowding, and with these came increased habitat for the vector mosquito of dengue, Aedes aegypti, which selectively breeds in water containers associated with human habitations. These breeding sites can include water jugs, Dengue and dengue haemorrhagic fever 221 blocked drain spouts, decorative plant containers, discarded refuse that may retain water, and many other items associated with modern life (Gratz 1993a). A pattern of dengue transmission then became evident, especially in major cities of South East Asia, first with increasingly frequent epidemics of classic dengue, then merging to continuous endemic transmission and appearance of DHS/DSS, initially sporadically but becoming more common within the last 20 years, until these too are now endemic (Akiyama 1993, Gratz 1993b, Gubler 1993, Itoh 1993, Kojima 1993, Lam 1993). This pattern has been seen in varying degrees throughout much of tropical Asia, from the southern provinces of China to the west coast of India, and south to northern Australia and many Pacific island nations. The same evolutionary pattern is now being repeated in the urban centres of Latin America (Gubler 1993, Halstead 1992, 1993a). In the 1960s, a hemisphere-wide Aedes aegypti eradication programme was in existence, and nearly succeeded in total eradication of the vector. However, the decision was made to change the strategy from eradication to control. As a result, Aedes aegypti has now reinfested virtually all areas where it was once eliminated, and has even expanded to new habitats. As a result of this reinfestation, transmission of dengue has recently occurred in almost all parts of Latin America and the Caribbean (Pan American Health Organization 1993). It is currently limited to increasingly frequent epidemics of dengue fever, often coupled with sporadic cases or outbreaks of DHF/DSS. The pattern is, however, identical to that experienced in Asia during the 1950s and 1960s, and it is clear that DHF/DSS will become increasingly common in the years to come (Gubler 1993). Indeed, in 1994 a major outbreak of dengue and DHF occurred in north-eastern Brazil, and some experts have suggested that as many as 300 000 cases might have occurred. In Africa, similar conditions exist, but reporting has been fragmentary, and there are no reliable estimates of the number of persons infected or the virus serotypes in circulation (Monath et al. 1974, Fagbami, Monath & Fabiyi 1977, Roche et al. 1983). As a typical example, however, in 1993 an outbreak of dengue occurred in Comoros, during which an estimated 60 000 cases occurred (Pan American Health Organization 1993). Review of major studies and data available Because of the variation in clinical presentation and lack of readily available, inexpensive diagnostic tests, dengue fever is generally not formally reported. In some endemic countries, such as Thailand, DHF/DSS may be systematically recorded and reports of such surveillance are periodically published. Both WHO and the Rockefeller Foundation have monitored formal country reports of DHF/DSS, and these reports formed the basis of our estimates for incidence of disease (Halstead 1992, 1993b, Okabe 1993, Pan American Health Organization 1993). The published scientific literature is generally limited to reports of specific outbreaks and most 222 Global Epidemiology of Infectious Diseases often does not offer a quantitative historical perspective of dengue or DHF/DSS. It is thus of limited usefulness in preparing burden of disease analyses. The problem is further complicated by the fact that dengue and DHF/DSS frequently occur as outbreaks or epidemics, often above a background of endemic transmission. Consequently, larger outbreaks may be recognized and reported; indeed, estimates of the number of cases may be inflated, since many febrile diseases may be reported then as dengue. Between outbreaks, uncomplicated dengue may be merged with all other causes of febrile illness, and thus there may be general underreporting of the disease. The most comprehensive cost-benefit analysis of dengue and DHF/DSS was prepared by Shepard & Halstead (1993), and their estimates are incorporated into the present model whenever possible. It is generally believed that millions of people suffer from dengue fever each year in Asia, Africa, the Pacific Islands, and the Americas (Pan American Health Organization 1993). It is virtually impossible, however, to document the global impact of dengue fever given the general absence of surveillance. For the purposes of this publication, we have focused on the more severe complications of dengue infection, DHF/DSS, and severe dengue fever cases. These conditions often require hospitalization and form the basis for the limited number of officially reported cases. Thus, our calculations are based on reported numbers of cases whenever possible. As these estimates reflect only the most clinically severe cases, they almost certainly underestimate the true burden of disease. We also used average incidence rates whenever possible. These were usually determined based on the number of cases reported over the past decade. This is justified since dengue and DHF are usually epidemic diseases, where the number of cases reported during an outbreak may well be 100 times greater than during inter-epidemic periods. Averaging over a 10-year period serves to compensate for these wide annual fluctuations. The trend, however, in nearly all endemic areas, and especially in the Americas, is towards more frequent and larger outbreaks. Unfortunately, our estimates fail to reflect this upward trend, and thus also lead to an underestimate of the true disease burden. International Classification of Diseases The specific categories related to dengue in the ninth and tenth revisions of the International Classification of Diseases (ICD) are listed in Box 8.1. Ninth revision Uncomplicated dengue fever, also listed as breakbone fever, is coded as 061 in the ninth revision of the International Classification of Diseases. Alternative codings could include 066: other arthropod-borne viral diseases, and 780.6: fever of unknown origin. Dengue haemorrhagic fever is coded as 065.4, with the following synonyms—Bangkok, Chikungunya, Dengue and dengue haemorrhagic fever Box 8.1 International Classification of Diseases (ICD) codes related to dengue and dengue haemorrhagic fever ICD 9 061 223 ICD 10 Uncomplicated dengue fever (also called breakbone fever) A90 Classical dengue fever A91 Dengue haemorrhagic fever A90–99 Anthropod-borne viral haemorrhagic fevers and viral haemorrhagic fevers mosquito-borne, Philippine, Singapore, South-East Asia and Thailand haemorrhagic fever—all sharing the same code, even though Chikungunya is both antigenically and taxonomically a completely different virus. Other classifications could result from coding based on clinical disease rather than viral etiology, and might include 785.5: shock, hypovolaemic without mention of trauma, and 448.9: haemorrhage, capillary, diseases of capillaries, other, and unspecified. Tenth revision Classical dengue fever is coded as A90 in the tenth revision of the International Classification of Diseases, and dengue haemorrhagic fever is classified as A91. The codes for arthropod-borne viral fevers and viral haemorrhagic fevers are A90–A99. Related codes that might include dengue or dengue haemorrhagic fever include B33: other viral diseases, not elsewhere classified, and B34: viral infection of unspecified site. Codes based on clinical disease rather than specific etiology might include: D65: disseminated intravascular coagulation (defibrination syndrome); D69: purpura and other haemorrhagic conditions, especially D69.9: haemorrhagic condition, unspecified; R04: haemorrhage from respiratory passages, especially R04.8: haemorrhage from other sites in respiratory passages; R21: rash and other nonspecific skin eruption; R50: fever of unknown origin; R57: shock, not elsewhere classified, especially R57.1: hypovolaemic shock, and R57.9: shock, unspecified; and H58: haemorrhage, not elsewhere classified. Estimated Burden of Dengue and Dengue Haemorrhagic Fever The most comprehensive data available for dengue and DHF/DSS originate in South East Asia, in the Other Asia and Islands region. We have based our estimates on data from this region, then tried to estimate the situation in other areas by comparison with Other Asia and Islands. 224 Global Epidemiology of Infectious Diseases Other Asia and Islands Population at risk. Countries within the region were reviewed to exclude those nations not thought to be at risk of severe dengue/DHF. An estimate of 86.4 per cent of the regional population was considered at risk of this disease. Countries or areas considered not at risk included: Bhutan, Democratic Peoples’ Republic of Korea, Mongolia, Nepal, Republic of Korea, and Hong Kong (total population 92 652 000). The age-specific and sex-specific population at risk was determined by multiplying the regional age-specific and sex-specific population estimates by 0.864. Overall incidence rates. The average number of reported cases in the decade prior to 1992 was calculated based on statistics summarized in recent publications (Halstead 1993a, 1993b) or those obtained by Gratz (unpublished data). Those averages were used to determine country-specific annual incidence rates per 1000 population based upon population estimates for those countries. These incidence rates were then used to estimate an overall incidence rate for the regional populations at risk of severe dengue/DHF. The overall incidence rate among the population at risk was estimated to be 0.3 per 1000. Because of general underreporting and differences in reporting criteria, this estimate is likely to be low. Furthermore, it is based on a 10year average, and in most parts of this region the incidence of dengue and DHF/DSS is increasing, another reason to suspect that the estimate is low. Nonetheless, there are no better estimates available for this or any other region, and we therefore have based our calculations on this estimate. Sex-specific incidence rates. Published reports indicate that female cases of severe dengue/DHF exceed male cases by a ratio of 1.2:1.0 (Shepard & Halstead 1993). Thus, we estimated the contribution of male cases at an incidence rate of 0.27 per 1000, and females at 0.33 per 1000. These rates were used to determine the total estimated number of cases. Age-specific incidence rates. Published reports indicate that approximately 95 per cent of severe dengue/DHF occurs in children under 15 years old (Shepard & Halstead 1993). Among these children, 75 per cent of cases are estimated to occur in the 5–14 year age group, and 25 per cent in the 0–4 year age group. We estimated the contribution from the remaining age groups as 3 per cent for 15–44 year, 1.5 per cent for 45–59 year, and 0.5 per cent for 60 years and over. We found no published reports to validate the estimates for adults. Estimated number of cases. Age-specific incidence rates were used to estimate the age-specific number of cases by multiplying the fraction by the total sex-specific estimated number of cases. Estimated number of deaths. Case fatality rates for dengue/DHF were Dengue and dengue haemorrhagic fever 225 reviewed for countries within the region where information was available. Case fatality rates clustered into two groups: those with rates that averaged 0.5 per cent, and those with rates that averaged 5.2 per cent. We included Malaysia, Singapore, Taiwan, Thailand and Viet Nam as “low threat areas”, and Bangladesh, Cambodia, Indonesia, Lao People’s Democratic Republic, Myanmar, Papua New Guinea, Philippines, and Sri Lanka, and all islands as “high threat areas”, even though reliable estimates of disease attributable to dengue/DHF do not exist for some of these countries. A total of 27.7 per cent of the at-risk population lived in “low-threat areas,” and 72.3 per cent in “high-threat areas.” We calculated the estimated number of age-specific and sex-specific cases in these areas by multiplying these factors by the total estimated number of cases. We did not have sufficient information to divide populations into risk areas in the other regions examined. The number of deaths in “high-threat areas” is estimated by multiplying the estimated number of cases in that area by the average case fatality rate, 5.2 per cent. The number of deaths in “low threat areas” was estimated by multiplying the estimated number of cases in that area by the average case fatality rate, 0.5 per cent. These values were summed to estimate the total number of deaths for the region, and are shown in Table 8.1. India Population at risk. Based on very limited information presented at a meeting on dengue held in 1994 in Pune, India, and cosponsored by the National Institute of Virology, the Rockefeller Foundation and the World Health Organization, we estimated that roughly 80 per cent of the population of Table 8.1 Sex and age (years) Estimates of the burden of dengue haemorrhagic fever, Other Asia and Islands Regional population (thousands) Population at risk (thousands) Estimated number of cases Estimated number of deaths 0–4 5–14 15–44 45–59 60+ 43 763 84 032 160 825 34 138 20 208 37 811 72 604 138 953 29 495 17 460 19 002 57 005 2 400 1 200 400 736 2 208 93 46 15 Total 342 966 296 323 80 007 3 099 0–4 5–14 15–44 45–59 60+ 41 988 80 217 159 611 35 090 22 662 36 278 69 307 137 904 30 318 19 580 22 994 68 983 2 905 1 452 484 891 2 672 112 56 19 Total 339 568 293 387 96 818 3 750 Males Females 226 Global Epidemiology of Infectious Diseases India is at risk of infection with dengue/DHF (World Health Organization 1994c). Overall incidence rates. We found no reliable estimates of dengue incidence in India; however, it was apparent from discussions during the meeting that the problem may approach or exceed the magnitude of that seen in the Other Asia and Islands Region. Consequently, we used the same estimate for overall incidence as applied for that region, 0.3 per 1000. Sex-specific incidence rates. We used the same sex-specific incidence rates as used for the Other Asia and Islands Region. Age-specific incidence rates. We used the same age-specific incidence rates as used for the Other Asia and Islands Region. Estimated number of cases. Age-specific incidence rates were used to estimate the age-specific number of cases by multiplying the fraction by the total sex-specific estimated number of cases. Estimated number of deaths. A presentation made at the meeting on dengue indicated that there is a wide spread need for educational programmes for medical staff on the proper management of DHF/DSS patients (World Health Organization 1994c). Existing mortality rates attributable to DHF may be approximately the same as in areas considered as high threat in the Other Asia and Islands Region. We therefore used a case fatality rate of 5.2 per cent, realizing that additional information is clearly needed to form the basis for a more accurate estimate. The number of deaths was estimated by multiplying the estimated number of cases by the case fatality rate, 5.2 per cent. Estimates for the disease burden attributable to DHF in India are shown in Table 8.2. China Population at risk. Only the southern provinces of China are at risk for dengue/DHF; dengue does not occur in areas where hard freezes occur in winter. Consequently, we estimated that 10 per cent of the total population of China is at risk for dengue/DHF. Overall incidence rates. The incidence of dengue/DHF in southern China is less than in India or the Other Asia and Islands Regions, but nonetheless thought to be substantial. We found no quantitative estimates of dengue/ DHF incidence rates in China, so we estimated an incidence rate of 0.2 per 1000, somewhat less than that used for the Other Asia and Islands Region. One might suspect that the incidence in southern China would approximate that seen in adjacent areas of the Lao People’s Democratic Republic and Viet Nam. Dengue and dengue haemorrhagic fever Table 8.2 Sex and age (years) 227 Estimates of the burden of dengue haemorrhagic fever, India Regional population (thousands) Population at risk (thousands) Estimated number of cases Estimated number of deaths 0–4 5–14 15–44 45–59 60+ 59 789 101 752 200 525 47 567 29 768 47 831 81 402 160 420 38 054 23 814 22 541 67 624 2 847 1 424 475 1 172 3 516 148 74 25 Total 439 401 351 521 94 911 4 935 0–4 5–14 15-44 45–59 60+ 56 679 95 263 183 242 46 005 28 924 45 343 76 210 146 594 36 804 23 139 25 714 77 142 3 248 1 624 541 1 337 4 011 169 84 28 Total 410 113 328 090 108 270 5 630 Males Females Sex-specific incidence rates. We used the same ratio of sex-specific incidence rates as used for the Other Asia and Islands Region; thus the incidence rate for males was 0.18 per 1000 and for females 0.22 per 1000. Age-specific incidence rates. We used the same age-specific incidence rates as for the Other Asia and Islands Region. Estimated number of cases. Age-specific incidence rates were used to estimate the age-specific number of cases by multiplying the fraction by the total sex-specific estimated number of cases. Estimated number of deaths. We used an intermediate case fatality rate of 3.7 per cent for the entire population at risk. The number of deaths was estimated by multiplying the estimated number of cases by this rate. These estimates are presented in Table 8.3. Sub-Saharan Africa Population at risk. There is very little quantitative information regarding the actual population at risk of severe dengue/DHF in Africa. Periodic reports have been published that clearly indicate that dengue viruses are present in many different parts of Africa, and recent outbreaks of dengue have occurred in Comoros and among United Nations personnel deployed to Somalia (Halstead 1992, World Health Organization 1993). Few serological surveys have been attempted and interpretation of those done is hampered by the presence of other flaviviruses, which cross-react with dengue in many serological tests. Dengue viruses are present, particularly around urban centres where Aedes aegypti may be abundant. Based on very 228 Table 8.3 Sex and age (years) Global Epidemiology of Infectious Diseases Estimates of the burden of dengue haemorrhagic fever, China Regional population (thousands) Population at risk (thousands) Estimated number of cases Estimated number of deaths 0–4 5–14 15–44 45–59 60+ 60 243 96 997 306 305 72 674 48 980 6 024 9 700 30 631 7 267 4 898 2 502 7 505 316 158 53 93 278 12 6 2 Total 585 199 58 520 10 534 390 0–4 5–14 15-44 45–59 60+ 57 946 90 401 284 078 64 403 51 666 5 795 9 040 28 408 6 440 5 167 2 866 8 598 362 181 60 106 318 13 7 2 Total 548 494 54 849 12 067 446 Males Females crude estimates, we considered that 20 per cent of the residents of Africa were at risk of dengue/DHF, even though we could locate no confirmed record of DHF in sub-Saharan Africa. Overall incidence rates. The limited information currently available indicates that dengue/DHF is less abundant than in India or the Other Asia and Islands Region. We estimated the incidence to be 0.1 per 1000, with little factual basis for this figure. Sex-specific incidence rates. We used the same ratio of sex-specific incidence rates as for the Other Asia and Islands Region, thus we used an incidence rate of 0.09 per 1000 males and 0.11 per 1000 females. Age-specific incidence rates. We used the same age-specific incidence rates as for the Other Asia and Islands Region. Estimated number of cases. Age-specific incidence rates were used to estimate the age-specific number of cases by multiplying the fraction by the total sex-specific estimated number of cases. Estimated number of deaths. We used a high case fatality rate of 5.2 per cent, based on the fact that the incidence rate is relatively low, thus giving clinicians little experience with the disease, and realizing that virtually no special training has been provided to clinicians in Africa for management of DHF/DSS. We used only a single case fatality rate for all of the SubSaharan Africa Region. The estimated number of deaths was calculated by Dengue and dengue haemorrhagic fever 229 multiplying the case fatality rate by the estimated number of cases. These estimates are presented in Table 8.4. Latin America and the Caribbean Population at risk. Dengue and DHF/DSS are rapidly expanding throughout Latin America and the Caribbean, closely following the reinfestation of most urban centres and many rural areas by the vector mosquito of dengue, Aedes aegypti. As a result, the general trend has been to increasing incidence of disease, especially over the past decade. Limited quantitative information is available to support specific country estimates, although the Pan American Health Organization has attempted to monitor dengue and DHF for more than a decade (Pan American Health Organization 1993). The most dramatic outbreak of dengue and DHF/DSS occurred in Cuba in 1981, primarily between the months of June and October, when 344 203 cases of dengue were reported, of which 116 151 were hospitalized (Kouri, Guzman & Bravo 1986). Of the hospitalized cases, 9203 were considered “serious,” and an additional 1109 were considered “very serious.” A total of 158 deaths were recorded (the case fatality rate was 1.5 per cent of serious and very serious cases, or 0.14 per cent of hospitalized cases). While Cuba has subsequently virtually eliminated Aedes aegypti mosquito populations, the dramatic potential impact of epidemic dengue or DHF on a nation’s health care system is nonetheless evident from these figures. Indeed, at the peak of the Cuban epidemic, approximately 10 000 cases were hospitalized per week, enough to swamp even the most sophisticated medical system. Considering the expanding dengue incidence throughout most of Latin America and the Caribbean in both urban and rural settings, Table 8.4 Sex and age (years) Estimates of the burden of dengue haemorrhagic fever, Sub-Saharan Africa Regional population (thousands) Population at risk (thousands) Estimated number of cases Estimated number of deaths 0–4 5–14 15–44 45–59 60+ 47 484 70 258 103 764 20 308 10 508 9 497 14 052 20 753 4 062 2 102 1 079 3 236 136 68 23 56 168 7 4 1 Total 252 322 50 464 4 542 236 0–4 5–14 15-44 45–59 60+ 47 030 69 818 106 257 22 117 12 730 9 406 13 964 21 251 4 423 2 546 1 035 1 536 2 338 487 280 54 80 122 25 15 Total 257 952 51 590 5 675 295 Males Females 230 Global Epidemiology of Infectious Diseases we have estimated that 60 per cent of the population of this region is at risk of severe dengue/DHF. Overall incidence rates. While the incidence of dengue is less in Latin America and the Caribbean than in Asia, it is nonetheless significant and increasing. From 1988 to 1992, an average of 102 162 cases of dengue fever were reported in the region. During that same period, an average of 2100 DHF cases were reported annually, with a mean of 31 deaths attributed to the disease (Pan American Health Organization 1993), as shown in Table 8.5. Based on these reported rates, we have estimated the annual incidence of DHF to be 0.006 per 1000. Sex-specific incidence rates. We used the same ratio of sex-specific incidence rates as for the Other Asia and Islands Region; thus the rate for males was 0.0056 per 1000 and for females 0.0063 per 1000. Age-specific incidence rates. We used the same age-specific incidence rates as for the Other Asia and Islands Region. Estimated number of cases. Although dengue fever is rapidly increasing in most countries of the region, DHF remains relatively uncommon, and thus we used the same annual incidence rate for all populations at risk. Estimated number of deaths. We used a case fatality rate of 1.5 per cent based on the average number of deaths reported from 1988 to 1992. A summary of estimates for the population at risk, number of cases and deaths is found in Table 8.6. Middle Eastern Crescent, Formerly Socialist Economies of Europe, and Established Market Economies In these regions, dengue and DHF/DSS are generally limited to tourists travelling to and from endemic areas in other regions, although transmisTable 8.5 Reported cases of dengue fever, dengue haemorrhagic fever, and deaths attributable to dengue haemorrhagic fever in Latin America and the Caribbean, 1988 to 1992 Year Dengue fever Dengue haemorrhagic fever Deaths 1988 1989 1990 1991 1992 47 783 89 138 116 389 155 543 101 958 86 2 682 3 646 1 760 2 319 6 33 62 31 23 Average 102 162 2 099 31 Source: Pan American Health Organization (1993). Dengue and dengue haemorrhagic fever Table 8.6 Sex and age (years) 231 Estimates of the burden of dengue haemorrhagic fever, Latin America and the Caribbean Regional population (thousands) Population at risk (thousands) Estimated number of cases Estimated number of deaths Males 0–4 5–14 15–44 45–59 60+ 28 721 52 124 104 286 22 249 14 231 22 977 41 699 83 429 17 799 11 385 236 707 30 15 5 2 7 0 0 0 Total 221 611 177 289 993 10 0–4 5–14 15-44 45–59 60+ 27 676 50 744 104 085 23 355 16 824 22 141 40 595 83 268 18 684 13 459 267 800 34 17 6 4 12 1 0 0 Total 222 684 178 147 1 122 17 Females sion may exist in a few population centres. For example, in the Middle Eastern Crescent Region, there is growing evidence that active dengue transmission probably occurs in some locations. Nonetheless, the number of cases in these regions is relatively small, and no quantitative information exists upon which incidence estimates could be based. Validity of incidence estimates Clearly the estimate of incidence rates is the single most critical value in calculating the global burden of disease. Unfortunately, the data from which this estimate must be drawn, at least for dengue/DHF, are very weak. The incidence of dengue/DHF depends upon several factors, including the abundance of potential mosquito vectors, especially Aedes aegypti, the number of susceptible humans in the population at risk, and the presence of dengue viruses circulating in or introduced into those populations. Other factors, especially environmental and sociological, also influence estimates of the incidence of dengue/DHF. The fact that distinct viruses may cause dengue, the confusion among clinicians about what actually distinguishes dengue from DHF, and the limitations in availability of laboratory confirmation, all contribute to the ambiguity of incidence estimates. Further, dengue and DHF tend to occur in epidemics, so that tremendous numbers of cases may occur during one period, followed by subsequent years when relatively few cases may be recorded. Finally, as with all diseases, the estimates are only as good as the surveillance and reporting systems from which the estimates are drawn. In general, for dengue and DHF, these are not well developed. Between outbreaks, there is likely to be underreporting, dengue 232 Global Epidemiology of Infectious Diseases being grouped with undifferentiated febrile illness or haemorrhagic diseases of undetermined origin. During outbreaks, there may be overreporting, since all febrile illnesses may be reported as dengue, and any febrile disease with bleeding manifestations as DHF. To overcome these potential problems, we have concentrated our efforts at estimation on incidence in the Other Asia and Islands Region, where dengue and DHF are best known clinically and historically, and where we feel that the reporting systems are most reliable. To overcome the large annual variations in incidence of dengue and DHF, we have attempted to average the incidence rates for the most recent 10-year period for which complete information is available. The actual number of reported DHF cases by selected countries are presented in Table 8.7. We have attempted to calculate the overall regional incidence rates by taking into account the specific rates for those countries for which information is available, and estimating for other countries in the region where DHF is known to occur, but no published information on incidence rates is available. Whenever possible, we used these same techniques in the other areas considered; however, the data are generally much less reliable for the other regions. Consequently, we used the rate for the Other Asia and Islands Region as a reference point, then determined whether other regions had greater, equal, or lower incidence rates, based on informal reporting and published information from recent outbreaks. The gross estimates are summarized in Table 8.8. We suspect that the final product of this exercise grossly underestimates the true burden of DHF. The question that remains to be answered is the order of magnitude that this underestimation represents. Attempts are in progress to improve dengue diagnosis and surveillance in hopes of generating more accurate information on the true global burden of this increasingly important disease. In reviewing the data available to estimate incidence, a striking trend is consistently seen. Dengue and DHF are increasing in virtually every location where they are found, and are rapidly appearing in new areas, following infestations by the vector mosquito, Aedes aegypti. This rapid increase is most notable in the Other Asia and Islands Region, where DHF has become endemic and a constant threat in many countries, and in Latin America and the Caribbean, where dengue epidemics are increasingly common and DHF is seen with increasing frequency. Given that the estimates for incidence were based on 10-year averages, they are most likely underestimates of the true incidence of disease. The global burden of this disease is likely to be much greater in the coming years. Other considerations in estimating disease burden The focus of this discussion has been on severe dengue and DHF, conditions likely to require hospitalization and thus be captured by a national surveillance system. Uncomplicated dengue fever has not been considered although it is acknowledged that there are many more dengue fever patients Cases 5 909 4 665 13 875 12 710 13 875 12 710 22 765 47 573 10 362 13 043 21 120 17 620 196 2270 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 Total 6 994 231 255 491 382 460 600 1 039 1 527 464 458 578 509 Deaths Indonesia 16 152 486 3 052 790 702 384 1 403 2 025 1 448 2 401 2 071 741 649 Cases 211 14 36 10 5 11 9 8 2 14 38 39 25 Deaths Malaysia 38 822 1 524 1 706 2 856 2 323 2 666 2 191 7 292 1 181 1 196 6 318 8 055 1 514 Cases 1 372 90 49 83 39 134 111 222 65 52 182 305 40 Deaths Myanmar 13 979 123 305 1 684 2 545 2 096 687 859 2 922 305 588 1 865 n/a Cases 654 8 31 130 89 210 30 27 68 14 27 20 n/a Deaths Philippines 9 535 133 216 205 86 126 354 436 245 944 1 733 2 179 2 878 Cases 180 0 0 2 0 2 0 0 0 0 4 6 4 Deaths Singapore 681 013 25 641 22 250 30 022 69 597 80 076 29 030 170 630 26 926 69 204 113 855 43 782 n/a Cases 3 689 194 159 231 451 542 206 896 189 280 422 119 n/a Deaths Thailand Dengue haemorrhagic fever reported, selected countries, Other Asia and Islands, 1982–1992 Year Table 8.7 1 009 640 35 323 39 806 149 519 30 498 45 107 46 266 354 517 85 160 40 205 37 569 94 630 51 040 Cases 7 455 408 361 1 798 368 399 511 1 566 826 289 255 403 271 Deaths Viet Nam Dengue and dengue haemorrhagic fever 233 234 Table 8.8 Global Epidemiology of Infectious Diseases Crude estimate for the global burden of disease attributable to dengue haemorrhagic fever, 1994 Region Cases Deaths Other Asia and Islands India China Sub-Saharan Africa Latin America and Caribbean 180 000 205 000 23 000 11 000 3 000 7 000 11 000 1 000 600 100 Total 422 000 19 700 for each case of DHF. Consequently, the nearly half a million cases of DHF estimated here are likely to reflect 5–50 million or more dengue infections annually. Certainly in epidemic years, this could well be an underestimation. The cost of uncomplicated dengue is perhaps most noticeable among the adult population, where the disease is more often clinically apparent, frequently leading to acute illness of about a week’s duration, and occasionally with prolonged convalescence. Those at greatest risk are residents of areas recently infested with vector mosquitoes and exposed for the first time as adults to dengue, as is happening now in many parts of Latin America, and travellers (including deployed military personnel, business people, tourists and others). No attempt has been made to quantify either the economic or health impact of dengue on these groups, but it is undoubtedly substantial and growing. Risk factors Risk factors for dengue and DHF are primarily those associated with the potential to be fed upon by infected mosquitoes. Because the primary vector feeds mostly during daylight hours, and is well adapted to the urban environment, the socioeconomic conditions that favour the spread of other vector-borne disease, such as malaria, are not as important with dengue. Certainly lower-income neighbourhoods may be at increased risk of infection, especially if litter and refuse that may retain sufficient water to support mosquito larval development are abundant; however, this mosquito species breeds equally efficiently in flower pots and other ornamental containers that hold water, and may be abundant in any area, including middle-income or upper-income neighbourhoods, or in public places such as schools and parks. This, coupled with its day-biting activity, means that people may easily be exposed to dengue-infected mosquitoes during their daily travels, even though their personal dwellings have screens and they live in a “nice” neighbourhood. A second major risk factor is prior infection with a dengue virus, since heterologous anti-dengue antibodies may enhance dengue infection and predispose an individual to much greater risk of DHF on subsequent infection. Similarly, maternally-acquired antibodies, as titres wane and are no longer able fully to inactivate dengue virus, place infants at increased risk Dengue and dengue haemorrhagic fever 235 of antibody-enhanced dengue infection leading to increased risk of DHF. Other risk factors include the human population density, and thus the availability of susceptible individuals to sustain transmission, residence in tropical environments, where vector species survive best, and residence in an area where infected travellers may introduce the virus and precipitate epidemic transmission. The risk of dying from dengue or DHF depends to a large extent on the skills of local attending physicians and nurses. Properly trained and experienced health care professionals can expertly manage critically ill patients and reduce mortality rates to below 1 per cent; however, those not familiar with the disease may find mortality rates of 5 per cent or greater among their seriously ill patients. Other diseases Two other vector-borne diseases frequently occur under conditions similar to those which support dengue virus transmission: yellow fever and Chikungunya fever. Yellow fever is well known as a historical scourge of the tropics; thousand of individuals died, especially in the temperate and tropical zones of the Americas, often seriously limiting national development projects, as was the case with the Panama Canal. What is often overlooked today is the fact that the same conditions that originally supported urban yellow fever transmission in the Americas are once again developing, with widespread and abundant populations of vector mosquitoes, Aedes aegypti, and large, susceptible human populations. All that is needed to initiate urban transmission is the introduction of infected humans into a situation where they may be fed upon by vector mosquitoes, and an outbreak could ensue. Currently, active yellow fever transmission is limited to endemic areas of sub-Saharan Africa and northern South America; however, with the frequency of air travel to and from virtually all parts of the world, it is not inconceivable that an individual could be infected in a remote endemic area, and shortly thereafter be fed upon by potential vector mosquitoes in a major urban centre far removed from the site of original infection. Chikungunya is not as well known as yellow fever, but it too is transmitted by Aedes aegypti, and may cause massive outbreaks in urban settings. This virus is not known to cause fatal human infection, but it does cause a serious febrile illness similar to dengue fever, which may incapacitate infected individuals for about a week. Large epidemics have been reported in India, Indonesia and the Philippines, and the virus is known to be present in much of Africa. Cost-effectiveness of interventions At present, countries where dengue and DHF are endemic spend millions of dollars on vector control. This includes direct health care costs for treatment and indirect costs resulting from lost tourism and lost work. 236 Global Epidemiology of Infectious Diseases The latter arises not solely in relation to infected individuals; since those infected are often children, parents may have to stop working while they care for their sick child. The average hospital stay for DHF patients is 5 to 10 days. Shepard & Halstead (1993) have summarized recent cost estimates for dengue epidemics and found the total for direct health care interventions, associated vector control activities, and indirect costs to be from several million to over US$100 million per outbreak. Unfortunately, the frequency of such outbreaks is increasing as new areas are infested with Aedes aegypti. Future research priorities Substantial research efforts are required to manage the problem of dengue/ DHF. An immediate need exists for expanded medical training to teach physicians and nurses the proper management of DHF patients, especially in areas where the disease is just emerging and clinicians have little experience in managing these seriously ill patients. Significant reductions in the mortality rates in endemic areas can be realized with minimal costs through such efforts, and clearly this should be among the highest priority activities in dengue control efforts. Likewise, support is required in the clinical laboratory to ensure that dengue/DHF is properly and promptly diagnosed. Again, this can be accomplished with relatively little investment since common laboratory techniques are used, and all that may be required is provision of the appropriate diagnostic reagents. Accurate diagnosis should then feed into a well-organized surveillance and reporting system that could allow health managers to monitor disease trends and initiate outbreak control efforts at the first indication of a pending epidemic. Several organizations are attempting to develop a vaccine for dengue/ DHF. A tetravalent vaccine would be especially beneficial. Since humans represent the primary amplifying host of these viruses, a fully protective vaccine that was widely administered could interrupt transmission and significantly reduce the risk of infection. This is an especially challenging task, however, since four distinct viruses may cause the disease, and the confounding factor of immune-enhancement where there is partial protection, as would be elicited in a vaccine against only one, two or three of the four dengue viruses, may in fact place individuals at greater risk of DHF than unvaccinated persons. A live, attenuated, tetravalent vaccine for dengue viruses has entered into Phase 2 clinical trials, and appears to be safe and immunogenic. At present, the only avenue for effective prevention of dengue/DHF is through vector control. Unfortunately, sustained vector control is costly, and as a consequence, frequently limited in success. Research is needed into better ways of controlling Aedes aegypti populations. These must not only be economically feasible, but also sustainable. Further, they need to take into account the movement of large segments of the populations from rural areas to urban centres in many affected countries. Current control efforts Dengue and dengue haemorrhagic fever 237 are often directed towards routine pest management, and may include community involvement coupled with emergency insecticide applications when outbreaks are recognized. These efforts have met with marginal success, and it is clear that there is room for improvement. Historically, Aedes aegypti was controlled by highly regimented mosquito control programmes that involved both source reduction and effective use of insecticides (Soper et al. 1943). Such programmes are, however, extremely expensive and difficult to maintain. Today, some countries have succeeded in controlling vector mosquito populations using these techniques (for example, Cuba after the 1981 outbreak), but in most cases countries have relied on insecticide applications after outbreaks are recognized, generally with much less success. It is clear that innovative measures are needed, which will be identified only through multidisciplinary research efforts that consider both the sociological aspects of the local conditions leading to vector breeding, and of the biology of the vector mosquito itself. Equally challenging are the characteristics of the viruses that cause dengue. Wide variations in the pathogenicity of dengue viruses have been observed. Some strains of the same serotype may cause outbreaks with little or no serious illness, while others may elicit significant numbers of severe or fatal DHF cases. The molecular basis for this variation must be determined. Fortunately, the complete nucleotide sequence is now available for strains of all four dengue serotypes, and infectious clones have been produced for some serotypes. These tools should help in the future definition of virulence factors of dengue viruses. Conclusions Dengue, and especially DHF/DSS, are significant global diseases that are increasing in incidence and economic importance. Four distinct viruses may cause classic dengue fever, and while infections with one serotype leads to lifelong homologous protection, there is no lasting cross-protection from other dengue serotypes; indeed, subsequent dengue infections are associated with increased risk of DHF/DSS. All four dengue viruses are transmitted primarily by the urban mosquito vector, Aedes aegypti, a species that is well adapted to life in close association with humans, and one whose distribution and abundance is rapidly expanding throughout tropical and subtropical cities, suburbs and villages of the world. Vector control is expensive and inefficient, and as a result, conditions are prime for epidemic dengue virus transmission in these areas. The result is that today much of the world’s population is at risk of dengue, either as an immediate threat in areas where the primary vector is abundant and dengue viruses are endemic, or to travellers to these areas. We have attempted to estimate the global burden of disease attributable to severe dengue and DHF/DSS, based on reported numbers of cases. The greatest number of cases and deaths are thought to occur in India and the Other Asia and Islands Region; however, the fastest growing 238 Global Epidemiology of Infectious Diseases region in terms of dengue fever is the Latin America and the Caribbean Region. Little information is available to document the burden of disease for the Sub-Saharan Africa Region. These analyses conclude that nearly half a million cases of dengue sufficiently severe to reach formal reporting occur annually. Of these, approximately 20 000, or about 4 per cent, end in death. These values clearly underestimate the true impact of severe dengue because: they are based on reported cases only, and this information is fragmen- tary and incomplete; they are based on annual estimates averaged over a decade and thus minimize the impact of the increasing incidence of DHF/DSS in some regions; in some regions, little information is available on which to base any estimate. References Akiyama J (1993) Review of multicentre project to control DHF vectors by community participation in WHO, SEARO. Tropical Medicine, 35:277–283. Bosio CF et al. (1992) Variation in the efficiency of vertical transmission of dengue1 virus by strains of Aedes albopictus (Diptera: Culicadae). Journal of Medical Entomology, 29:985–989. Clark DH, Casals J (1965) Arboviruses, Group B. In: Horsfall FL, Tamm I, eds. Viral and rickettsial infections of man. Philadelphia, Lippincott, pp. 606–658. Cordellier R et al. (1983) Circulation silvatique du virus Dengue 2 en 1980, dans les savannes sub-soudaniennes du Côte d’Ivoire [Silvian circulation of the Dengue 2 virus in 1980 in the savannahs of Côte d'Ivoire]. Cahier ORSTOM Série Entomologie Médicale Et Parasitologie, 21:165–172. Craven RB (1991) Flaviviruses. In: Belshe RB ed. Textbook of human virology, 2nd ed., St. Louis, MO, Mosby-Year Book, Inc. pp. 633–662. de Souza M, Freier JE (1991) Vertical transmission of dengue 1 virus by Haemagogus equinus mosquitoes. Journal of the American Mosquito Control Association, 7: 118–20. Ehrenkranz NJ et al. (1971) Pandemic dengue in Caribbean countries and the southern United States: past, present and potential problems. New England Journal of Medicine, 285:1460–1469. Fagbami AH, Monath TP, Fabiyi A (1977) Dengue virus infections in Nigeria: a survey for antibodies in monkeys and humans. Transactions of the Royal Society of Tropical Medicine and Hygiene, 71:60–65. Fauran P, Laille M, Moreau JP (1990) Etude sur la transmission verticale des virus de la dengue dans le Pacifique Sud [Study on the vertical transmission of the dengue virus in the South Pacific]. Bulletin de la Société de Pathologie Exotique, 83:311–316. Dengue and dengue haemorrhagic fever 239 Freier JE, Rosen L (1987) Vertical transmission of dengue viruses by mosquitoes of the Aedes scutellaris group. American Journal of Tropical Medicine and Hygiene, 37:640–647. Freier JE, Rosen L (1988) Vertical transmission of dengue viruses by Aedes mediovittatus. American Journal of Tropical Medicine and Hygiene, 39:218–222. Gratz N (1993a) Lessons of Aedes aegypti control in Thailand. Medical and Veterinary Entomology, 7:1–10. Gratz N (1993b) What must we do to effectively control Aedes aegypti? Tropical Medicine, 35: 243–251. Gubler DJ (1993) Emergent and resurgent arboviral diseases as Public Health Problems. Concepts in Virology, 24:257–273. Halstead SB (1970) Observations related to the pathogenesis of dengue hemorrhagic fever: VI hypotheses and discussion. Yale Journal of Biology and Medicine, 42: 350–362. Halstead SB (1980) Immunological parameters of togavirus disease syndromes. In: Schlesinger RW, ed. The Togaviruses. New York, Academic Press, pp.107–174. Halstead SB (1992) The XXth century dengue pandemic: need for surveillance and research. World Health Statistics Quarterly, 45:292–298. Halstead SB (1993a) Dengue virology: fifty years of problems and progress. Concepts in Virology, 25:275–281. Halstead SB (1993b) Global epidemiology of dengue: health systems in disarray. Tropical Medicine, 35:137–146. Halstead SB, Nimmannitya S, Margiotta MR (1969) Dengue and chikungunya virus infection in man in Thailand, 1962–1964: II observations on disease in outpatients. American Journal of Tropical Medicine and Hygiene, 18:972–983. Hammon WMcD et al. (1960) New hemorrhagic fevers of children in the Philippines and Thailand. Transactions of the Association of American Physicians, 73:140–155. Hammon WMcD, Rudnick A, Sather GE (1960) Viruses associated with epidemic hemorrhagic fevers of the Philippines and Thailand. Science, 131:1102–1103. Hare GE (1898) The 1897 epidemic of dengue in North Queensland. Australasian Medical Gazette, 17:98–107. Henchal EA et al. (1986) Identification of an antigenic and genetic variant of dengue 4 virus from the Caribbean. American Journal of Tropical Medicine and Hygiene, 35:393–400. Hull B et al. (1984) Natural transovarial transmission of dengue 4 virus in Aedes aegypti in Trinidad. American Journal of Tropical Medicine and Hygiene, 33: 1248–1250. Itoh T (1993) Control of DF/DHF vector, Aedes Mosquito, with insecticides. Tropical Medicine, 35:259–267. Kerschner JH et al. (1986) Genetic and epidemiologic studies of dengue type 2 viruses by hybridization using synthetic deoxyoligonucleotides as probes. Journal of General Virology, 67:2645–2661. 240 Global Epidemiology of Infectious Diseases Khin MM, Than KA (1983) Transovarial transmission of dengue 2 virus by Aedes aegypti in nature. American Journal of Tropical Medicine and Hygiene, 32: 590–594. Kliks SC et al. (1989) Antibody-dependent enhancement of dengue virus growth in human monocytes as a risk factor for dengue hemorrhagic fever. American Journal of Tropical Medicine and Hygiene, 40:444–451. Kojima S (1993) A rapporteur’s summary: vector control. Tropical Medicine, 35: 331–334. Kouri G, Guzman MG, Bravo J (1986) Haemorrhagic dengue in Cuba: history of an epidemic. Bulletin of the Pan American Health Organization, 20:24–30. Kuberski T et al. (1977) Clinical and laboratory observations on patients with primary and secondary dengue type 1 infections with hemorrhagic manifestations in Fiji. American Journal of Tropical Medicine and Hygiene, 26:775–783. Lam SK (1993) A rapporteur’s summary: health system. Tropical Medicine, 35:335. Lopez-Correa RH et al. (1979) The 1977 dengue epidemic in Puerto Rico: epidemiologic and clinical observations. In: Dengue in the Caribbean, 1977. Washington, DC, Pan American Health Organization (Scientific Publication, No. 375, pp. 60–67). Mitchell CJ, Miller BR (1990) Vertical transmission of dengue viruses by strains of Aedes albopictus recently introduced into Brazil. Journal of the American Mosquito Control Association, 6:251–253. Monath TP et al. (1974) Arbovirus studies in Nupeko Forest, a possible natural focus of yellow fever virus in Nigeria: I. – description of the area and serological survey of humans and other vertebrate hosts. Transactions of the Royal Society of Tropical Medicine and Hygiene, 68:30–38. Okabe N (1993) Situation on dengue fever and dengue haemorrhagic fever in the western Pacific region. Tropical Medicine, 35:147–160. Pan American Health Organization (1982) Dengue laboratory diagnostic procedures, Washington, DC, Pan American Health Organization. Pan American Health Organization (1993) Dengue in the Americas: an update. Epidemiological Bulletin, 14:1–3. Pan American Health Organization (1994) Dengue and dengue haemorrhagic fever in the Americas: guidelines for prevention and control. Washington, DC, Pan American Health Organization. Quintos FN et al. (1954) Hemorrhagic fever observed among children in the Philippines. Philippine Journal of Pediatrics, 3:1–9. Ramirez-Ronda CH et al. (1979) Dengue hemorrhagic shock in the western hemisphere. Tropical and Geographical Medicine, 31:127–131. Repik PM et al. (1983) RNA finger-printing as a method for distinguishing dengue 1 virus strains. American Journal of Tropical Medicine and Hygiene, 32: 577–589. Roche JC et al. (1983) Isolement de 96 souches de virus dengue 2 à partir de moustiques capturés en Côte d’Ivoire et Haute-Volta [Isolation of 96 samples Dengue and dengue haemorrhagic fever 241 of the dengue 2 virus from mosquitoes captured in Côte d’Ivoire and HauteVolta]. Annales de Virologie, 134E:233–238. Rosen L (1987) Mechanism of vertical transmission of the dengue virus in mosquitoes. Comptes Rendus de l’Académie des Sciences Série III, Sciences de la Vie, 304:347–50. Rosen L (1988) Further observations on the mechanism of vertical transmission of flaviviruses by Aedes mosquitoes. American Journal of Tropical Medicine and Hygiene, 39:123–126. Rosen L et al. (1983) Transovarial transmission of dengue viruses by mosquitoes: Aedes albopictu and Aedes aegypti. American Journal of Tropical Medicine and Hygiene, 32:1108–1119. Rudnick A (1965) Studies of the ecology of dengue in Malyasia: a preliminary report. Journal of Medical Entomology, 2:203–208. Rudnick A (1984) The ecology of the dengue virus complex in Peninsular Malaysia. In: Pang T and Pathmanathan R, eds. Proceedings of the International Conference on Dengue/DHF, Kuala Lumpur, University of Malyasia Press, pp. 7. Rush B (1789) An account of the bilious remitting fever as it appeared in Philadelphia in the summer and autumn of the year 1780. In: Medical inquiries and observations, Philadelphia, pp.89–100. Russell PK, McCown JM (1972) Comparison of dengue-2 and dengue-3 virus strains by neutralization tests and identification of subtype of dengue-3. American Journal of Tropical Medicine and Hygiene, 21:97–99. Russell PK, Brandt WE, Dalrymple JM (1980) Chemical and antigenic structure of flaviviruses. In: Schlesinger RW, ed. The Togaviruses, New York, Academic Press, pp. 503–530. San Juan Laboratories (1979) 1979 annual report. San Juan, Puerto Rico, Centers for Disease Control. Serufo JC et al. (1993) Isolation of dengue virus type 1 from larvae of Aedes albopictus in Campos Altos City, State of Minas Gerais, Brazil. Memorias do Instituto Oswaldo Cruz, 88:503-504. Shepard DS, Halstead SB (1993) Dengue (with notes on yellow fever and Japanese encephalitis). In: Jamison DT et al., eds. Disease control priorities in developing countries, New York, Oxford University Press, 303–320. Shroyer DA (1990) Vertical maintenance of dengue-1 virus in sequential generations of Aedes albopictus. Journal of the American Mosquito Control Association, 6:312–314. Soper FL et al. (1943) The organization of permanent nation-wide anti-Aedes aegypti measures in Brazil, New York, Rockefeller Foundation. Tsai CJ et al. (1991) Upper gastrointestinal bleeding in dengue fever. American Journal of Gastroenterology, 86:33–35. Watts DM et al. (1987) Effect of temperature on the vector efficiency of Aedes aegypti for dengue 2 virus. American Journal of Tropical Medicine and Hygiene, 36:143–152. 242 Global Epidemiology of Infectious Diseases World Health Organization (1986) Dengue haemorrhagic fever: diagnosis treatment and control, Geneva, World Health Organization. World Health Organization (1993) Dengue epidemic. Weekly Epidemiological Record, 68:191. World Health Organization (1994a) Dengue and dengue haemorrhagic fever. Weekly Epidemiological Record, 69:265–266. World Health Organization (1994b) Dengue and dengue haemorrhagic fever. Weekly Epidemiological Record, 69:237–239. World Health Organization (1994c) International conference on dengue haemorrhagic fever. Geneva, World Health Organization (WHO/CDS/MIM/94.1). Chapter 9 Intestinal nematode infections DAP Bundy, MS Chan, GF Medley, D Jamison, L Savioli Introduction Effective control of the major intestinal nematode infections of humans involves relatively low-cost interventions (Savioli, Bundy & Tomkins 1992). Informed decisions about the need for investment in control programmes, however, also require useable estimates of the scale of morbidity (Guyatt & Evans 1992). Direct estimates of morbidity would be preferred, but the currently available techniques suffer from major limitations, examples of which follow. Active case detection can be useful at the local level (Pawlowski & Davis 1989), but demands a level of resources that makes its widespread use impractical. Passive case detection, which has more appropriate resource demands (Guyatt & Evans 1992), is unreliable for geohelminthiases because the symptoms are non-specific; in one prospective study of trichuris colitis, only 2 per cent of cases observed in the community had self-presented to the local health services (Cooper, Bundy & Henry et al. 1986). Furthermore, there is increasing evidence that the most common and potentially important consequences of infection are insidious effects on nutritional status (Nesheim 1989, Tomkins & Watson 1989) and on physical and intellectual development (Stephenson 1987, Cooper et al. 1990, Nokes et al. 1992a, 1992b). Such effects are unlikely to result in self-presentation to passive case detection clinics, and will be grossly underestimated by survey procedures based on the active detection of clinical signs. This is especially so, as the infections are ubiquitous; both passive and active detection are usually based on deviations from local norms, which may be greatly influenced by the infections. There are no direct estimates of the community morbidity caused by intestinal helminths. Instead, most studies have focused on estimating the prevalence of infections, only a fraction of which will be associated with disease. Perhaps the first estimate of the global prevalence of intestinal nematode infections was presented by Stoll (1947). His technique 244 Global Epidemiology of Infectious Diseases of extrapolation from survey data has proven robust and has been used to derive many of the more recent estimates which suggest that there are some billion infections with Ascaris lumbricoides and only slightly fewer infections with hookworm (both Necator americanus and Ancyclostoma duodenale) and Trichuris trichiura (World Health Organization 1987, Crompton 1988, 1989, Bundy & Cooper 1989). In the absence of reliable procedures for estimating disease directly, estimates of burden have been based on extrapolation from the infection prevalence data to provide an estimate of the proportion of infections that are likely to be associated with disease (Warren et al. 1993, Chan et al. 1994). This requires a careful and conservative approach because of the ubiquity of infection; extrapolating from such large numbers has the consequence that even small errors in estimating the disability attributable to individual cases can result in considerable inaccuracy in estimating the total burden. Because of the inherent potential for error in the use of any extrapolation procedure, the major aims of this chapter are to encourage independent scrutiny of the methodology and to highlight areas that particularly warrant further empirical study. This chapter builds on the studies of Chan et al. (1994) and Warren et al. (1993), and attempts to provide improved estimates of burden of disease based on extrapolation from observed estimates of prevalence of infection. The size of the population at risk of disability is estimated using epidemiological methods developed to describe the relationship between prevalence and mean intensity, and between intensity and potential morbidity (Guyatt et al. 1990, Lwambo, Bundy & Medley 1992), but modified to incorporate the heterogeneity between communities and age classes (Chan et al. 1994). The disability per case at risk is estimated using procedures described elsewhere in this series (Murray & Lopez 1996), and the regional and global burden of disease is obtained by extrapolation from these estimates. Basic biological characteristics of intestinal nematodes The analyses presented here focus on the four species of intestinal nematode that are of circumglobal distribution and occur at high prevalence: Ascaris lumbricoides, Trichuris trichiura, and the two major hookworm species, Necator americanus and Ancylostoma duodenale. A. lumbricoides is the large roundworm (15 cm) and lies free in the human duodenum where it feeds on lumenal contents (see Crompton, Nesheim & Pawlowski (1989) for further details of the biology of this parasite). Like all the other nematodes considered here, the worms are dioecious, which is to say that they exist as male and female. The female produces some 100 000 eggs per day which pass out in the faeces of the host and embryonate externally at a rate determined by local environmental factors. The eggs hatch on ingestion, releasing a larva that undergoes a Intestinal nematode infections 245 tissue migration involving the cardiovascular and pulmonary systems. The larva moults as it migrates and ultimately is coughed up from the lungs, swallowed, and becomes established as the adult in the small intestine. The cycle from egg deposition to female patency, when the female is able to produce eggs, has a duration of some 50 days. T. trichiura, the human whipworm, is a much smaller worm (25 mm) and inhabits the colon (Bundy & Cooper 1989). The anterior two-thirds of the worm is thin and thread-like and is laced through the mucosal epithelium, upon which the worm is believed to feed, leaving the blunt posterior projecting into the colonic lumen for excretion and oviposition. The female produces some 2000 eggs per day which pass out in the host faeces and embryonate externally. The infectious eggs hatch on ingestion and undergo a specifically local migration, via the crypts of Lieberkühn to the mucosal surface. The development cycle takes some 60 days. The two major hookworm species, which are of similar magnitude to the whipworm, inhabit the small intestine, where they attach to villi with biting mouthparts (Schad & Warren 1990). The worms feed on host blood and move frequently to new sites, leaving multiple, bleeding petechial haemorrhages on the mucosal surface. A. duodenale is believed to be more voracious, consuming some 0.14–0.40 ml of blood per day compared with 0.01–0.10 ml by N. americanus. The eggs pass out in host faeces and embryonate. Unlike the roundworm and whipworm, the eggs hatch externally releasing a mobile infective larva that actively infects the human host by dermal penetration. A. duodenale is also able to infect by the oral route, and there is evidence for vertical transmission of this species either transplacentally or in maternal milk. Having entered the host, the larva undergoes a tissue migration to the lungs and is coughed up and swallowed to moult to the adult in the small intestine. The cycle takes approximately 60 days for both species. A. duodenale is apparently able to extend this period by arresting development to the adult stage; a mechanism which may allow avoidance of seasonally hostile external conditions. Basic epidemiological characteristics of intestinal nematodes Understanding the epidemiology of helminth infections requires a fundamentally different approach from that required for all other infectious agents. Each worm’s establishment in a host is the result of a separate infection event, and the number of infective stages shed (the infectiousness of the host) is a function of the number of worms present. In population dynamic terms this implies that the individual worm is the unit of transmission for helminths, while the individual host is the unit for microparasites (Anderson & May 1982). The size of the worm burden (the intensity of infection) is therefore a central determinant of helminth transmission dynamics, and is also the major determinant of morbidity since the pathology is related to the size of the worm burden, usually in 246 Global Epidemiology of Infectious Diseases a non-linear fashion (Stephenson 1987, Cooper & Bundy 1987, 1988). Since the size of the worm burden varies considerably between individuals, and infection implies only that worms are present, a population of “infected” people will exhibit considerable variation in the severity of disease manifestations. The intuitive assumption that all infections are equal may help to explain the historical confusion over the pathogenicity and public health significance of helminth infection (Bundy 1988, Cooper and Bundy 1989). From these considerations it is apparent that an understanding of helminth epidemiology centres around an understanding of the patterns of infection intensity. Worm burden distributions Worm burdens are neither uniformly nor randomly distributed amongst individuals, but are highly over-dispersed such that most individuals have few worms while a few hosts harbour disproportionately large worm burdens. This pattern has been described for Ascaris lumbricoides (Croll et al. 1982), both species of hookworm (Schad & Anderson 1985) and Trichuris trichiura (Bundy et al. 1985). Most studies suggest that approximately 70 per cent of the worm population is harboured by 15 per cent of the host population. These few heavily infected individuals—the “wormy persons” described by Croll & Ghadirian (1981)—are simultaneously at highest risk of disease and the major source of environmental contamination. Studies of reinfection suggest that individuals are predisposed to a high or low intensity of infection; the size of the worm burden reacquired after successful treatment is positively associated with the intensity of infection before treatment. This association has been shown for all the major geohelminths (Anderson 1986, Elkins, Haswell-Elkins & Anderson 1986, Bundy & Cooper 1988, Holland et al. 1989), and persists over at least two reinfection periods (Chan, Kan & Bundy 1992). Longitudinal studies of T. trichiura (Bundy et al. 1988) and A. lumbricoides (Forrester et al. 1990) confirm that this positive association reflects a direct relation between the rate of reinfection and initial infection status. Thus in an endemic community it appears that there is a consistent trend for an individual to have an above (or below) average intensity of infection. This trend is more apparent in children with A. lumbricoides (Haswell-Elkins, Elkins & Anderson 1987) and T. trichiura (Bundy & Cooper 1988) infection, and more apparent in adults with Necator americanus (Bradley & Chandiwana 1990) infection, perhaps reflecting the different age-intensity patterns of these species. This pattern is also apparent at the family level (Chai et al. 1985, Forrester et al. 1988, 1990, Chan, Bundy & Kan 1994). Worm burden and age of host Over-dispersed distributions, where the highest intensity infections are aggregated in a few individuals of infection intensity, are observed in the community as a whole and also in individual age-classes. The degree of over-dispersion, however, shows some age-dependency. In hookworm Intestinal nematode infections 247 infection the distribution becomes more over-dispersed in adults (Bradley & Chandiwana 1990), while in A. lumbricoides and T. trichiura there is evidence that dispersion increases to a peak in the child age groups (Bundy et al. 1987b) and then declines in adults (Chan, Kan & Bundy 1992). These changes reflect age-specific trends in the proportion infected (the size of the zero class in the frequency distribution) and in the mean intensity of infection (the mean of the distribution). The age-dependent pattern of infection prevalence is generally rather similar amongst the major helminth species, exhibiting a rise in childhood to a relatively stable asymptote in adulthood. Maximum prevalence is usually attained before 5 years of age for A. lumbricoides and T. trichiura, and in young adults with hookworm infection. In A. lumbricoides there is often a slight decline in prevalence during adulthood, but this is less common with the other major nematode species. Prevalence data indicate the proportion of individuals infected and do not provide a simple indication of the number of worms harboured. The marked non-linearity of this relationship is a direct statistical consequence of the over-dispersed pattern of intensity (Guyatt et al. 1990). If worm burdens were normally distributed, there would be a linear relationship between prevalence and intensity (Anderson & May 1985). This is an important relationship since it is central to the method of extrapolation from infection prevalence to infection intensity described in this chapter. It is worth emphasising, therefore, that the relationships are firmly based on empirical studies of the major species of intestinal helminths (Guyatt et al. 1990, Lwambo et al. 1992, Booth 1994). The lack of simple correspondence between prevalence and intensity has the consequence that the observed age-prevalence profiles provide little indication of the underlying profiles of age-intensity. For most helminth species the initial rise in intensity with age closely mirrors that of prevalence but occurs at a slightly slower rate. Maximum intensity occurs at a host age which is parasite species-specific and dependent on parasite longevity, but independent of local transmission rates (Anderson 1986). For A. lumbricoides and T. trichiura maximum worm burdens occur in human populations at 5–10 years of age and for hookworms at 20–25 years. The most important differences in the age-intensity profiles of these species become apparent after peak intensity has been attained. A. lumbricoides and T. trichiura both exhibit a marked decline in intensity to a low level which then persists throughout adulthood. Age-profiles based on egg density in stool had suggested that there was considerable variation in the patterns seen in hookworm (Behnke 1987, Bundy 1990), but it now appears, from studies where burdens have been enumerated by anithelminthic expulsion, that the intensity attains a stable asymptote, or rises marginally, in adulthood (Pritchard et al. 1990, Bradley et al. 1992). Thus, for those species with convex age-intensity profiles, but asymptotic age-prevalence profiles, a similar proportion of children and adults are infected but the adults have substantially smaller worm burdens. With 248 Global Epidemiology of Infectious Diseases hookworm infection, where both prevalence and intensity are asymptotic, more adults are infected and they have larger worm burdens. Definition and measurement For most helminthiases, the relationship between infection and disease is likely to be non-linear and complex. If we accept, for the moment, the simple premise that only heavy worm burdens cause disability, then it is apparent that disability will have an age-dependent distribution, since intensity is age-dependent, and also that disability will have an over-dispersed pattern even within the susceptible age-classes. The relationship is further complicated by the non-linear relationship between the severity of disease and the intensity of infection, and by the interaction between symptomatology and the chronicity of infection. Helminthic infection does not inevitably lead to disease: a failure to appreciate this has led to apparently contradictory results from morbidity studies and may be the major contributor to the under-recognition of the public health significance of helminthiasis (Cooper & Bundy 1987, 1988). An essential prerequisite, therefore, to assessing the global burden of disease is the estimation of the sub-population of the infected population that has a sufficiently large worm burden to put them at risk of disability. This requires some method of relating prevalence of infection data, which are available from empirical studies in all geographical regions, to intensity of infection data, which are not. In this section we describe procedures for extrapolating intensity from prevalence survey data, and for partitioning the risk of disability. Worm burdens and morbidity There is a general acceptance of the simple view that very intense infection results in illness, a view that reflects both clinical experience of overwhelming infection and, perhaps equally importantly, an atavistic repugnance at the insidious invasion of the body by large numbers of worms. Such extremes of infection result in the severe anaemia of necatoriasis and the intestinal obstruction of ascariasis (Stephenson 1987), and the chronic colitis of classical trichuris dysentery syndrome (Cooper & Bundy 1988). That helminth morbidity is dependent on infection intensity is, from this perspective, uncontroversial. An understanding of the pattern of the relationship between infection intensity and clinical signs has proven more elusive. This appears to be the result of two main factors. First, intensity and pathogenesis are non-linearly related. Studies of the anaemia associated with hookworm infection indicate that there is a disproportionate reduction in plasma haemoglobin concentration after some threshold worm burden is exceeded. Although profound anaemia is associated with thousands of worms, clinically significant anaemia can be induced by a few hundred worms, the precise threshold depending on the host’s iron status (Lwambo, Bundy & Hedley 1992). Note that this Intestinal nematode infections 249 occurs despite the constant per capita blood loss attributable to hookworm feeding (Martinez-Torres et al. 1967), which might intuitively be expected to give a linear relationship between burden and anaemia. Studies of protein-losing enteropathy in trichuriasis also indicate a non-linear relationship with worm burden (Cooper et al. 1990). The rate of gut clearance of a-1-antitrypsin at first rises rapidly with increasing worm burden to a threshold and rises more slowly thereafter. The relationship is markedly non-linear. This implies that significant intra-lumenal leakage of protein can occur with T. trichiura burdens of a few hundred worms. The clinical consequences of this loss will be determined by the nutritional and dietary status of the host and by the chronicity of infection (Cooper, Bundy & Henry 1986). In one study, children with the classical Trichuris Dysentery Syndrome had all experienced mucoid dysentery and rectal prolapse for more than three years (Cooper & Bundy 1987, 1988). The second reason for the lack of understanding of the relationship between intensity and disease is the difficulty of measuring and attributing morbidity. This is in part the classical epidemiological problem of identifying specific morbidity in an endemic population subjected to multiple insults (Walsh 1984). It is exacerbated for helminth infection, however, by the absence of pathognomonic signs in moderate, but clinically significant, infection. This problem has been addressed by intervention trials using specific antihelminthic therapy. Such studies have shown, for example, significant improvements in linear growth after treatment of moderately infected and stunted children with hookworm, A. lumbricoides or T. trichiura infection (Stephenson 1987, Stephenson et al. 1990, Cooper et al. 1990). Even more subtle consequences of infection are suggested by recent double-blind placebo trials, which show significant improvement after antihelminthic treatment in the cognitive ability of school-children moderately infected with T. trichiura (Nokes et al. 1991, 1992a). These results suggest that even moderate helminthic infection may have insidious consequences that are unlikely to be attributed to helminthiasis in public health statistics. In one study of a village population with hyperendemic geohelminthiasis (Cooper, Bundy & Henry 1986), only 2 per cent of actual morbidity had been reported to the health authorities. Some insights into the relationship between infection and disability can be provided by data analysis (Guyatt et al. 1990, Lwambo, Bundy & Medley 1992). Empirical studies can provide estimates of the threshold number of worms associated with risk of disability (see above). Then, using models which describe the empirical relationship between infection intensity and prevalence, it is possible to estimate the proportion of individuals in whom the threshold worm burden is exceeded (and are thus are likely to suffer disability) at a given prevalence of infection. Non-linear relationships of this form have been shown in studies of schistosome infections, for which adequate country based morbidity data are available (Jordan & Webbe 1982). There is a need to obtain similar data for intestinal nematodes, but this is currently confounded by the difficulties in estimating morbidity 250 Global Epidemiology of Infectious Diseases directly for these infections. One important conclusion of this analysis, however, is that the threshold need not be precisely defined, since the form of the relationship between infection and disease is relatively insensitive to the threshold value, provided the value is relatively large (> 20 worms). These analyses indicate that the proportion at risk of disability increases dramatically as infection prevalence rises. For example, if 25 A. lumbricoides worms are associated with disease, then 20 per cent of the population will be at risk at an infection prevalence of 80 per cent, and less than 2 per cent at an infection prevalence of 70 per cent. It is worth noting that both these infection prevalences could reasonably be considered to be high, which may help explain why studies of helminth morbidity in “high” prevalence areas often reach very different conclusions about the public health significance of helminthiasis (for example, Keusch 1982). The relationships described here form the basis for the extrapolation procedure developed in this chapter. This procedure involves the use of infection prevalence survey data from individual countries to give an estimate of regional and global prevalence of infection. Empirical data are then used to estimate the fraction of this infected population which is at risk of disability. The following sections describe this procedure and discuss the validity of its assumptions. Estimation of risk of disability from infection prevalence data The risk of disability is estimated on the basis of the relationship between worm burden and prevalence of infection. The frequency distribution of worm burdens between individuals has been consistently shown to be highly over-dispersed. Within a community, the majority of people have few or no worms and a few people have very high worm burdens. Observed distributions can be represented empirically by a negative binomial distribution (Anderson & Medley 1985, Guyatt et al. 1990, Lwambo, Bundy & Medley 1992, Bundy & Medley 1992). This theoretical distribution has two parameters, the mean worm burden, m, and the aggregation parameter, k (see Anderson and Medley, 1985 for explanation of this parameter). General values used in this study are k = 0.54 for A. lumbricoides (Guyatt et al. 1990), k = 0.23 for T. trichiura (Booth 1994) and k = 0.34 for hookworms (Lwambo, Bundy & Medley 1992). These values are estimated from empirical data from field surveys that estimated the distribution of intensity in human populations. The basis for estimating potential disability is that a) the risk of disability is higher in individuals with higher worm burdens and b) there is some threshold worm burden above which disability is more likely to occur (Table 9.1). The proportion of the population with worm burdens higher than this threshold can then be estimated from the negative binomial distribution (Guyatt & Bundy 1995, Lwambo, Bundy & Medley Intestinal nematode infections Table 9.1 251 Worm burden thresholds for morbidity used in the model Higher estimate, lower threshold Lower estimate, higher threshold 0–4 5–9 10–14 15+ 10 15 20 20 20 30 40 40 T. trichiura 0–4 5–9 10–14 15+ 90 130 180 180 250 375 500 500 Hookworms 0–4 5–9 10–14 15+ 20 30 40 40 80 120 160 160 Species Age (years) A. lumbricoides Note: The lower thresholds are based on empirical observations of worm numbers associated with developmental deficits. The higher thresholds are a more conservative value intended to provide a lower bound to the estimate of morbidity. The thresholds were estimated for children in the 5–10 year age group, and then adjusted for other age group using the procedure described in the text. The lower threshold estimate for Ascaris lumbricoides assumes that 30 000 eggs per gram (epg) is associated with deficits in growth and fitness (see Stephenson et al. 1989,1990) and that worm fecundity is equivalent to approximately 3000 epg for female worm. The higher threshold assumes that morbidity is associated with twice this burden. The Trichuris trichiura lower threshold is taken from studies of growth stunting in 5–10 year old children (Cooper et al. 1990), while the higher threshold is estimated from studies of clinical colitis (ES Cooper, personal communication). The hookworm thresholds are based on upper and lower bound estimates of the relationship between infection intensity and anaemia (Lwambo, Bundy & Medley 1992). 1992, Medley, Guyatt & Bundy 1993). This approach is necessarily an approximation, since the effects of a given worm burden on an individual will be modified by that individual’s condition (e.g. nutritional status and concurrent infections) and the chronicity of the infection. Studies indicate that developmental effects of infection (e.g. cognitive and growth deficits) occur at lower worm burdens than the more serious clinical consequences. We therefore use two sets of thresholds for each species, where the lower threshold corresponds to a higher estimate of potential disability from cognitive and growth deficits. These thresholds are shown in Table 9.1. The thresholds for A. lumbricoides and T. trichiura correspond to those given by Chan et al. 1994. The thresholds for hookworm have been adjusted downwards. This reflects the analyses presented by Crompton & Whitehead (1993), and the observation that the thresholds reviewed by Lwambo, Bundy & Medley (1992) apply to adults rather than children, as had been assumed in the previous calculations. Because children are smaller than adults, it is assumed that a similar worm burden is associated with more morbidity in children. There is a lack of data on the relationship between disability threshold and age. However, since a given worm burden is more likely to cause disability in children than in adults, the use of a single, age-independent threshold would tend to considerably overestimate the potential disability. 252 Global Epidemiology of Infectious Diseases In order to approximate this age effect, in the absence of empirical data, the same proportional changes with age are used for all worm species. The threshold for children under 5 years of age was taken as 50 per cent of that for adults (i.e. adults require twice the worm burden of pre-school children before suffering ill effects), for 5–9 year old children it was taken as 75 per cent of that for adults, and the adult threshold was used for 10–14 year old children (Table 1). These estimates are, we believe, conservative; but they are not firmly based on empirical observation. Other age-dependent differences are also incorporated into the model. A. lumbricoides and T. trichiura infections are usually more prevalent in children, whereas hookworm infections are more prevalent in adults. For simplicity, a single age-prevalence relationship (one for each species) was used, based on the typical age-prevalence relationship observed in field studies (Table 9.2). The use of different prevalences for different age groups implies that a separate negative binomial distribution must be calculated for each age group. In order to ensure that the overall prevalence remains unchanged by this procedure, the observed demographic age distribution of the population must be taken into account. The effect of host age on the aggregation parameter (k) remains undefined, and the present framework uses a single (species-specific) aggregation parameter for all age groups. Incorporating geographical heterogeneity The prevalence of infection is non-linearly related to the mean intensity of infection in a community, such that the proportion of the population potentially suffering morbidity is disproportionately greater at higher levels of prevalence (Guyatt & Bundy 1991). If the average prevalence among communities is used as a basis for estimating intensity (and potential disability) for a geographical region, this will grossly underestimate the actual morbidity. It is therefore necessary to incorporate geographical Table 9.2 Age weights for prevalences used in the model Species Age (years) A. lumbricoides 0–4 5–9 10–14 15+ Age weight 0.75 1.2 1.2 1 T. trichiura 0–4 5–9 10–14 15+ 0.75 1.2 1.2 1 Hookworms 0–4 5–9 10–14 15+ 0.2 0.5 0.9 1 Intestinal nematode infections 253 heterogeneity in prevalence within the estimation procedure (Chan et al. 1994). Spatial heterogeneity in intestinal nematode infection is a relatively neglected area of study (Bundy et al. 1991, Booth & Bundy 1992) but its potential importance has been convincingly demonstrated for microparasitic infections (Anderson 1982, May & Anderson 1984). Heterogeneity was considered at several geographical levels in the extrapolation procedure. The highest level is the regional level, as defined for the eight World Bank regions. Two regions (EME and FSE) were excluded from the analysis since the prevalences of intestinal nematode infections are very low. The remaining six regions were then divided into “population units” of 20 to 100 million people for which mean prevalence values, based on empirical data, were input into the model. Generally these population units coincide with politically defined countries (which is the unit for which prevalence data are most readily available) but some countries with exceptionally large populations, especially China and India, were subdivided into smaller units based on states or provinces for prevalence estimates. The subdivision of a region into population units captures some geographical heterogeneity, but this does not include the heterogeneity within the population unit (or country). Literature searches yielded suitable data for examining intra-country variation in the six geographical regions. The level of heterogeneity among communities within the same country was assessed using community-level estimates of prevalence from different studies. Variation in prevalence within countries was estimated for 9 countries for A. lumbricoides, 8 countries for T. trichiura, and 10 countries for hookworm. The number of prevalence surveys available within each country ranged from 21 to 115. A total of 1600 prevalence surveys were examined in this analysis. The within-country distributions differed between worm species and between countries but were not markedly skewed or asymmetrical. Good correspondence was found between the data and the theoretical normal distribution which was therefore used. Highly significant positive correlations between mean and standard deviation were observed for A. lumbricoides and T. trichiura distributions. Therefore, an estimate of “typical” geographical heterogeneity within a population unit could be estimated for these species from the regression. Hookworm distributions show a wider range of standard deviations and there was no significant correlation between these and the means. This may reflect the fact that the hookworm data include undifferentiated estimates for two quite different parasite species (A. duodenale and N. americanus). This additional source of variation was captured by taking the mean of the standard deviations as an estimate of heterogeneity within the population unit. These estimates of the geographical variation in prevalence within a population unit were assumed characteristic of each species and were incorporated in the model for all subsequent calculations. 254 Global Epidemiology of Infectious Diseases Summary of estimation procedure assumptions The framework for estimation of the population at risk of disability involves a set of assumptions about the patterns of infection observed at both the community and the regional level. Community-level assumptions Worm burden frequency distributions are adequately described by the negative binomial distribution. This probability distribution is the most widely accepted empirical description of observed worm burden distributions (Anderson & May 1991). However, it has been shown in some parasite species to underestimate the proportion of very low worm burdens and thus, potentially, overestimate the number of people in the higher worm burden classes. This could lead to an overestimation of the population at risk. The frequency distribution of worm burdens is species-specific and largely independent of geographical region, infection prevalence or age group. Analyses of the available empirical data suggest that the degree of aggregation, as assessed by the negative binomial parameter, k, is remarkably consistent between studies and largely independent of geographical region for all the nematode species considered here (Guyatt et al. 1990, Lwambo, Bundy & Medley 1992, Booth 1994). There is some evidence that k increases slightly with prevalence of A. lumbricoides (Guyatt et al. 1990) and even more marginally with hookworm prevalence (Lwambo, Bundy & Meadley 1992). Exclusion of this effect would lead to overestimation of potential morbidity. There is conflicting and limited evidence for the relationship between worm burden distribution and age, with some studies showing an increase and others a decrease in aggregation with age (Bundy et al. 1987b). The current assumption of an age-independent distribution could lead to either overestimation or underestimation of potential morbidity. Disability occurs above a worm burden threshold that is higher for adults than children. Thresholds were estimated from empirical data (Table 9.1). Two sets of thresholds were used for each species. An overestimate of the threshold worm burden would lead to an underestimate of potential morbidity and vice versa. No information is available on the variation of the threshold with age. The age prevalence profile can be generalized between communities. A similar general pattern is seen when age prevalence profiles from different studies of the same species are compared (Anderson & May 1985). The effect on the estimates of changing the shape of the age-prevalence profile are likely to be complex but, in general, the larger the differences in prevalence between different age groups, the higher the estimate of population at risk. Intestinal nematode infections 255 Regional-level assumptions Infection prevalences between communities in the same country are normally distributed. Examination of the actual distributions suggested that a normal approximation would be appropriate (Chan et al. 1994). The use of a symmetrical distribution is the most conservative assumption since with a skewed distribution with the same mean (such as the negative binomial distribution), the frequency of very high prevalences will be increased. The current assumption may therefore tend to underestimate the population at risk. The standard deviations of these normal distributions increase linearly with the mean prevalence for A. lumbricoides and T. trichiura and are independent of mean prevalence for hookworms. These assumptions are the best available estimates of the effect of spatial heterogeneity on the prevalence distribution, and are based on data presented by Chan et al. (1994). The mean standard deviation relationship is the same in different geo- graphical regions of the world. The data available suggest a consistent relationship but there are insufficient data to assess this relationship by region. It is not known if there are any regional differences nor whether these might increase or decrease the estimates. A population unit of 20 to 100 million people is a sufficiently fine- grained spatial stratification to capture geographical variation in a population of 4.1 billion people. The size of this unit is constrained by the availability of empirical data. Larger units would reduce the precision of the estimates of potential morbidity. Estimation procedure The method used for estimation is essentially an integration of all the processes described in the text. Fuller details are given by Chan et al. (1994). In summary, the procedure involved the following steps. 1. Country (or other population unit) prevalence data were obtained and divided into prevalence classes. Five prevalence classes were defined and an intermediate prevalence value (S) was used for calculation purposes. These classes are shown in Table 9.3. In this round of calculations the highest two prevalences classes were combined and the set prevalence of the lower class was used for estimation (67 per cent) of community morbidity. This was necessary because with very high prevalences, the negative binomial distribution greatly overestimates potential morbidity. This calculation is not exactly equivalent to merging the two highest prevalence classes since countries in class 5 have a different distribution of community prevalences with a greater proportion in the class 4+5. The estimated morbidity for these countries will therefore be higher. 256 Global Epidemiology of Infectious Diseases Table 9.3 Prevalence classes and reference (set) prevalences used in the model Class Prevalence range (percentage) Reference prevalence (percentage) 1 2 3 4 5 4+5a < 25 25–45 45–60 60–75 75–100 60–100 10 35 52 67 80 67 a. Combined classes 4 and 5 were used for community prevalence distribution. 2. The total populations in each prevalence class were multiplied by the transition matrix (M) to give an estimated community distribution of prevalences for the region (C). The matrix is derived using the following procedure. For each reference prevalence class in the estimation, a normal distribution for the individual community prevalence class distribution in the countries concerned was calculated. For A. lumbricoides and T. trichiura a standard deviation that increased with mean was used, whereas with hookworms a constant standard deviation given by the average standard deviation of the data sets was used. The regional community distribution of prevalences can then be obtained by multiplying a vector of the country mean prevalence distribution with a species-specific transition matrix which consists of the calculated normal distributions: C = R.M. The community prevalence distribution has one zero prevalence class and four non-zero-prevalence classes equivalent to the four lower prevalence classes of the country prevalence distribution (Table 9.3). Note that in the previous estimation (Chan et al. 1994) a fifth prevalence class (> 75 per cent) was included. This is now considered to over-emphasize the top end of the prevalence range and thus to overestimate the population at risk, and was therefore not used in the present calculations. The transition matrices used for each of the species are shown respectively, in Tables 4a, 4b and 4c. 3. Using regional demographic data, the population is divided into classes of community prevalence and age group and the age-weighted prevalence is calculated for each of these classes. The age weights (A) are shown in Table 9.2. Given a community prevalence si for prevalence class i, the prevalence in adults (pi,15+) is given by: pi,15+ = ∑ j si (a jd j) Intestinal nematode infections Table 9.4a 257 Transition matrix for Ascaris lumbricoides Community prevalence distribution Reference class 1 2 3 4 5 0 1 2 3 4+5 0.251 0.081 0.057 0.021 0 0.590 0.264 0.149 0.081 0.048 0.149 0.310 0.214 0.149 0.097 0.010 0.186 0.175 0.169 0.133 0 0.159 0.405 0.580 0.722 Note: A zero class community distribution denotes a prevalence of 0%. Table 9.4b Transition matrix for Trichuris trichiura Community prevalence distribution Reference class 1 2 3 4 5 0 1 2 3 4+5 0.288 0.097 0.068 0.028 0.011 0.509 0.259 0.153 0.087 0.047 0.177 0.288 0.200 0.149 0.101 0.026 0.180 0.170 0.157 0.125 0 0.176 0.409 0.579 0.716 Note: A zero class community distribution denotes a prevalence of 0%. Table 9.4c Transition matrix for hookworms Community prevalence distribution Reference class 1 2 3 4 5 0 1 2 3 4+5 0.309 0.040 0 0 0 0.464 0.269 0.089 0.018 0 0.187 0.382 0.274 0.118 0.040 0.04 0.203 0.292 0.227 0.119 0 0.106 0.345 0.637 0.841 Note: A zero class community distribution denotes a prevalence of 0%. where aj is the age weight for age group j and dj is the proportion of the population in age group j. The age specific prevalence in the other age groups are then given by: pij = pi,15+ a j . 4. For each of the classes, using the species-specific aggregation parameter (k) and age specific morbidity threshold (ttj), the potential morbidity estimate (eij) is calculated using the negative binomial distribution. This theoretical distribution has two parameters, the mean worm burden µ, and the aggregation parameter k. These are related to the prevalence (P) in the following way: µ −k P = 1 − (1 + ) k 258 Global Epidemiology of Infectious Diseases The basis for the estimation of the population at risk of disability is that morbidity is mainly confined to the fraction of the population with high worm burdens. A threshold worm burden (T ) is defined over which morbidity effects are potentially observed (Table 9.1). The individual terms of the negative binomial, p(x),(the proportion of individuals with x worms) are given by: µ π (x) = 1 + k −k Γ(k + x) µ x! Γ(k) µ + k x where Γ represents the gamma function. The proportion of the population with more than (T) worms (the morbidity function Morb(P,T)) is therefore given by: x =T Morb(P, T ) = 1 − ∑ π (x ) x=0 The potential morbidity estimate (eij) is obtained by multiplying the morbidity function by the population in each class (nij): e ij = nij Morb(pij , t j) 5. The above estimates are summed to give the age specific population at risk estimates for each region (ffj): fj = ∑ eij i These procedures were followed for each of the parasitic infections under consideration. Estimates of population infected and at risk of disability The estimates of population infected and at risk or disability are shown in Tables 9.5 to 9.7. For each nematode species, these estimates are given for different age classes and for different geographical regions. Two estimates of population at risk were used for each infection, based on different estimates of worm burden thresholds. Both thresholds are based on empirical data and were chosen to be relatively conservative. The lower estimate of worm burden reflects probable developmental consequences of infection such as impaired growth or fitness, while the higher estimate is intended to reflect the likelihood of more serious consequences of infection. The estimates of population infected are all slightly lower than those presented by Chan et al. (1994), because of the exclusion of the highest reference prevalence class from the present extrapolation procedure. This has had an even greater effect on reducing the estimated size of the Intestinal nematode infections Table 9.5a Region 259 Estimates of numbers infected and at risk of disability for Ascaris lumbricoides, by World Bank region Population (millions) Infections (millions) Infections (as a percentage) Population at risk (millions) Population at risk (percentage) SSA 512 104 20.25 3.04 0.78 0.59 0.15 LAC 441 171 38.76 8.04 1.92 1.82 0.44 MEC 503 96 19.15 2.94 0.74 0.58 0.15 IND 850 188 22.13 6.60 1.62 0.78 0.19 1 160 532 45.85 25.320 5.72 2.18 0.49 OAI 654 291 44.53 16.380 3.99 2.50 0.61 Total 4 120 1 382 CHN Table 9.5b Age group (years) 62.320 14.760 Estimates of numbers infected and at risk of disability for Ascaris lumbricoides, by age group Population (millions) Infections (millions) Infections (percentage) Population at risk (millions) Population at risk percentage) 0–4 553 130 24 1.84 0.05 0.33 0.01 5–9 482 182 38 27.810 9.45 5.77 1.96 10–14 437 169 39 17.990 4.65 4.12 1.06 15+ 2 647 901 34 14.680 0.61 0.55 0.02 Total 4 120 1 382 34 62.320 14.760 1.51 0.36 population at risk of disability from ascariasis and trichuriasis, and on the estimated population above the high threshold for hookworm infection. The estimated population above the lower threshold of hookworm burdens has, however, increased as a result of the change in threshold. The estimates for A. lumbricoides are shown in Table 9.5. The estimated total number of A. lumbricoides infections is 1382 million, slightly higher than estimates from other sources (World Health Organization 1987, Crompton 1988, Bundy 1990). The number at risk of morbidity is estimated in the range 15–63 million. The overall infection prevalence is 34 per cent in the exposed population while the prevalence of those at risk 260 Global Epidemiology of Infectious Diseases Table 9.6a Region Estimates of numbers infected and at risk of disability for Trichuris trichiura, by World Bank region Population (millions) Infections (millions) Infections (percentage) Population at risk (millions) Population at risk (percentage) SSA 512 85 16.57 2.21 0.97 0.43 0.19 LAC 441 146 33.16 7.27 2.97 1.65 0.67 MEC 503 64 12.64 0.04 0.00 0.01 0.00 IND 850 134 15.78 3.02 1.25 0.36 0.15 1 160 340 29.29 17.290 6.65 1.49 0.57 OAI 654 238 36.41 15.700 6.54 2.40 1.00 Total 4 120 1 007 24.43 45.530 18.390 1.11 0.45 CHN Table 9.6b Age group (years) Estimates of numbers infected and at risk of disability for Trichuris trichiura, by age group Population (millions) Infections (millions) Infections (percentage) Population at Population at risk risk (millions) (percentage) 0–4 553 96 17.38 0.04 0.00 0.01 0.00 5–9 482 135 28.04 19.920 10.400 4.13 2.16 10–14 437 125 28.60 15.930 7.69 3.65 1.76 15+ 2 647 650 24.57 9.63 0.30 0.36 0.01 Total 4 120 1 007 24.43 45.530 18.390 1.11 0.45 in the exposed population is between 0.4 per cent and 1.5 per cent. The highest prevalences are found in China and the Other Asia and Islands region. The model assumes that prevalence of A. lumbricoides infection is slightly higher in children of 5 to 15 years old than in adults and younger children (Anderson & May 1985). The results show that this difference is greatly magnified in the estimates of population at risk, such that potential morbidity is significantly higher in school age children than in any other Intestinal nematode infections Table 9.7a Region 261 Estimates of numbers infected and at risk of disabililty for hookworm, by World Bank region Population (millions) Infections (millions) Infections (percentage) Population at risk (millions) Population at risk (percentage) SSA 512 138 26.93 18 8 2.46 1.06 LAC 441 130 29.40 15 3 3.44 0.70 MEC 503 95 18.81 8 2 1.54 0.47 IND 850 306 36.00 47 11 5.48 1.32 1 160 340 29.29 29 3 2.47 0.24 OAI 654 242 37.04 35 8 5.29 1.29 Total 4 120 1 250 30.34 151 36 3.66 0.87 CHN Table 9.7b Age group (years) Estimates of numbers infected and at risk of disability for hookworm, by age Population (millions) Infections (millions) Infections (percentage) Population at risk (millions) Population at risk (percentage) 0–4 553 41 7.35 0 0 0.00 0.00 5–9 482 89 18.54 0 0 0.00 0.00 10–14 437 145 33.31 10 1 2.29 0.16 15+ 2 647 975 36.81 141 35 5.33 1.32 Total 4 120 1 250 30.34 151 36 3.67 0.87 age group. This is a result of the non-linear relationship between prevalence of infection and potential morbidity (Guyatt et al. 1990). The estimates for T. trichiura are shown in Table 9.6. The totals show a lower number of infections than for A. lumbricoides, there being 1007 million infections resulting in a global prevalence of 24 per cent. The total population at risk is also lower for T. trichiura, in the range 18–45 million. As with A. lumbricoides, the highest prevalences of disability risk are found in Other Asia and Islands, in China and in children of school age. The estimates for hookworm infections are shown in Table 9.7. There 262 Global Epidemiology of Infectious Diseases are an estimated 1250 million hookworm infections and between 36 million and 151 million people at risk of disability. The distributions of infection and morbidity, both by age and region, are different from those for A. lumbricoides and T. trichiura. The disability risk attributable to hookworm infections is much higher in adults. The regional distribution is also different, with the highest prevalence of estimated disability in India. Review of empirical databases by region Direct estimates of the community morbidity attributable to intestinal helminthiases are unavailable. Hence the disability-adjusted life year (DALY) estimates are based on extrapolating the population at risk (intensely infected) from empirical observations of the proportion of the population infected. The data on prevalence of infection used in the current estimation are from a database held at Oxford University, derived from field survey data compiled for a UNESCO report on global prevalence of helminth infection (Bundy & Guyatt 1990). They are based on an extensive search of the original literature and, as far as possible, represent data collected within the last 20 years; older data were used only if no other data were available for a particular country. Additional criteria for data selection include large sample size and community-based studies (i.e. not hospital records or institutional data). The number of studies available for each country varied, hence the reliability of the estimate for mean prevalence also varies. For the majority of countries, the sample size was at least several thousand individuals. Note that the data presented in Table 9.8 exclude a substantial number of unpublished surveys used in the actual analysis. The survey estimates are based on the microscopic detection of parasite eggs in faecal specimens. While A. lumbricoides and T. trichiura eggs can be readily identified, the eggs of the two hookworm species, Ancyclostoma duodenale and Necator americanus, cannot be distinguished by normal diagnostic methods and are recorded here as the combined prevalence of both species. The stool examination procedure also fails to detect light infections, particularly when single examinations are made as in the case of field surveys (Hall 1982). Furthermore, the procedure will not detect non-fecund infections (e.g. single worm or single sex) which may represent a significant minority of infections (Guyatt 1992, Guyatt & Bundy 1995). Thus the survey data are conservative and underestimate the true prevalence of infection. Estimation of DALYs The analyses to this point have produced estimates of the size of the populations with worm burdens that are likely to result in some form of disability. This section focuses on the estimation of the proportion of the population at risk who are likely to be disabled and the degree of disability, and how 214 126 88 32 276 1 059 5 506 2 684 684 422 690 217 3 313 Central African Republic Gabon Gambia Ghana Liberia Madagascar Madeira Ethiopia 1 572 423 135 913 27 040 5 040 Cape Verde Burkina Faso Cameroon Benin Sample size 21 3 4 8 1 2 5 3 2 290 41 95 1 1 2 530 18 Number of sites 39 29 52 17 61 7 40 19 72 0.5 45 60 Ascaris 30 19 20 0.5 68 40 17 53 35 Trichuris Prevalence estimate (percentage) 17 60 A.d = 30 A.d = 18 N.a = 0.1 N.a = 20 A.d = 1 43 N.a = 34 A.d = 83 N.a = 49 N.a = 19 Hookworma Garin 1978, Richard-Lenoble 1982 McGregor & Smith 1952; Marsden 1963 Annan et al. 1986 Sturchler et al. 1980 Cerf, Burgess & Wheeler 1981 Santos 1952 Tedla & Ayel 1986 Taticheff, Abdulahi & Haile-Meskal 1981 continued Pampliglione & Ricciardi 1971 Faucher et al. 1984 Ratard et al. 1991 Carrie 1982, Ripert et al. 1978, Ripert, Leugueun-Ngougbeou & Same-Ekobo 1982 de Meira, Nogueira & Simoes 1947 Barbosa 1956, Nogueira & Coito1950 Ricciardi 1972 Brumpt et al.,1972 Source Summary of prevalence studies of helminth infections, by World Bank region and country or area Sub-Saharan Africa Country or area Table 9.8 Intestinal nematode infections 263 China India Togo Uganda United Republic of Tanzania Zambia Zimbabwe Nigeria Réunion Senegal Somalia Sudan Malawi Mali Country or area Table 9.8 1 477 742 2 843 59 48 17 4 920 38 484 595 23 949 1 270 6+ 3 1 10 4 5 3 6 10 4 5 3 93 504 157 276 289 2 174 2 974 1 797 1 266 2 803 359 556 319 Sample size Number of sites 47 8 0.5 7 0 2 67 49 36 13 4 0.5 Ascaris 19 16 0.5 1 0 43 46.5 85 15 34 0.5 Trichuris Prevalence estimate (percentage) 17 13 10 0 A.d. =7 38 N.a. = 62 A.d. = 8 N.a. = 58 31 24 N.a = 9 15 Hookworma Yu 1994 Saxena & Prasad 1971 Bidinger, Crompton & Arnold 1981; Arora 1975; Chowdury 1968; Chatterjee & Mukhopadhyay 1985; Mukerji & Mathen 1950; Lane et al. 1917; Bagchi, Prasad & Mathur 1964; Sanyal et al. 1972 Wenlock & Ash 1977 Goldsmid et al. 1976 Blackie 1932; Goldsmid 1976 Oduntan 1974 Bonnefoy & Lsautier 1978 Juminer, Diallo & Laurens 1971 Bianchini et al. 1981 Ilardi et al. 1980, 1981, 1987 Kuntz, Lawless & Mansour 1955 Lapierre, Tourte-Schaeffer 1982 Sorvillo 1982 Sturrock 1964 Burgess, Burgess & Wheeler 1973 Rougement et al. 1974 Ranque 1982 Source Summary of prevalence studies of helminth infections, by World Bank region and country or area (continued) 264 Global Epidemiology of Infectious Diseases Guatemala Mexico Surinam Costa Rica Dominica Ecuador Brazil Chile Latin America & Caribbean Viet Nam Singapore Taiwan, China Thailand 2 511 34 799 90 277 33 161 1 000 3 970 1 568 15 383 604 854 10 38 81 3 2 14 12 1 1 10 30 33 32 4 3 7 7 85 101 8 96 075 1 304 1 023 1 300 286 614 622 67 300 68 153 409 Philippines 1 11 398 2 493 Republic of Korea Lao People’s Democratic Republic Malaysia Papua New Guinea 77 51 364 Indonesia Other Asia & Islands 73 62 62 12 38 50 60 19 33 45 91 6 54.8 48 24 14 49 71 33 4 N.a. = 11 26 28 92 60 1819 67 45 27 35 57 36 14 37 7 72 22 2 69 72 44 5.8 49 0.5 A.d = 41 29 55 42 50 48 continued Vinha 1971 Neghme & Silva 1952, 1963, 1983; Schenone et al. 1981; Ramirez et al. 1972; Puga et al. 1980 Zamora et al.1978; Rojas & Zumbado 1980 Grell & Zumbado 1981 Peplow 1982 Ortiz 1969 Aguilar 1981 Stoopen & Beltran 1964 Asin & van Thiel 1963 Kan 1982, 1985; Russell 1928, 1934; Dunn 1972 Sinniah et al. 1978 Ashford et al. 1981 Shield et al. 1986 Cross et al. 1977c Tatengco, Marxan & Castro 1972 Desowitz 1963 Bergner 1964; Chang, Sun & Chin 1973 Jones 1976 Preuksaraj et al. 1981; Papasarathorn et al. 1975 Sadun 1955 Colwell et al. 1971 Cross et al. 1970, 1975, 1976, 1977a & b; Clarke et al. 1973; Carney et al. 1974, 1977; Stafford et al. 1980; Joseph et al. 1978 Seo et al. 1983 Sornmani et al. 1974 Intestinal nematode infections 265 1 137 4 000 61 995 62 004 295 835 24 047 Iran Iraq Morocco 7 16 3 3 3 1 6 8 5 61 0.1 11 1 22 41 16 Ascaris 2 55 0.1 c. Hookworm only among Jasmin workers. 49 0 A.d = 29 1.3 A.d. = 1 17 0.5 13 Hookworma 29 Trichuris Prevalence estimate (percentage) b. For hookworm, Ancylostoma duodenale accounts for 50% and Necator americanus makes up 2%. a. N.a = Necator americanus; A.d. = Aneaslostoma duodenale. Pakistan Saudi Arabia Tunisia 565 41 476 39 574 Algeria Egypt Sample size Number of sites Kuntz 1960 El-Rahimy et al. 1986 Thiers, Lassoued & Abid 1976 Pampiglione & Hadjeres 1965 Mohamed et al. 1985 Farag 1985; Tadros 1973; Wells & Blagg 1956; Mohamed et al. 1988; Lawless, Kuntz & Strome 1956; Colett 1966; Nazari & Massoud 1980 Niazi et al. 1975 Cadi-Soussi et al. 1982 Source Summary of prevalence studies of helminth infections, by World Bank region and country or area (continued) Middle Eastern Crescent Country or area Table 9.8 266 Global Epidemiology of Infectious Diseases Intestinal nematode infections 267 this varies with such factors as age and sex. It also attempts to determine the relatively rare mortality attributed to intestinal nematode infection. The calculation of disability-adjusted life years (DALYs) is based on the population at risk of disability. It is assumed that there are three sources of DALY loss: contemporaneous disability, which occurs in individuals with worm burdens above the higher threshold in Table 9.1 and which persists as long as the individual remains infected; chronic disability, which occurs in a small proportion of children with worm burdens above the lower threshold and which is life-long; and life years lost from mortality. The sum of these three effects gives the overall DALY estimate. Population at risk of disability Estimates of the populations at risk of disability are shown in Tables 9.9 to 9.11. Note that these populations are not the populations infected but some fraction of these which are at risk of disability because their worm burdens exceed a threshold that has been shown to be associated with some disability. For each species there are two populations considered to be at risk, based on two estimates of threshold burden: a larger population with worm burdens exceeding the lower threshold (associated with chronic or developmental disability); and a smaller population with burdens exceeding the higher threshold (associated with acute or contemporaneous disability). The proportion of the population at risk (not not the prevalence of infection) is given in Tables 9.9 to 9.11. These estimates have been recalculated and differ from the original estimates (World Bank 1993, Chan et al. 1994). For A. lumbricoides and T. trichiura they are substantially lower than the earlier estimates as a result of removing the highest prevalence reference class from the extrapolation analysis. This change avoids overemphasis on the highest prevalence classes, which contribute disproportionately to the at risk population, and so gives a more conservative estimate. For the hookworm estimates, there was the same change in procedure but also a change in the worm burden thresholds. The change in thresholds was necessary to correctly assign the thresholds to the appropriate age classes in the light of new information. The effects of these changes for hookworm infection are to substantially reduce the estimated size of the population above the higher threshold and to slightly increase the population above the lower threshold. Contemporaneous effects of infection Given that people in endemic areas are continuously infected and reinfected throughout life it can be assumed, for present purposes, that incidence is numerically equivalent to prevalence and that infection duration is one year. The contemporaneous effects are assumed to occur in 100 per cent of the population with worm burdens above the higher threshold, and persist as long as an individual remains infected. With A. lumbricoides, the most common consequences of infection are insidious effects, which are often manifested as effects on development 268 Global Epidemiology of Infectious Diseases (reviewed by Crompton, Nesheim & Pawlowski 1989). They are, however, contemporaneous effects in that they can be partially reversed on treatment; that is, they occur only while infection persists. Such effects include reduction in growth rate (height-for-age and weight-for-age), physical fitness and appetite, for school-age and younger children (Stephenson et al. 1989, 1990, 1993). There is also evidence that this infection has consequences for congnitive ability in school-age children. Cognitive consequences have yet to be sought in adults, but it would be surprising if adults responded differently from children since the effect is on ability rather than development. We assume here that adults with above-threshold worm burdens are affected, although the proportion of adults in this category is very small (0.02 per cent). Table 9.9 Data for DALY calculation, Ascariasis, both sexes, 1990 Proportion at risk Region Age Group Acute Permanent Average age at onset Duration of risk Duration of disability Acute Permanent SubSaharan Africa 0–4 5–14 15–44 45–59 60+ 0 525 10 10 10 120 1 720 170 170 170 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 Latin America & Caribbean 0–4 5–14 15–44 45–59 60+ 10 1 790 20 20 20 410 5 950 580 580 580 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 Middle Eastern Crescent 0–4 5–14 15–44 45–59 60+ 0 555 10 10 10 120 1 815 180 180 180 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 India 0–4 5–14 15–44 45–59 60+ 0 770 10 10 10 170 2 530 250 250 250 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 China 0–4 5–14 15–44 45–59 60+ 20 2 810 40 40 40 600 8 870 890 890 890 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 Other Asia & Islands 0–4 5–14 15–44 45–59 60+ 10 2 345 30 30 30 540 7 850 770 770 770 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 Intestinal nematode infections 269 The disabling consequences of reduced physical fitness and cognitive ability have yet to be empirically quantified, as is the case for many of the morbid effects for which DALY estimation is attempted. We assume here, by default, that the contemporaneous effects of moderate ascariasis result in disability at the lowest disability weight (Class 1) (Murray & Lopez 1996). There are also more serious consequences of infection, largely associated with obstruction of ducts and intestinal lumen by these large worms. Systematic data on these acute complications are lacking, but the numerous reports based on inpatient records suggest that ascariasis is an important cause of hospitalization in endemic areas (reviewed by Pawlowski & Davies 1989). Ascariasis was the cause of 2.6 per cent of all hospital admissions in Kenya in 1976, and 3 per cent in a children’s hospital in Myanmar between 1981 and 1983 (Stephenson, Lathan & Oduori 1980, Thein-Hlaing 1987). Complications resulting from ascariasis accounted for 0.6 per cent of all admissions to a paediatric surgery department in South Africa in 1987, 5.8 per cent of emergency admissions to a hospital in Mexico in 1975, 10.6 per cent of admissions for acute abdominal emergency to a children’s hospital in Myanmar, and between 0.8 and 2.5 per cent of admissions in a survey of hospitals in China (Flores & Reynaga 1978, World Health Organization 1987, Thein-Hlaing et al. 1990). The most common abdominal emergencies presenting are intestinal obstruction and biliary ascariasis, the proportions varying geographically, perhaps because of differences in diagnostic procedures (Maki 1972). The classical surgical presentation is in patients between 3 and 10 years of age, although adults also may be affected (Davies and Rode 1982, Chai et al. 1991). Laparotomy attributable to ascariasis was the second most common cause of all laparotomies in 2–4 year old children in Durban, Lishiu and Sao Paulo, and the fifth or sixth cause in adults in Myanmar, China and Nigeria (World Health Organization 1987). Reports using unstandardized indicators indicate that between 0.02 and 0.9 per cent of infections may require hospitalization (Pawlowski & Davies 1989), the proportion presumably varying with the local intensity of infection. Thus of the 4.5 per cent of infected children under 10 years of age who are here estimated to exceed the higher threshold of infection, between 0.4 and 20 per cent are likely to require hospitalization. These children will suffer a severely disabling condition, which may be life-threatening (see below), but which can be alleviated by appropriate clinical management. If it is assumed that such cases are managed appropriately, then the duration of disability is likely to be a few weeks. Complicated ascariasis has a reported history of over 10 days followed by 5 days of management, while the management of biliary ascariasis involves 4–6 weeks of observation before opting for surgical intervention (Davies & Rode 1982). The disability therefore is considered, for the present analyses, to be contemporaneous with infection, to have a duration of 4 weeks, to have a severity of Class III (Murray & Lopez 1996), and to affect 5 per cent of children under 15 years 270 Global Epidemiology of Infectious Diseases Table 9.10 Data for DALY calculation, Trichuriasis, both sexes, 1990 Proportion at risk Region Age Group Sub-Saharan 0–4 Africa 5–14 15–44 45–59 60+ Average age Duration at onset of risk Acute Permanent 0 680 0 0 0 0 1 340 110 110 110 2 10 30 50 70 Duration of disability Acute Permanent 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 Latin America & Caribbean 0–4 5–14 15–44 45–59 60+ 0 2 855 10 10 10 10 5 795 470 470 470 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 Middle Eastern Crescent 0–4 5–14 15–44 45–59 60+ 0 0 0 0 0 0 30 0 0 0 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 India 0–4 5–14 15–44 45–59 60+ 0 620 0 0 0 0 1 240 100 100 100 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 China 0–4 5–14 15–44 45–59 60+ 0 3 380 20 20 20 10 6 475 570 570 570 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 Other Asia & Islands 0–4 5–14 15–44 45–59 60+ 0 3 980 20 20 20 20 8 050 650 650 650 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 81.3 71.7 51.9 32.5 14.9 of age with burdens exceeding the higher threshold. Note that this is a conservative assumption since it excludes the documented, though rare, occurrence of complications in adults. Furthermore, the case rates for children are based on records from tertiary facilities to which a substantial proportion of the most disadvantaged and heavily infected children may have limited access. With T. trichiura there is evidence that moderate intensity infections result in growth deficits that can be reversed by the antihelminthic removal of the worms (Cooper et al. 1990), and that these infections result in a protein losing enteropathy and anaemia (Cooper et al. 1991, 1992, MacDonald et al. 1991, Ramdath et al. 1995). In young children there are also Intestinal nematode infections 271 Table 9.11a Data for DALY calculation, Hookworm disease, males, 1990 Proportion at risk Region Average age Duration at onset of risk Age Group Acute Permanent Sub Saharan Africa 0–4 5–14 15–44 45–59 60+ 0 180 2 650 2 650 2 650 0 1 260 5 760 5 760 5 60 2 10 30 50 70 Latin America & Caribbean 0–4 5–14 15–44 45–59 60+ 0 30 1 090 1 090 1 090 0 945 5 030 5 030 5 030 Middle Eastern Crescent 0–4 5–14 15–44 45–59 60+ 0 30 770 770 770 India 0–4 5–14 15–44 45–59 60+ China Other Asia & Islands Duration of disability Acute Permanent 1 1 1 1 1 1 1 1 1 1 80.0 70.4 50.5 31.0 13.6 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 80.0 70.4 50.5 31.0 13.6 0 480 2 400 2 400 2 400 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 80.0 70.4 50.5 31.0 13.6 0 70 2 070 2 070 2 070 0 1575 8150 8 150 8 150 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 80.0 70.4 50.5 31.0 13.6 0–4 5–14 15–44 45–59 60+ 0 10 330 330 330 0 540 3 260 3 260 3 260 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 80.0 70.4 50.5 31.0 13.6 0–4 5–14 15–44 45–59 60+ 0 70 2 060 2 060 2 060 0 1545 7 980 7 980 7 980 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 80.0 70.4 50.5 31.0 13.6 effects on development quotient (Griffiths locomotor subscale), anaemia and growth that are at least partially reversible by antihelmintic therapy (Callender et al. 1994). There is also an increasing body of evidence that both cognitive function (Tables 9.12a and 9.12b) and educational achievement are impaired by moderate intensity infections, and that at least some of these effects can be reversed by antihelmintic treatment (Nokes et al. 1992a, 1992b, Simeon, Grantham-McGregor & Wong 1995, Simeon et al. 1995). As for ascariasis, it is assumed that these are contemporaneous effects of trichuriasis and result in the lowest disability weight (Class 1). Particularly large burdens of T. trichiura may result in the “classical” dysenteric form of trichuriasis, synonymous with Trichuris Dysentery Syn- 272 Global Epidemiology of Infectious Diseases Table 9.11b Data for DALY calculation, Hookworm disease, females, 1990 Proportion at risk Region Age Group Average age Duration at onset of risk Acute Permanent Sub Saharan 0–4 Africa 5–14 15–44 45–59 60+ 0 180 2 650 2 650 2 650 0 1 260 5 760 5 760 5 760 2 10 30 50 70 Latin America & Caribbean 0–4 5–14 15–44 45–59 60+ 0 30 1 090 1 090 1 090 0 945 5 030 5 030 5 030 Middle Eastern Crescent 0–4 5–14 15–44 45–59 60+ 0 30 770 770 770 India 0–4 5–14 15–44 45–59 60+ China Other Asia & Islands Duration of disability Acute Permanent 1 1 1 1 1 1 1 1 1 1 82.5 73.0 53.3 34.0 16.2 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 82.5 73.0 53.3 34.0 16.2 0 480 2 400 2 400 2 400 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 82.5 73.0 53.3 34.0 16.2 0 70 2 070 2 070 2 070 0 1 575 8 150 8 150 8 150 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 82.5 73.0 53.3 34.0 16.2 0–4 5–14 15–44 45–59 60+ 0 10 330 330 330 0 540 3 260 3 260 3 260 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 82.5 73.0 53.3 34.0 16.2 0–4 5–14 15–44 45–59 60+ 0 70 2 060 2 060 2 060 0 1 545 7 980 7 980 7 980 2 10 30 50 70 1 1 1 1 1 1 1 1 1 1 82.5 73.0 53.3 34.0 16.2 drome (Ramsey 1962) and Massive Infantile Trichuriasis (Kouri & Valdes Diaz 1952). This typically occurs in children between 3 and 10 years of age and is associated with burdens involving at least several hundreds of worms carpeting the colonic mucosa from ileum to rectum. The colon is inflamed, oedematous and friable, and often bleeds freely (Venugopal et al. 1987). Reviews of case histories suggest that the mean duration of disease at the time of presentation is typically in excess of 12 months and that relapse after treatment frequently occurs (Gilman et al. 1983, Cooper et al. 1990, Callender et al. 1994). The probability of relapse, and of a child experiencing multiple episodes, is greatly enhanced because a proportion of heavily infected children are predisposed to reacquire heavy infection Intestinal nematode infections 273 even after successful treatment (Bundy et al. 1987a, 1987b). The typical signs of the syndrome (see Bundy & Cooper 1989a for a review of 13 studies involving 697 patients) are rectal prolapse, tenesmus, bloody mucoid stools (over months or years), growth stunting, and a profound anaemia, which may lead to a secondary anaemia. The complete spectrum of clinical features associated with the syndrome occurs in some 30 per cent of children with intense trichuriasis. Many of the major clinical effects are reversible by appropriate therapy (Cooper, Bundy & Henry 1986, Gilman et al. 1983), hence the disability is considered here to be a contemporaneous consequence of infection. For the present analyses it is assumed that the disability is contemporaneous with infection, has a duration of over 12 months, has a severity of Class II (Murray & Lopez, 1996), and affects 20 per cent of children under 15 years of age experiencing the higher threshold of intensity. With hookworm the major consequence of infection is anaemia (see Schad & Banwell 1984 and Crompton & Stephenson 1990 for reviews of the extensive literature in this area). Anaemia is associated with: reduced worker productivity; reduced adult and child fitness; reduced fertility in women; reduced intrauterine growth rate, prematurity and low birth weight; and cognitive deficits (Fleming 1982, Stephenson et al.1993, Pollitt et al. 1986, Boivin et al. 1993) (Tables 9.12a and 9.12b). Since the higher threshold for the intensity of hookworm infection was selected on the basis of the development of anaemia, it is here assumed that 100 per cent of those exceeding this threshold suffer at least Class I disability. As discussed elsewhere (World Bank 1993), the consequences of anaemia will be more serious for a subset of the affected population, resulting in Class II and Class III disability. The disability weight distribution for anaemia was used for the present analyses, 70 per cent in Class II, 24 per cent in Class III and 6 per cent in Class IV. For all species it is assumed here that the effects are independent of host sex. It is also assumed that the disability weight is age-independent since the method of extrapolating the population at risk incorporates ageweights in both the prevalence of infection and the threshold associated with disability by age. Chronic effects of infection In addition to the contemporaneous effects of infection, there is evidence that some consequences of infection are irreversible. This is the case for some cognitive deficits (Table 9.12a), some elements of development quotient (Callendar et al. 1994), and some growth effects during childhood (Stephenson et al. 1993). In all studies, some forms of disability in a proportion of children do not respond to therapy. We estimate that in any annual cohort of heavily infected children some 5 per cent suffer these permanent consequences. Studies of reinfection show that children are predisposed to a particular intensity of infection (Keymer & Pagel 1990, Hall, Anwar & Tomkins 1992), such that some 30 per cent of heavily infected children Tests in battery = 9 x Pegboard non-dominant hand x Analogical reasoning C = 63 Albendazole x Digit-span backwards x Digit-span forwards P = 67 Placebo controlled 9–11 T = 66 Intervention 10 weeks Tests in battery = 8 C = 48 Albendazole * Silly sentences x Letter search 10–11 P = 48 Placebo controlled 14 weeks T = 49 Intervention Tests in battery = 6 * Letter search C = 100 Albendazole x Fluency x French learning (initial) 8–10 P = 93 14 weeks T = 96 Placebo controlled Tests in battery = 4 x School Attendance * Arithmetic * Reading C = 206 7–11 x Spelling 26 weeks Baseline differences between infected and uninfected group None significant None significant Underweight children improved in fluency. Treatment interaction effects: • stunted and received treatment improved in school attendance. • heavily infected and received treatment improved in spelling; Treatment interaction effects: Intervention differences Significant effects of helminth infection on mental processing P = 201 T = 206 Sample size Age of subjects (years) Intervention Albendazole Placebo controlled Intervention Jamaica Trichuris trichiura Moderate intensity Study design Country Parasite Time between baseline and follow-up (if appropriate) Table 9.12a Selected studies investigating the effects of parasitic helminth infections on mental processing Sternberg, Powell & McGrane, 1995 Baddeley, Meeks-Gardner & GranthamMcGregor 1995 Simeon et al. 1994 Simeon et al. 1994 Investigators 274 Global Epidemiology of Infectious Diseases presence + Iron status Hookworm presence Hookworm Moderate– heavy intensity Trichuris trichiura Zaire Zaire Jamaica Levamizole iron supplements Intervention, placebo controlled Levamizole + malaria treatment if infected only Case control Intervention, Albendazole Placebo controlled Intervention P + Iron; T only; P only; = 30 T + Iron = 17 C = 50 T = 47 4 weeks 4 weeks Fem = 8.0 Male = 7.7 8.75 ± 1.6 Tests in battery = 10 Not reported Tests in battery = 10 * Mental processing total Word order Sequential processing, including number recall Tests in battery = 8 Matching familiar figures: hard Matching familiar figures: easy Comprehension Coding Arithmetic Digit-span backwards Fluency C = 56 9–12 Digit-span forwards 63 ± 8 days P = 41 T = 62 Simultaneous total Sequential total Mental processing total Matrix analogies Gestalt Face recognition Word order Number recall Spatial memory Effect = iron + hookworm trnt Spatial memory Digit-span backwards Digit-span forwards Fluency 1993 continued Boivin & Giordani Boivin et al. 1993 Nokes et al.1992a, 1992b Intestinal nematode infections 275 presence Ascaris lumbricoides (Study actually designed to investigate T. trichiura) Presence Ascaris lumbricoides Ecuador Jamaica Jamaica Trichuris trichiura TDS – very heavy intensity Country Parasite Cross-section Cross-section, matched for class Albendazole given to infected children every 4 months T = 19 Case control, interventions pair-matched Uninfd C = 27 Infd = 103 (Infd with T. trichiura = 409) Uninfd = 207 Infd = 196 C = 19 Sample size Study design — — 12 months Time between baseline and follow-up (if appropriate) 9–13 7–11 3–6 Age of subjects (years) Interaction with nutrition in * Wisconsin Card Sorting Test * Digit Symbol * Peabody raw score * Stroop words Tests in battery = 4 * School attendance * Arithmetic * Spelling * Reading Performance Hearing and speech Eye hand coordination Locomotor Griffiths DQ subscales Baseline differences between infected and uninfected group Not done Locomotor Intervention differences Significant effects of helminth infection on mental processing Table 9.12a Selected studies investigating the effects of parasitic helminth infections on mental processing (continued) Levav et al. 1995 Simeon et al. 1994 Callender et al. 1991 Investigators 276 Global Epidemiology of Infectious Diseases intensity Low-Moderate and heavy Trichuris trichiura Intensity A. lumbricoides, T. trichiura, hookworm, Schistosome spp. Protozoa Polyparasitism intensity A. lumbricoides, T. trichiura, Hookworm Polyparasitism Italy South Africa Jamaica Cross-section all infected with different spp. Cross-section Cross-section Infd = 124 Uninf = 232 Subjects = 110 Uninfd = 170 Infd = 473 — — — 6–10 10 9–11 Other outcome = promotion. x Lower school grade correlated with intensity. No relationship when controlled for social and hygienic conditions. Other test = Memory (pathogenicity of parasitic species more important than prevalence/ intensity at predicting performance) * Sustained attention Only test in battery Academic stream (children streamed by teachers according to academic ability; children with least ability more likely to be infected and heavily infected) Tests in battery = 13 All sample had high EEG * = controlled for nutrition status Peabody Test Not done (Intervention disrupted by flooding) Not done Not done continued de Carneri 1968 Kvalsvig Cooppan & Connolly 1991 Nokes et al. 1991 Intestinal nematode infections 277 presence Hookworm Intensity Hookworm Intensity United States of America Australia Cross-section Cross-section Cross-section Italy Trichuris trichiura Low–Moderate & heavy Study design Country Parasite 78 Infected: high intensity= 159 Infected: low intensity = 65 Uninfd = 116 Infd = 233 Uninf = 125 Sample size — — — Time between baseline and follow-up (if appropriate) 6–17 6.5– 15.5 6–11 Age of subjects (years) * Grade advancement (deficit of 0.23 grades/ year) Binet-Simon test and Porteus Mazes correlated with intensity of infection. Other outcome = absenteeism No relationship when controlled for social and hygienic conditions. x Slower promotion, poor grades correlated with intensity. Baseline differences between infected and uninfected group Not done Not done Not done Intervention differences Significant effects of helminth infection on mental processing Table 9.12a Selected studies investigating the effects of parasitic helminth infections on mental processing (continued) Stiles 1915 Waite & Neilson 1919 de Carneri Garofano & Grassi 1967 Investigators 278 Global Epidemiology of Infectious Diseases Definitions Intervention Placebo controlled Case control Pair-matched Cross-section Presence Intensity Light, moderate or heavy intensity presence Hookworm A. lumbricoides T. trichiura Polyparasitism Presence Hookworm A. lumbricoides Harvard Step Test Albendazole Intervention placebo controlled Harvard Step Test Albendazole Intervention placebo controlled P=26 T=27 P=15 T=18 Sample size 4 months 7 weeks 7–13 6–12 No difference between infected groups. Fitness negatively correlated with haemoglobin/ iron status. Age of Baseline differences subjects between infected and (years) uninfected group Stephenson et al.1993 Stephenson et al.1990 No change in fitness score and pulse rate in placebo group. Significant improvement in fitness in treatment group. Improvement significantly related to reduction in hookworm and A. lumbricoides egg counts. No change in fitness score and pulse rate in placebo group. Significant improvement in fitness in treatment group. Improvement significantly related to reduction in hookworm, weight gain and increase in energy intake during the 4 month study period. Investigators Intervention differences continued One group receives treatment. Groups tested pre-intervention and post-intervention. Infected group randomly assigned to treatment or placebo. Groups tested pre-intervention and post-intervention. Infected group compared to an uninfected group. Groups matched for age as a minimum. Infected group receives treatment. Uninfected group receives no treatment. Infected group and uninfected group pair matched for confounding variables other than just for age. Infected group compared against uninfected group. Neither group receives treatment. Groups tested at baseline only. Subjects selected for the study on the basis of the presence of infection and not on the basis of the intensity of infection. Analysis of results takes into consideration the intensity of infection. The intensity of infection of subjects recruited to the study. Kenya Kenya Polyparasitism T. trichiura Country Parasite Study design and Measure of activity Time between baseline and follow-up (if appropriate) Table 9.12b Selected studies investigating the effects of parasitic helminth infections on physical fitness Intestinal nematode infections 279 Key X Controlled for confounding variables and infection did not remain significant. * Controlled for confounding variables and infection remained significant. No mark No attempt to control for confounding variables. Abbreviations T Infected group treated with antihelminthic. P Infected group treated with antihelminthic placebo. NoT Infected group receiving no intervention, i.e. no treatment or no placebo. C Uninfected control receiving no treatment or placebo. Infd Infected with parasite spp. Uninfd Uninfected with parasite spp. Table 9.12b Selected studies investigating the effects of parasitic helminth infections on physical fitness (continued) 280 Global Epidemiology of Infectious Diseases Intestinal nematode infections 281 in an annual cohort would be expected to reacquire heavy infection. Thus each year the proportion of children exceeding the threshold worm burden will consist of some 70 per cent of individuals who have not previously experienced heavy infection, which implies a cumulative increase in the proportion suffering permanent disability. We therefore assume that each year 3 per cent of newly heavily infected children, and children only, suffer life-long consequences of infection. The disability attributable to these effects has yet to be empirically determined. Stunted children may be disadvantaged in education (Moock & Leslie 1986, Jamison 1986, Glewwe & Jacoby 1995), as are children with low development quotients or cognitive impairment (Pollitt 1990). On the other hand, physical and mental maturation may eventually compensate, to some degree, for initial retardation (Pollitt et al. 1986). In a recent study of two years nutritional supplementation of stunted children, locomotor development improved in the first year, as seen with antihelminthic treatment of trichuriasis (Callender et al. 1994), while other areas of development did not improve until the second year (Grantham-McGregor et al. 1991). Given this uncertainty we assume that the permanent consequences of infection result in disability at the lowest weight (Class 1). Thus in calculating the DALYs for all the helminth infections it is assumed here that 3 per cent of children experiencing worm burdens above the lower threshold suffer permanent disability of Class 1. Mortality This is the weakest area of DALY estimation because of the lack of empirical data. Ascariasis is the best documented helminthiasis in terms of mortality. There are numerous studies of case fatality rates in hospitals (reviewed by Pawlowski & Davies 1989). These indicate that the outcome of acute complications of ascariasis is modified by the general health status of the patient, the intensity of infection and the medical procedure (see Pawlowski & Davies 1989). In one hospital in Sao Paulo, for example, the case fatality rate was 1.35 per cent for conditions which could be managed conservatively, and 26.1 per cent in patients undergoing surgery (Okumura et al. 1974). These studies confirm that death is a not infrequent outcome of complications of ascariasis, but provide little insight into mortality rates in the community. An extrapolation from central hospital data in Myanmar suggests there are 0.008 deaths per 1000 infections per year (Thein-Hliang 1987), but this is considered to be a considerable underestimate since only a small proportion of children with severe complications are likely to have access to the hospital (Pawlowski & Davies 1989). Only two population based estimates are available: one for the Darmstadt epidemic (0.1 deaths per 1000 infected per year) (Krey 1949) and one for Japan prior to national control efforts (0.061 deaths per 1000 infected per year) (Yokogawa 1976). Based on the present estimate of global infection, these rates suggest that between 8 000 and 14 000 children die each year. The final estimate of global mortality due to ascariasis in the Global 282 Global Epidemiology of Infectious Diseases Burden of Disease project was 11 000 concentrated in the age group 5–14 years (Murray and Lopez 1996). Mortality was distributed by region in proportion to the region-specific population at risk (above higher threshold) and to the region-specific probability of dying between birth and 4 years (World Bank 1993). This last quantity was added to take account of regional variations in access to acute medical services. No population-based mortality estimates have been published for T. trichiura infection. Prior to the advent of safe and effective therapy for T. trichiura infection in the late 1970s a number of reports described paediatric inpatients with Trichuris Dysentery Syndrome who, despite clinical efforts, died as a result of profuse haemorrhage and secondary anaemia (Wong and Tan 1961, Fisher and Cremin 1970) or of intussusception (Reeder, Astacio & Theros 1968). Although there continue to be reports of the syndrome, a fatal outcome in a clinical setting today would suggest inappropriate management. The picture in the community, however, may be rather different since, in the absence of specific diagnosis, the aetiology of chronic bloody dysentery may be unrecognized. Nevertheless, mortality is undoubtedly a rare consequence of trichuriasis. The profound anaemia of hookworm infection is life-threatening and has been estimated, although the means of estimation are not described, to result in 65 000 deaths per year (World Health Organization 1992). Again there is a lack of empirical data, presumably in this case because of the difficulty in identifying the etiology of anaemia-related deaths. A figure of 4 300 deaths was used by Murray and Lopez (1996) and was distributed to ascribe the highest proportion of mortality to women of childbearing age (15–44 years) and to older age groups. The distribution of deaths between regions was divided in the same way as for the other infections. In including these estimates of mortality we recognize that they are unsupported by vital registration statistics. But it should also be recognized that intense infection is most prevalent in the poorest regions of the poorest countries. In such areas mortality may be most likely because of limited access to appropriate management, while both the diagnosis of cause of death and its registration may be least reliable. There is clear evidence that deaths do occur. What is unclear is the extent of this mortality. Disability Much of the disability associated with helminth infection is insidious and would be unlikely to be brought to clinical attention. As such it is difficult to compare with the more classical clinical signs. On the other hand, cognitive deficits may have profound consequences for educational outcomes and growth stunting is one of the best characterized correlates with underachievement, so these insidious effects may have far reaching societal consequences. Achieving some realistic balance between the clinical consequences for the individual and developmental consequences for Intestinal nematode infections 283 society goes beyond the scope of the present exercise, and would require a more sophisticated weighting system, if indeed the effects could be quantified. For present purposes, the very low proportional weighting selected for the chronic effects of infection is influenced by the view that the Class 1 disability weight may be an overestimate of the effects of infection from a clinical perspective. Other considerations in calculating burden There are three main reasons why the burden of intestinal helminths may not be adequately captured by the present calculations. First, there are no direct measures of morbidity against which the extrapolation procedure could be conclusively validated. Although each step in the extrapolation was independently assessed against empirical data as far as possible, there must remain uncertainty until observed data become available. Second, the mortality data are largely unsubstantiated. Mortality has the potential to significantly alter the overall burden estimates. It is therefore unsatisfactory that mortality data have received so little research attention. Third, there is a particular lack of information on the morbid consequences of infection in young children. It is possible that even very low worm burdens may have disproportionately severe effects on developing physiologies and organ systems. Anecdotal evidence suggests that the current estimates of threshold and disability weights may significantly underestimate the burden in children under 10 years of age, and particularly those under 5 years. The societal consequences of growth stunting and educational underachievement may be of substantially greater relevance than the disability in the individual. Intestinal nematode infections as risk factors for other diseases Intestinal nematode infection is associated with malabsorption and is a potentially important predisposing factor for malnutrition in communities on marginal diets. These effects may relate broadly to protein-energy malnutrition, or to specific deficiencies. For example, A. lumbricoides infection has been associated with malabsorption of vitamin A. Hookworm infection and to a lesser extent trichuriasis are associated with iron loss predisposing to anaemia. It is self evident that the risk of anaemia is dependent on iron balance and thus that infection may be an important contributing factor. The attributable contribution to global malnutrition is potentially considerable given the ubiquity of infection and its specifically high prevalence in the poorest societies with the least adequate diets. 284 Global Epidemiology of Infectious Diseases Burden and intervention The major intestinal helminth infections can be effectively treated simultaneously with single dose oral therapy. The treatment is widely available, safe, simple and inexpensive. Prevention of reinfection requires reduction in transmission, which can be achieved by synchronized treatment programmes and by improvements in sanitation. In the absence of currently financed health interventions there would be some increase in the current burden. For example, there would be a greater number of deaths from intestinal obstruction in the absence of operative procedures, and from severe anaemia or malnutrition in the absence of rehabilitation therapy. However, with some important exceptions such as Japan and the Republic of Korea, control of intestinal helminth infections is only rarely a component of national public health programmes. It could be argued that the entire burden could eventually be avoided by appropriate application of currently available interventions. For example, evidence from the Republic of Korea and Japan indicates that reduction in the prevalence of A. lumbricoides infection at the national level results in a significant decline in acute complications of ascariasis requiring hospitalization (Chai et al. 1991) and in ascariasis related mortality (Yokogawa 1976). Curative treatment would mitigate the contemporaneous or acute effects of infection, while measures to control transmission would avoid chronic developmental disability, although neither could reverse the deficits that are already present in the population. Economic analyses suggest that carefully targeted community treatment programmes are exceptionally cost-effective (Warren et al 1993, Guyatt & Evans 1992, World Bank 1993). This arises because the therapy is inexpensive (US$0.20 or less per dose), is required at infrequent intervals (of the order of one year), can have community-wide effects even if targeted at some fraction of the population which makes the greatest contribution to transmission (such as school-age children), and can be delivered through existing infrastructures (such as schools or the primary health care system). Conclusions The global morbidity attributable to intestinal nematode infections, although generally accepted to be large, has proven difficult to quantify. The method presented here provides a framework whereby the potential global burden may be estimated in the absence of any direct measures of morbidity. The estimates are intended to give some indication of the potential burden of intestinal helminthiases rather than to provide absolute values. It would of course be possible to seek further refinement of the approach, but our view is that the most pressing need is to obtain reliable community data on the observed levels of morbidity and on the consequences of disability. The present analyses indicate that even low Intestinal nematode infections 285 levels of individual disability can sum to a considerable burden with such ubiquitous infections; the important question is what this implies for communities in practice. The analysis has revealed important lacunae in our knowledge of these infections and it is hoped that this might guide future applied research. The present estimates suggest that the potential morbidity attributable to geohelminthiases is much greater than previously supposed. This reflects the inclusion of both the traditionally recognized clinical effects of helminthiases (see World Health Organization 1992) and more recently recognized developmental effects, which rarely result in clinical presentation but which may have major consequences for the individual and the community. Another general implication of the results arises from the similarity of the age and regional distributions for A. lumbricoides and T. trichiura infection and morbidity. This observation has been made before for prevalence data (Booth & Bundy 1992), and strong positive correlations between A. lumbricoides and T. trichiura prevalences in communities have been demonstrated. This suggests that these two infections could be controlled within a single programme. The age distribution of the burden also supports the conclusion that there are particular benefits in targeting control of A. lumbricoides and T. trichiura infection at school age children (Bundy et al. 1985, Savioli, Bundy & Tomkins 1992). The results suggest that the vast majority of the morbidity attributable to intestinal nematodes is readily avoidable or reversible using existing and cost-effective approaches, and that the mortality is also a largely avoidable consequence of acute infection. These observations, and the scale of the burden of current disease, argue for greater public health emphasis on the control of intestinal helminthiases. References Aguilar FJ (1981) Helmintiasis humanas transmitidas a traves del suelo en Guatemala [Soil-transmitted human helminthiasis in Guatemala]. Boletin Chileno de Parasitologia, 36:6–9. Anderson RM (1982) Transmission dynamics and control of infectious disease agents. In: Anderson RM, May RM eds., Population biology of infectious diseases. Berlin, Springer, pp. 149–76. Anderson RM (1986) The population dynamics and epidemiology of intestinal nematode infections. Transactions of the Royal Society of Tropical Medicine and Hygiene, 80:686–696. Anderson RM, May RM (1982) Population dynamics of human helminth infections: control by chemotherapy. Nature, 297:557–563. Anderson RM, May RM (1985) Helminth infections of humans: mathematical models, population dynamics, and control. Advances in Parasitology, 24:1–101. 286 Global Epidemiology of Infectious Diseases Anderson RM, May RM (1991) Infectious diseases of humans: dynamics and control. Oxford, Oxford University Press. Anderson RM, Medley GF (1985) Community control of helminth infections of man by mass and selective chemotherapy. Parasitology, 90:629–660. Annan A et al. (1986) An investigation of the prevalence of intestinal parasites in pre-school children in Ghana. Parasitology, 92:209–217. Arora RR et al. (1975) Prevalence of intestinal parasites in the village Baskripal Nagar, District Alwar, Rajasthan. Journal of Communicable Diseases, 7:264–268. Ashford RW, Hall AJ, Babona D (1981) Distribution and abundance of intestinal helminths in man in western Papua New Guinea with special reference to Strongyloides. Annals of Tropical Medicine and Parasitology, 75:269–279. Asin HRG, van Thiel PH (1963) On the distribution of intestinal helminths in the urban and bushland population in Surinam. Tropical and Geographical Medicine, 15:257–267. Baddeley A, Meeks-Gardner J, Grantham-McGregor S (1995) Cross-cultural cognition: developing tests for developing countries. Unpublished. Bagchi SC, Prasad BG, Mathur GB (1964) A survey of hookworm infection in the area of rural health training centre, Sarojini Nagar, Lucknow. Indian Journal of Medical Research, 52:411–417. Barbosa JCL (1956) Contribucao para o conhecimento do estado de nutricao da populacao de Cabo Verde. IV – Inquerito realizado em S. Nicolau em 1954 [Contribution to the knowledge of the nutritional status of the population of Cape Verde, study carried out in S Nicolau in 1954]. Anais do Instituto de Medicina Tropical, 13:689–809. Behnke JM (1987) Do hookworms elicit protective immunity in man? Parasitology Today, 3:200–206. Bergner JF (1964) Intestinal parasites in an Aborigine village in Southeast Taiwan. American Journal of Tropical Medicine and Hygiene, 13:78–81. Bianchini C et al. (1981) Reperti parassitologici in corso di una epidemia di enterocolite in un villaggio del distretto di Jowhar (Somalia) [Parasitological reports collected in the course of an epidemic of Enterocolitis in a village of the Jowhar district (Somalia)]. Parassitologia, 23:122–124. Bidinger PD, Crompton DWT, Arnold S (1981) Aspects of intestinal parasitism in villagers from rural peninsular India. Parasitology, 83:373–380. Blackie WK (1932) A helminthological survey of Southern Rhodesia, London, London School of Hygiene and Tropical Medicine. Boivin MJ, Giordani B (1993) Improvements in cognitive performance for schoolchildren in Zaire, Africa, following an iron supplement and treatment for intestinal parasites. Journal of Pediatric Psychology, 18:249–264. Boivin MJ et al. (1993) Effects of treatment for intestinal parasites and malaria on the cognitive abilities of schoolchildren in Zaire, Africa. Health Psychology, 12:220–226. Bonnefoy X, Isautier H (1978) Variation de l’incidence des helminthiases a la Reunion en fonction de différents paramètres [Variation in the incidence of Intestinal nematode infections 287 worm infection in Réunion in relation to various parameters]. Bulletin de la Société de Pathologie Exotique et de ses Filiales, 71:70–78. Booth M (1994) The epidemiology and population biology of multiple infections with ascaris lumbricoides, Trichuris trichiura and hookworms. University of London (unpublished doctoral thesis). Booth M, Bundy DAP (1992) Comparative prevalences of Ascaris lumbricoides, Trichuris trichiura and hookworm infections and the prospects for combined control. Parasitology, 105:151–157. Bradley M et al. (1992) The epidemiology and population biology of Necator americanus infection in a rural community in Zimbabwe. Transactions of the Royal Society of Tropical Medicine and Hygiene, 86:73–76. Bradley M, Chandiwana SK (1990) Age-dependency in predisposition to hookworm infection in the Burma valley area of Zimbabwe. Transactions of the Royal Society of Tropical Medicine and Hygiene, 84:826–828. Brumpt LC, Sang HT, Jaeger G (1972) Quelques reflexions a propos du parasitisme sanguin et intestinal dans deux villages d’Afrique Centrale. Bulletin de la Société de Pathologie Exotique et de ses Filiales, 65(2):263–270. Bundy DAP (1988) Population ecology of intestinal helminth infections in human communities. Philosophical Transactions of the Royal Society of London, Series B, 321:405–420. Bundy DAP (1990) Is the hookworm just another geohelminth? In: Schad GA, Warren KS eds., Hookworm disease, current status and new directions. London, Taylor and Francis, pp. 147–164 (Chapter 10). Bundy DAP et al. (1985) Population distribution of Trichuris trichiura in a community of Jamaican children. Transactions of the Royal Society of Tropical Medicine and Hygiene, 79:232–237. Bundy DAP et al. (1987a) Epidemiology and population dynamics of Ascaris lumbricoides and Trichuris trichiura in the same community. Transactions of the Royal Society of Tropical Medicine and Hygiene, 81:987–993. Bundy DAP et al. (1987b) Age-related prevalence and intensity of Trichuris trichiura infection in a St. Lucian community. Transactions of the Royal Society of Tropical Medicine and Hygiene, 88:85–94. Bundy DAP et al. (1988) Effect of age and initial infection intensity on the rate of reinfection with Trichuris trichiura after treatment. Parasitology, 97:469–476. Bundy DAP et al. (1991) The epidemiological implications of a multiple-infection approach to the control of human helminth infections. Transactions of the Royal Society of Tropical Medicine and Hygiene, 85:274–276. Bundy DAP, Cooper ES (1988) The evidence for predisposition to trichuriasis in humans: comparison of institutional and community studies. Annals of Tropical Medicine and Parasitology, 82:251–256. Bundy DAP, Cooper ES (1989) Trichuris and trichuriasis in humans. Advances in Parasitology, 28:107–173. Bundy DAP, Guyatt HL (1990) Global distribution of parasitic worm infections. Report to the Division of Science, Technical and Environmental Education. 288 Global Epidemiology of Infectious Diseases UNESCO Project to Increase School Performance through Improved Nutrition and Health. Bundy DAP, Medley GF (1992) Immuno-epidemiology of human geohelminthiasis: ecological and immunological determinants of worm burden. Parasitology, 104:S105–S119. Burgess HJ, Burgess A, Wheeler EF (1973) Results and appraisal of a nutrition study in Malawi. Tropical and Geographical Medicine, 25:372–380. Cadi-Soussi M et al.(1982) Épidémiologie des nématodes intestinaux au Maroc. Medecine d’Afrique Noire, 29:563–565. Callender JEM et al. (1991) Treatment effects in Trichuris dysentery syndrome. Acta Paediatrica, 83:1182–1187. Callender JEM et al. (1992) Trichuris infection and mental development in children. Lancet, 339:181. Carney WP et al. (1974) The Napu Valley, a new schistosomiasis area in Sulawesi, Indonesia. Southeast Asian Journal of Tropical Medicine and Public Health, 5:246–251. Carney WP et al. (1977) Intestinal and blood parasites in the North Lore District, Central Sulawesi, Indonesia. Southeast Asian Journal of Tropical Medicine and Public Health, 8:165–172. Carrie J (1982) Les nematodoses en Afrique Centrale: etude epidemiologique au Cameroun [Worm infections in central Africa: epidemiological study in Cameroon]. Médecine d’Afrique Noire, 29:557–562. Cerf P, Moyroud J, Coulanges P (1981) Enquête epidemiologique et problèmes sanitaires dans un village du centre est de Madagascar [Epidemiological survey and sanitary problems in a village in East Central Madagascar]. Archives de l’Institut Pasteur de Madagascar, 48:151–161. Chai JY et al. (1985) Prevalence, worm burden and other epidemiological parameters of Ascaris lumbricoides infection in rural communities in Korea. Korean Journal of Parasitology, 23:241–246. Chai JY et al. (1991) Reduction in the incidence of biliary and other surgical complications of ascariasis according to the decrease of its national egg prevalence in Korea. Korean Journal of Parasitology, 29:101–111. Chan L, Bundy DAP, Kan SP (1994) Aggregation and predisposition to Ascaris lumbricoides and Trichuris trichiura at the familial level. Transactions of the Royal Society of Tropical Medicine and Hygiene, 87:46–48. Chan L, Kan SP, Bundy DAP (1992) The effect of repeated chemotherapy on age-related predisposition to Ascaris lumbricoides and Trichuris trichiura. Parasitology, 104:371–377. Chan MS et al. (1994) The evaluation of potential global morbidity attributable to intestinal nematode infections. Parasitology, 109:373–87. Chang KC, Sun YL, Chiu JK (1973) A survey on parasitic infections among Aborigines in remote mountain area of Chien-Shih District, Hsinchu County, Taiwan. Journal of the Formosan Medical Association, 72:297–303. Intestinal nematode infections 289 Chatterjee BN, Mukhopadhyay BB (1985) A study of prevalence of bacterial infections and parasitic infestations in an urban slum community. Indian Journal of Public Health, 29:9–14. Chowdhury AB, Schad GA, Schiller EL (1968) The prevalence of intestinal helminths in religious groups of a rural community near Calcutta. American Journal of Epidemiology, 87:313–317. Clarke MD et al. (1973) A parasitological survey in the Jogjakarta area of Central Java, Indonesia. Southeast Asian Journal of Tropical Medicine and Public Health, 4:195–201. Colett S, Jalayer T, Kohout E (1966) Parasitic infection in the Jewish community of Shiraz, Iran. Israel Journal of Medical Sciences, 2:380–383. Colwell EJ et al. (1971) Intestinal parasitism in residents of the Mekong Delta of Vietnam. Southeast Asian Journal of Tropical Medicine and Public Health, 2: 25–28. Cooper ES et al. (1990) Growth suppression in the Trichuris dysentery syndrome. European Journal of Clinical Nutrition, 44:285–291. Cooper ES et al. (1991) Immediate hypersensitivity in colon of children with chronic Trichuris trichiura. Lancet, 338:1104–1107. Cooper ES et al. (1992) Intestinal nematode infections in children: the pathophysiological price paid. Parasitology, 104:S91–S103. Cooper ES, Bundy DAP (1987) Trichuriaisis. In: ZS Pawlowski (ed.) Intestinal helminthic infections, Bailliere’s Clinical Tropical Medicine and Communicable Disease, 2(3):629–643. Cooper ES, Bundy DAP (1988) Trichuris is not trivial. Parasitology Today, 4: 301–306. Cooper ES, Bundy DAP, Henry FJ (1986) Chronic dysentery, stunting, and whipworm infestation. Lancet, (ii):280–281. Croll NA, Ghadirian E (1981) Wormy persons: contributions to the nature and patterns of overdispersion with Ascaris lumbricoides, Ancylostoma duodenale, Necator americanus and Trichuris trichiura. Tropical and Geographical Medicine, 33:241–248. Croll NA et al. (1982) The population biology and control of Ascaris lumbricoides in a rural community in Iran. Transactions of the Royal Society of Tropical Medicine and Hygiene, 76:187–197. Crompton DWT (1988) The prevalence of ascariasis. Parasitology Today, 4:162–169. Crompton DWT (1989) Prevalence of ascariasis. In: Crompton DWT, Nesheim MC, Pawlowski ZS eds., Ascariasis and its prevention and control. London, Taylor and Francis, pp. 45–69 (Chapter 4). Crompton DWT, Nesheim MC, Pawlowski ZS (eds). (1989). Ascariasis and its prevention and control. London: Taylor and Francis. Crompton DWT, Stephenson LS (1990) Hookworm infection, nutritional status and productivity. In: Schad GA, Warren KS, eds., Hookworm disease, current status and new directions, pp. 231–65. London, Taylor and Francis. 290 Global Epidemiology of Infectious Diseases Crompton DWT, Whitehead RR (1993) Hookworm infections and human iron metabolism. Parasitology, 107:S137–S145. Cross JH et al. (1970) Survey for human intestinal and blood parasites in Bojolali, Central Java, Indonesia. Southeast Asian Journal of Tropical Medicine and Public Health, 1:354–360. Cross JH et al. (1975) Parasitology survey and seroepidemiology of amoebiasis in South Kalimantan (Borneo), Indonesia. Southeast Asian Journal of Tropical Medicine and Public Health, 6:52–60. Cross JH et al. (1976) Parasitology survey in Northern Sumatra, Indonesia. Journal of Tropical Medicine and Hygiene, 79:123–131. Cross JH et al. (1977a) Biomedical survey on the Minahasa Peninsula of North Sulawesi, Indonesia. Southeast Asian Journal of Tropical Medicine and Public Health, 8:390–399. Cross JH et al. (1977b) Biomedical survey in Irian Jaya (West Irian), Indonesia. Southeast Asian Journal of Tropical Medicine and Public Health, 8:532–536. Cross JH et al. (1977c) Biomedical survey in North Samar Province, Philippine Islands. Southeast Asian Journal of Tropical Medicine & Public Health, 8: 464–75. Davies MRQ, Rode H (1982) Biliary ascariasis in children. In: Rickham PP et al., eds., Pediatric surgery in tropical countries: progress in paediatric surgery (Vol. 15). Baltimore, Munich, Urban and Schwarzenberg, pp. 55–74. de Carneri I (1968) Indagine elmintologica quantitativa nella popolazione scolare di 21 centri rurali della provincia di Pavia [Quantitative helminthologic study of the school children in 21 rural centers in Pavia province]. Nuovi Annali d’Igiene e Microbiologia, 19:42–65. de Carneri I, Garofano M, Grassi L (1967) Il ruolo della Tricocefalosi nel ritardo dello sviluppo fisico e mentale infantile in base ad un’inchiesta quantitativa nel basso lodigiano. [The role of trichcephaliasis in infantile physical and mental retarded development studies by a quantitative survey in Basso Lodigiano]. Rivista di Parassitologia, 28:103–122. de Carneri I (1968) Indagine elmintologica quantitative nella popolazione scolare di 21 centri rurali della provincia di Pavia con riferimenti alla situazione socioeconomica e igienica familiare e al rendimento scolastico. Nuovi Annali d’Igiene e Microbiologia, 19:1–24. de Carneri I, Garofano M, Grassi L (1967) Il ruolo della Tricocefalosi nel ritardo dello sviluppo fisico e mentale infantile in base ad un’inchiesta quantitativa nel basso lodigiano. Rivista di Parassitologia, 28:103–122. de Meira MTV, Nogueira JFP, Simoes TS (1947) Contribucao para o estudo do parasitismo intestinal nas ilhas do Sal, Boa Vista, e S. Nicolau (Cabo Verde) [Contribution to a study of intestinal parasite infection in the islands of Sal, Boa Vista, S. Niolau (Cape Verde)]. Anais do Instituto de Medicina Tropical, 4:239–256. de Vlas SJ, Gryseels B (1992) Underestimation of Schistosoma mansoni prevalences. Parasitology Today, 8:274–277. Intestinal nematode infections 291 Desowitz RS (1963) Review of research in parasitology in Singapore. Singapore Medical Journal, 4:30–33. Dunn FL (1972) Intestinal parasitism in Malayan aborigines (Orang Asli). Bulletin of the World Health Organization, 46:99–113. Elkins DB, Haswell-Elkins M, Anderson RM (1986). The epidemiology and control of intestinal helminths in the Pulicat Lake region of Southern India. I. Study design and pre- and post-treatment observations on Ascaris lumbricoides infection. Transactions of the Royal Society of Tropical Medicine and Hygiene, 80:774–792. El-Rahimy H et al. (1986) Prevalence of intestinal parasitic infections in Macca, Saudi Arabia. Journal of the Egyptian Society of Parasitology, 16:185–7. Farag HF (1985) Intestinal parasitosis in the population of the Yemen Arab Republic. Tropical and Geographical Medicine, 37:29–31. Faucher P et al. (1984) Etude epidémiologique des helminthiases intestinales dans la région de Kaya (Haute-Volta) [Epidemiology of intestinal helminthiasis in the Kaya Region (Upper-Volta)]. Bulletin de la Société de Pathologie Exotique et de ses Filiales, 77:507–513. Fisher RM, Cremin BJ (1970) Rectal bleeding due to Trichuris trichiura. British Journal of Radiology, 43:214–215. Fleming AF (1982) Iron deficiency in the tropics. Clinics in Haematology, 11:365–388. Flores AL, Reynaga FJ (1978) Obstrucción intestinal por ascarides en la infancia [Intestinal obstruction ducts ascarides in childhood]. Boletin Medico del Hospital Infantil de Mexico, 35:197–204. Forrester JE et al. (1988) Clustering of Ascaris lumbricoides and Trichuris trichiura infections within households. Transactions of Royal Society of Tropical Medicine and Hygiene, 82:282–28. Forrester JE et al. (1990) Predisposition of individuals and families in Mexico to heavy infection with Ascaris lumbricoides and Trichuris trichiura. Transactions of the Royal Society of Tropical Medicine and Hygiene, 84:272–276. Garin Y et al. (1978) Le parasitisme intestinal au Gabon Oriental [Intestinal parasitism in eastern Gambia]. Bulletin de la Société de Pathologie Exotique et de ses Filiales, 71:157–164. Gilman RH et al. (1983) The adverse consequences of heavy Trichuris infection. Transactions of the Royal Society of Tropical Medicine and Hygiene, 77: 432–438. Glewwe P, Jacoby HG (1995) An economic analysis of delayed primary school enrollment in a low income country: the role of early childhood nutrition. The Review of Economics and Statistics, 77:156–169. Goldsmid JM (1976) Ascariasis in Rhodesia. Central African Journal of Medicine, 22:220–223. Goldsmid JM et al. (1976) The intestinal protozoa and helminths infecting Africans in the Gatooma region of Rhodesia. Central African Journal of Medicine, 22:91–95. 292 Global Epidemiology of Infectious Diseases Grantham-McGregor SM et al. (1991) Nutritional supplementation, psychosocial stimulation, and mental development of stunted children: the Jamaican study, Lancet, 388:1–5. Grell GAC et al. (1981) A survey of parasites in primary school children in Dominica, West Indies. Annals of Tropical Paediatrics, 1:155–160. Guyatt HL (1992) Parasite population biology and the design and evaluation of helminth control programmes. University of London, (unpublished doctoral thesis). Guyatt HL et al. (1990) The relationship between the frequency distribution of Ascaris lumbricoides and the prevalence and intensity of infection in human communities. Parasitology, 101:139–143. Guyatt HL, Bundy DAP (1991) Estimating prevalence of community morbidity due to intestinal helminths: prevalence of infection as an indicator of the prevalence of disease. Transactions of the Royal Society of Tropical Medicine and Hygiene, 85:778–782. Guyatt HL, Evans D (1992) Economic considerations for helminth control. Parasitology Today, 8:397–402. Guyatt HL, Bundy DAP (1993) Estimation of intestinal nematode prevalence: influence of parasite mating patterns. Parasitology, 107:99–105. Hall A (1982) Intestinal helminths of man: the interpretation of egg counts. Parasitology, 85:605–613. Hall A, Anwar KS, Tomkins AM (1992) Intensity of reinfection with Ascaris lumbricoides and its implications for parasite control. Lancet, 339:1253–1257. Haswell-Elkins MR, Elkins DB, Anderson RM (1987) Evidence for predisposition in humans to infection with Ascaris, hookworm, Enterobius and Trichuris in a South Indian fishing community. Parasitology, 95:323–337. Holland CV et al. (1989) The epidemiology of Ascaris lumbricoides and other soil-transmitted helminths in primary school children from Ile-Ife, Nigeria. Parasitology, 99:275–285. Ilardi I et al. (1980) Indagine epdiemiologica, con methodo diretto e immunodiagnostico, delle malattie-parasitarie nei residente di un villagio somalo situato in zona irrigata. [Epidemiological studies with direct and immunodiagnostic methods of parasitic diseases in residents of a Somalia village located in the irrigated zones]. Parassitologia, 22(1-2):141–148. Ilardi I et al. (1981) Indagine parassitologica con metodo diretto ed immunodiagnostico in un villaggio somalo (Mooda Moode) in zona arida. Parassitologia, 23:191–193. Ilardi I et al. (1987) The prevalence and intensity of intestinal parasites in two Somalian communities. Transactions of the Royal Society of Tropical Medicine and Hygiene, 81:336–338. Jamison DT (1986) Child malnutrition and school performance in China. Journal of Developmental Economics, 20:299–309. Jones HI (1976) A study of human intestinal helminthiases on Kar Kar Island, Madang Province. Papua New Guinea Medical Journal, 19:165–172. Intestinal nematode infections 293 Jordan P, Webbe G (1982) Schistosomiasis, epidemiology, treatment and control. London, William Heinemann. Joseph SW et al. (1978) Human parasitoses of the Malili Area, South Sulawesi (Celebes) Province, Indonesia. Southeast Asian Journal of Tropical Medicine and Public Health, 9:264–271. Juminer B, Diallo S, Laurens D (1971) Enquête parasitologique au sein d’une collectivité serère du Sine (Senégal). [Parasitologic investigation in a community in Sine (Senegal)]. Bulletin de la Société de Pathologie et de ses Filiales 1971, 64:901–912. Kan SP (1982) Soil-transmitted helminthiases in Selangor, Malaysia. Medical Journal of Malaysia, 37:180–190. Kan SP (1985) The effect of long-term deworming on the prevalence of soil-transmitted helminthiases in Malaysia. Medical Journal of Malaysia, 40:202–210. Keusch GT (1982) Summary and recommendations. The biology of parasitic infections: workshop on interactions of nutrition and parasitic diseases. Reviews of Infectious Diseases, 4:901–7. Keymer A, Pagel M (1990) Predisposition to helminth infection. In: Schad GA, and Warren KS, eds., Hookworm disease, London, Taylor and Francis, pp. 177–209. Kouri P, Valdos Diaz R (1952) Concepto actual sobre el papel patogeno del Tricocefalo dispar [Present situation of the pathogenic role of trichocephalus]. Revisto Cubano de la Medicina Tropicale y Parasitoligica, 8:37–41. Krey W (1949) Der Darmstaedter Spulwurmbefall und seine Bekaempfung [Infestation with the Darmstaedt worm and its treatment]. Zeitschrift fuer Hygiene, 129:507–518. Kuntz RE (1960) Intestinal protozoa and helminths in school children of Dacca, East Pakistan (East Bengal). American Journal of Tropical Medicine and Hygiene, 9:168–172. Kuntz RE, Lawless DK, Mansour NS (1955) A cursory survey of the intestinal parasites of natives living in Southwest Sudan. American Journal of Tropical Medicine and Hygiene, 4:895–900. Kvalsvig JD, Cooppan RM, Connolly KJ (1991) The effects of parasite infections on cognitive processes in children. Annals of Tropical Medicine and Parasitology, 85:551–568. Lane C (1917) An investigation into ankylostome infection in 11 000 inhabitants of the Darjeeling District of India. Indian Journal of Medical Research Lapierre J, Tourte-Schaeffer C (1982) Prevalence des principales nematodoses au Togo. Medicine d’Afrique Noire, 29:571–572. Lawless DK, Kuntz RE, Strome CPA (1956) Intestinal parasites in an Egyptian village of the Nile Valley with emphasis on the protozoa. American Journal of Tropical Medicine and Hygiene, 5:1010–1014. Levav M et al. (1995) Parasitic infection in malnourished school children: effects on behaviour and EEG. Parasitology, 110:103–111. 294 Global Epidemiology of Infectious Diseases Lwambo NJS, Bundy DAP, Medley GFH (1992) A new approach to morbidity risk assessment in hookworm endemic communities. Epidemiology and Infection, 108:469–481. MacDonald TT et al. (1991) Histopathology and immunohistochemistry of the caecum in children with the Trichuris dysentry syndrome. Journal of Clinical Pathology, 44:194–199. Maki T (1972) Surgical diseases due to Ascaris lumbricoides. In: Morishita K et al. eds., Progress in medical parasitological in Japan, (Vol. 4.), Tokyo, Meguro Parasitological Museum, pp. 221–270. Marsden PD (1963) Observations on the incidence, diagnosis and treatment of ascariasis in Gambian schoolchildren. African Child Health, 9:52. Martinez-Torres C et al. (1967) Hookworm infection and intestinal blood loss. Transactions of the Royal Society of Tropical Medicine and Hygiene, 61:373–383. May RM, Anderson RM (1984) Spatial heterogeneity and the design of immunisation programs. Mathematical Bioscience, 72:83–111. McGregor IA, Smith DA (1952) A health, nutrition and parasitological survey in a rural village (Keneba) in West Kiang, Gambia. Transactions of the Royal Society of Tropical Medicine and Hygiene, 46:403–427. Medley GF, Guyatt HL, Bundy DAP (1993) A quantitative framework for evaluating the effect of community treatment on the morbidity due to ascariasis. Parasitology, 106:211–221. Mohamed NH et al. (1985) Prevalence of parasitic diseases in Tahrir Province and Shubra el Kheima. Journal of the Egyptian Society of Parasitology, 15:535–542. Mohamed NH et al. (1988) Parasitic diseases in El Korean, Sharkyia Governorate with evaluation of some methods used in stool examination. Journal of the Egyptian Society of Parasitology, 18:305–311. Moock PR, Leslie J (1986) Childhood malnutrition and schooling deficit in the Terai rgion of Nepal. Journal of Development Economics, 20:33–52. Mukerji AK, Mathen KK (1950) Survey of hookworm infection in the Jharia coalfields settlement area. Indian Journal of Medical Research, 38:95–118. Nesheim MC (1989) Ascariasis and Human Nutrition. In: Crompton DWT, Nesheim MC and Pawlowski ZS, eds., Ascariasis and its Prevention and Control, London, Taylor and Francis, pp. 87–100. Murray CJL, Lopez AD (1996) Global health statistics: a compendium of incidence, prevalence and mortality estimates for over 200 conditions. Global Burden of Disease and Injury Series, Volume II. Harvard University Press, Boston, 1996. Nazari MR, Massoud J (1980) Hookworm infection and blood changes in the rural population of Dezful Area in Khuzestan, south-west of Iran. Bulletin de la Société de Pathologie Exotique et ses Filiales, 73:108–111. Neghme A, Silva R (1952) Estado actual de las infecciones por Ascaris lumbricoides y Trichuris trichiura en Chile [Current situation of infections with Ascaris lumbricoides and Trichuris trichiura in Chile]. Boletin Chileno de Parasitologia, 18(4):100–103. Intestinal nematode infections 295 Neghme A, Silva R (1963) Estado actual de las infecciones por Ascaris lumbricoides y Trichuris trichiura en la ciudad de Santiago de Chile[Current situation of infections with Ascaris lumbricoides and Trichuris trichiura in the city of Santiago, Chile]. Boletin Chileno de Parasitologia, 18:54–60. Neghme A, Silva R, Artigas J (1983) Nuevos aspectos sobre epidemiologia de la amebiasis y enteroparasitosis [New aspects of the epidemiology of amoebiasis and enteric parasitosis]. Revista Chilena de Hygiene y Medicina Preventa, 14: 243–257. Nesheim MC (1989) Ascariasis and Human Nutrition. In: Crompton DWT, Nesheim MC and Pawlowski ZS, eds., Ascariasis and its prevention and control, London, Taylor and Francis, pp. 87–100. Niazi AD et al. (1975) Pilot study on prevalence of ankylostomiasis in Iraq. Bulletin of Endemic Diseases, 16:105–142. Nogueira JFP, Coito AMF (1950) Sobre a ancilostomiase autoctona na ilha Bravaarquipelago de Cabo Verde [About autoctone anyclostomiasis in the island Brava – archipelago of Cape Verde]. Anais do Instituto de Medicina Tropical, 7:253–281. Nokes C et al. (1991) Geohelminth infection and academic assessment in Jamaican children. Transactions of the Royal Society of Tropical Medicine and Hygiene, 85:272–273. Nokes C et al. (1992a) Moderate to high infections of Trichuris trichiura affect cognitive function in Jamaican school children. Parasitology, 104:539–547. Nokes C et al. (1992b) Parasitic helminth infection and cognitive function in school children. Proceedings of the Royal Society of London – Series B: Biological Sciences, 247:77–81. Oduntan SO (1974) The health of Nigerian children of school age (6–15 years): II – Parasitic and infective conditions, the special senses, physical abnormalities. Annals of Tropical Medicine and Parasitology, 68:145–156. Okumura M et al. (1974) Acute intestinal obstruction by Ascaris: analysis of 455 cases. Revisto do Instituto de Medecine Tropical de Sao Paulo, 16:292–300. Ortiz RL (1969) Indice coproparasitario en la poblacion de Bahia de Caraquez y Zonas Aledanas, Prov de Manabi [Parasitic indices in the population of Bahia de Caráquez and Zonas Aledanas, Manabi province]. Revista Ecuatoriana de Higiene y Medicina Tropical, 26:133–137. Pampiglione S, Hadjeres S (1965) Su di un nuovo focolaio di anchilostomiasi scoperto nell’est Algerino. [On a new focus of anchylostomiasis discovered in eastern Algeria]. Rivista di Parassitologia 26:241–50. Pampiglione S, Ricciardi ML (1971) Indagini parassitoloiche tra la popolazione del villaggio palafitticolo di Ganvie (Dahomey, Africa Occidentale). [Parasitological studies of the inhabitants of the pile-built village of Ganvie (Dahomey, West Africa)]. Parassitologia, 13:271–280. Papasarathorn T et al. (1975) Epidemiology of hookworm infection in Soong Nern district, Korat Province, Thailand. Southest Asian Journal of Tropical Medicine and Public Health, 6:82–92. 296 Global Epidemiology of Infectious Diseases Pawlowski ZS, Davis A (1989) Morbidity and mortality in Ascariasis. In: Crompton DWT, Nesheim MC, Pawlowski ZS, eds., Ascariasis and its prevention and control, pp. 71–86. London, Taylor and Francis. Peplow D (1982) Parasitos intestinales en la poblacion de varias regiones de Ecuador: estudio estadistico [Intestinal parasites among the populations of various regions of Ecuador: statistical study]. Boletin de la Oficina Sanitaria Panamericana, 93:233–239. Pollitt E (1990) Malnutrition and infection in the classroom. Paris: UNESCO. Pollitt E et al. (1986) Iron deficiency and behavioral development in infants and preschool children. American Journal of Clinical Nutrition, 43:555–565. Pollitt E, Haas J, Levitsky DA (1989) International conference on iron deficiency and behavioural development. American Journal of Clinical Nutrition, 50: 565–705. Preuksaraj S et al. (1981) Studies on prevalence and intensity of intestinal helminthic infections in the rural population of Thailand, 1980–1981. Collected papers on the Control of Soil-Transmitted Helminthiases: APCO (Vol. 2). Pritchard DI et al. (1990) Epidemiology and immunology of Necator americanus infection in a community in Papua New Guinea: humoral responses to excretory-secretory and cuticular collagen antigens. Parasitology,100 :317–326. Puga S et al. (1980) Infeccion humana por protozoos y helmintos intestinales en dos poblaciones costaneras de la Provincia de Valdivia. [Human infection by protozoa and helminths in two small towns of the province of Valdivia]. Revista Medica de Chile, 108:608–611. Ramdath DD et al. (1995) Iron status of schoolchildren with varying intensities of Trichuris trichiura infection. Parasitology, 110:347–351. Ramirez R et al. (1972) Frecuencia en Chile de las infecciones humanas por protozoos y helmintos intestinales. (Incidence in Chile of human intestinal infections by protozoa and helminths]. Boletin Chileno de Parasitologia, 27:116–118. Ramsey FC (1962) Trichuris dysentery syndrome. West Indian Medical Journal, 11:235–239. Ranque P (1982) Aperçu épidémiologique des nematodoses intestinales au Mali [Observed epidemiology of intestinal nematodes in Mali]. Medecine d’Afrique Noire, 29:577–580. Ratard RC et al. (1991) Ascariasis and trichuriasis in Cameroon. Transactions of the Royal Society of Tropical Medicine and Hygiene, 85:84–88. Reeder MM, Astacio JE, Theros EG (1968) An exercise in radiologic-pathologic correlation. Radiology, 90:382–387. Ricciardi ML (1972) Differenze nella formula parassitaria intestinale tra popolazione indigena e popolazione Europea immigrata in una azienda forestale della Repubblica Centrafricana [Difference in the picture of intestinal parasitimia in indigenous populations and European immigrant populations in a forest area of the Central African Republic]. Parassitologia, 14:347–351. Richard-Lenoble D et al. (1982) Prevalences au Gabon des nematodoses intestinales pathogenes chez l’homme. [Prevalence of human pathogenic intestinal nematode infections in Gabon]. Medicine d’Afrique Noire, 29:581–585. Intestinal nematode infections 297 Ripert C et al (1978) Étude épidémiologique des nematod es intestinales dans cinq villages de la vallée de la Sanago (Cameroun): résultats de traitement de masse des populations par le pamoate de Pyrantel. [Epidemiological study of intestinal nematode infections on five villages in the Sanago valley (Cameroon). Results of the mass treatment of the population with pyrantel pamoate]. Bulletin de la Société de Pathologie Exotiqueet des ses Filialies, 71:361–369. Ripert C, Leugueun-Ngougbeou J, Same-Ekobo A (1982) Etude epidemiologique de la bilharziose et des nematodoses intestinales a Bafia (Cameroun). [An epidemiological study of intestinal bilharziasis and roundworm infections in Bafia (Cameroon)]. Bulletin de la Société de Pathologie Exotique et de ses Filiales, 75:55–61. Rojas ARV, Zumbado H (1980) Parasitosis intestinal en Cartago y Limon. Acta Medica Costarricense, 23:245–251. Rougement A et al. (1974) Prevalence des helminthiases intestinales dans la region de Bamako (Mali). Medecine Tropicale, 34:29–36. Russell PF (1928) Nematodes in man in the Straits Settlements: a preliminary report. The Malayan Medical Journal, 3:113–123. Russell PF (1934) Racial and age-group incidence of common intestinal helminths in the Straits Settlements. The Malayan Medical Journal, 9:17–22. Sadun EH (1955) Studies on the distribution and epidemiology of hookworm, Ascaris and Trichuris in Thailand. American Journal of Hygiene, 62:116–155. Santos AG(1952) Bilharziases e helmintiases intestinais nos indigenas da circunscricao de chimoio [Schistosomiasis and intestinal helminth infection in indigenous populations]. Anais do Instituto de Medicina Tropical, 9:1155–1174. Sanyal SC et al. (1972) Enteric infections in Sunderpur slum area. Indian Journal of Medical Research, 60:979–986. Savioli L, Bundy DAP, Tomkins AM (1992) Intestinal parasitic infections: a soluble public health problem. Transactions of the Royal Society of Tropical Medicine and Hygiene, 86:353–354. Saxena PC, Prasad BG (1971) A review of surveys on intestinal helminths in India. Indian Journal of Public Health, 15:31–38. Schad GA, Banwell JG (1984) Hookworms, In: Warren KS, Mahmoud AAF, eds., Tropical and Geographical Medicine. McGraw-Hill Book Company, New York, pp. 359–372. Schad GA, Anderson RM (1985) Predisposition to hookworm infection in humans. Science, 228:1537–1540. Schad GA, Warren KS, eds. (1990) Hookworm isease: current status and new directions. London, Taylor and Francis. Schenone H et al. (1981) Aspectos epidemiológicos de las infecciones humanas por protozoos y helmintos intestinales en Chile (1970–1980). [Epidemiological aspects of human infections produced by intestinal protozoa and helminths in Chile (1970–1980)]. Boletin Chileno de Parasitologia, 36:44–48. Seo BS et al. (1983) Efficacy of mebendazole in treatment and control of trichuriasis in Korea. Korean Journal of Parasitology, 21:95–101. 298 Global Epidemiology of Infectious Diseases Shield J et al. (1986) Intestinal helminthiasis and nutritional status including iron, zinc and copper in Papua New Guinea urban children aged 1 to 5 years and effect of anthelmintic treatment. Papua New Guinea Medical Journal, 29:317–331. Simeon D et al. (1994) School performance, nutritional status and Trichuriasis in Jamaican schoolchildren. Acta Pediatrica, 83:1188–1193. Simeon DT et al. (1995) Treatment of Trichuris trichiura infections improves growth, spelling scores and school attendance in some children. Journal of Nutrition, 125:1875–1883. Simeon DT, Grantham-McGregor SM, Wong MS (1995) Trichuris trichiura infection and cognition in children: results of a randomised clinical trial. Parasitology, 110:457–64. Sinniah B et al. (1978) Prevalence of parasitic infections in Malaysian oil palm estate workers. Southeast Asian Journal of Tropical Medicine and Public Health, 9:272–276. Sornmani S et al. (1974) An epidemiological survey of human intestinal parasites in Vientiane, Laos. Southeast Asian Journal of Tropical Medicine and Public Health, 5:541–6. Sorvillo F, Ash LR (1982) Parasitic diseases in Karamoja, Uganda. The Lancet, 8277:912–913. Stafford EE et al. (1980) Intestinal and blood parasites in the Torro Valley, Central Sulawesi, Indonesia. Southeast Asian Journal of Tropical Medicine and Public Health, 11:468–72. Stephenson LS (1987) Impact of helminth infections on human nutrition: schistosomes and soil transmitted helminths, New York, Taylor and Francis. Stephenson LS et al. (1989) Treatment with a single dose of albendazole improves growth of Kenyan schoolchildren with hookworm, Trichuris trichiura and Ascaris lumbricoides infections. American Journal of Tropical Medicine and Hygiene, 41:78–87. Stephenson LS et al. (1990) Improvements in physical fitness of Kenyan schoolboys infected with hookworm, Trichuris trichiura and Ascaris lumbricoides following a single dose of albendazole. Transactions of the Royal Society of Tropical Medicine and Hygiene, 84:277–282. Stephenson LS et al. (1993) Physical fitness, growth and appetite of Kenyan school boys with hookworm, Trichuris trichiura and Ascaris lumbricoides infections are improved four months after a single dose of albendazole. Journal of Nutrition, 123:1036–1046. Stephenson LS, Latham MC, Oduori ML (1980) Costs, prevalence and approaches for control of Ascaris infection in Kenya. Journal of Tropical Paediatrics, 26: 246–263. Sternberg RJ, Powell C, McGrane P (1995) Effects of a parasitic infection on cognitive functioning. Unpublished. Stiles CW (1915) Zooparasitic intestinal infections: an analysis of infections found among 1287 school children (776 white, 511 negro) of the city of X. Public Health Reports, 30:3738–3745. Stoll NR (1947) This wormy world. Journal of Parasitology, 33:1–18. Intestinal nematode infections 299 Stoopen M, Beltran FA (1964) Características epidemiológicas de las helmintiasis en Yancuictlalpan, Puebla (Mexico) [Epidemiological characteristics of helminth infections in Yancuictlalpan, Puebla (Mexico)]. Revista Medicina, 45:28–33. Sturchler D et al. (1980) Intestinal parasitoses in eight Liberian settlements: prevalences and community anthelmintic chemotherapy. Tropenmedizin und Parasitologie, 31:87–93. Sturrock RF (1964) Hookworm studies in Tanganyika: investigations at Machui in the Tanga Region. East African Medical Journal, 41:520–529. Tadros MB (1973) Parasitological, mensural and nutritional studies on 3000 Egyptian school children in Giza Province (Part I). Journal of the Egyptian Society of Parasitology, 46:2–3. Tantengco VO, Marzan AM, de Castro CR (1972) Helminthic infections and blood parameter studies in Filipino schoolchildren. Southest Asian Journal of Tropical Medicine and Public Health, 3:580–586. Taticheff S, Abdulahi Y, Haile-Meskal F (1981) Intestinal parasitic infection in pre-school children in Addis Ababa. Ethiopian Medical Journal, 19:35–40. Tedla S, Ayele T (1986) Ascariasis distribution in Ethiopia. Ethiopian Medical Journal, 24:79–86. Thein-Hlaing (1987) A profile of ascariasis morbidity in Rangoon Children’s hospital, Burma. Journal of Tropical Medicine and Hygiene, 90:165–9. Thein-Hlaing et al. (1990) Role of ascariasis in surgical abdominal emergencies in the Rangoon Children’s Hospital, Burma. Annals of Tropical Pediatrics, 10:53–60. Thiers G, Lassoued H, Abid A (1976) Étude d’un foyer d’ascaridiase dans une localite du Cap Bon (Tunisie). [Study on a focus of ascaridiasis in a locality of Cap Bon (Tunisia)]. Bulletin de la Société de Pathologie Exotique et de ses Filiales, 69:320–329. Tomkins AM, Watson F (1989) Malnutrition and infection: a review. Geneva: United Nations, Administrative Committee on Coordination/Subcommittee on Nutrition. Venugopal S et al. (1987) Sigmoidoscopy in children with chronic mucoid diarrhoea in rural St Lucia. Annals of Tropical Pediatrics, 7:104–108. Vinha C (1971) Incidencia, no Brasil, de helmintos transmitidos pelo solo – rotina coproscopica do ex-departamento nacional de endemias rurais. [Occurrence, in Brazil, of helminthiasis transmitted by the soil. Croposcopic routine of the ex-Departmento Nacional de Endemias Rurais]. Revista Brasileira de Malariologia e Doencas Tropicais, 23:3–17. Waite JH, Neilson IL (1919) A study of the effects of hookworm infection upon mental development of North Queensland school children. Medical Journal of Australia,1:1–7. Walsh JA (1984) Estimating the burden of illness in the tropics. In: Warren KS, Mahmoud AAF, eds., Tropical and geographical medicine. McGraw-Hill Book Company, New York, pp 1073–1085. 300 Global Epidemiology of Infectious Diseases Warren KS et al. (1993) Helminth Infections. In: Jamison DT et al. eds. Disease control priorities in developing countries, Oxford, Oxford University Press, pp. 131–60. Wells WH, Blagg W (1956) A survey of human intestinal parasites in a fishing village of Northern Egypt. American Journal of Tropical Medicine and Hygiene, 5:266–268. Wenlock RW (1977) The prevalence of hookworm and of S. haematobium in rural Zambia. Tropical and Geographical Medicine, 29:415–421. Wong HB, Tan KH (1961) Severe whipworm infestation in children. Singapore Medical Journal, 2:34–37. World Bank (1993) World development report 1993: investing in health. Washington DC, World Bank. World Health Organization (1987). Prevention and control of intestinal parasitic infections. Geneva, World Health Organization. (WHO Technical Report Series, No. 749). World Health Organization (1992). Global health situation and projections, estimates. Division of Epidemiological Surveillance and Health Situation Assessment, Geneva, World Health Organization (WHO/HST/92.1). Yokogawa RW (1976) World Health Organization informal document. Geneva, World Health Organization (WHO/HELM/67.76). Yu S-H et al. (1994) Nationwide survey of human parasites in China. Southeast Asian Journal of Tropical Medicine and Public Health, 25:4–10. Zamora AVC et al. (1978) Estado actual de parasitismo intestinal en la zona de Puriscal [Current situation of intestinal parasitosis in the Puriscal area]. Acta Medica Costarricense, 21:271–273. Chapter 10 Trachoma-Related Visual Loss B Thylefors, K Ranson, A-D Négrel, R Pararajasegaram Introduction Trachoma as a cause of blindness has been known to mankind for many centuries; it was documented in ancient China, Egypt and India. The disease probably came to Europe with the Napoleonic troops returning from Egypt and was endemic in most European countries well into the 20th century (Jitta & Luharis 1930). Trachoma, a communicable eye disease most commonly found amongst poor populations living in crowded conditions with insufficient personal and environmental hygiene, is the most important cause of preventable blindness in the world. The disease is still commonly found in many developing countries, usually in the poorest population groups in rural areas. Trachoma is a typical socioeconomically determined disease, with poverty, low living standards, and lack of hygiene as main risk factors. The disease is typically “clustered” in relation to community development and social class (Dawson, Jones & Tarizzo 1981). The trachomatous infection starts early in life, but the blinding outcome usually does not occur until adulthood, with a predominance of visual loss in women. The disease can constitute a significant cause of visual disability, adding to the already high burden of disease commonly found in poor rural populations. Trachoma is caused by Chlamydia trachomatis, a microorganism that is intracellular (like viruses) but susceptible to some antibiotics (like bacteria). Chlamydial infection may give systemic manifestations, including respiratory tract infections in infants and genital tract infections in adults. The relationship between ocular and genital chlamydial infections is complex and not the subject of this review. The epidemiology of trachoma is characterized by the early inflammatory phase, often during childhood, involving the tarsal conjunctive (the mucous membrane lining inside the eyelids). The upper eyelid is the main site of inflammation and the infection is invariably bilateral. The invasion of chlamydiae into the conjuctival cells leads to the formation of small 302 Global Epidemiology of Infectious Diseases follicles, which represent the inflammatory reaction to the presence of the infectious agent. The follicles gradually increase in size, up to 1–2 mm; as they “mature,” reaching their optimal size, they rupture and are eventually replaced by scar tissue. The scars can be seen as discrete white lines or stars on the tarsal conjuctival surface. With an increasing number of scars, there is traction inside the eyelid, leading eventually to an inturned lid (entropion), with eyelashes rubbing against the cornea (trichiasis). This causes superficial lesions in the corneal epithelium, which over time result in a gradual infiltrate and opacification of the cornea, often following severe secondary infections. The classical profile of trachoma is a severe blinding disease, usually hyperendemic amongst poor and underserved rural populations. The disease can, however, also exist in a non-blinding form with less intense infection, although still prevalent in the communities concerned. Under such circumstances, there is still typically a substantial burden of active inflammatory disease amongst children, whereas there are very few cases of trichiasis or corneal opacification in the adults. Trichiasis and corneal opacification usually do not occur until the third or fourth decade of life, but this varies in relation to the intensity of infection. Women tend to have more active inflammatory disease because of their close contact with children, who constitute a reservoir of infection. Accordingly, trichiasis is particularly common in middle-aged and elderly women (Dawson, Jones & Tarizzo 1981). The prevention of blindness from trachoma can be effectively achieved by: prevention of trachoma infection through health education and improved standards of living; antibiotic treatment of inflammatory disease; or surgery against trichiasis to prevent corneal opacification. Specific guidelines have been developed for the antibiotic treatment of trachoma, as well as for standard surgical procedures (Reacher et al. 1992). There have been few attempts in the past to measure the global disease burden of trachoma and the resulting visual loss. A WHO Scientific Group on Trachoma Research in 1959 referred to an estimated 400 million cases of trachoma (World Health Organization 1959). In the 1960s, the results of a number of surveys on trachoma and experience gained in several national campaigns led to a global estimate of 500 million cases of trachoma worldwide, with at least 2 million blind as a result of the disease (Bietti, Freyche & Vozza 1962). Subsequently, in the light of better epidemiological data on blindness, the figure for the estimated number of blind was adjusted to 6–9 million (World Health Organization 1984). In 1985, Dawson & Schachter (1985) estimated by means of a simple model that there could be some 360 million people with trachoma. A more comprehensive model was developed a few years later by the WHO, partially based on the results of a global questionnaire on trachoma (Thylefors, Négrel & Pararajasegaram 1992). Thus, it was estimated in 1992 that approximately 146 million people had active inflammatory disease. Trachoma-Related Visual Loss 303 That estimate took into account only cases of active disease in need of treatment, in accordance with a new and simplified grading system for trachoma (Thylefors et al. 1987). A separate estimate was subsequently put forward, taking into account the number of blind attributable to trachoma and cases with disabling, or potentially disabling, lesions, i.e. corneal opacification and trichiasis (Thylefors et al. 1995). The overall figure for this category of blind and those at immediate risk of blindness was 5.9 million. Definition A multitude of definitions of blindness were in use up to 1972, when a WHO study group on the prevention of blindness recommended a uniform definition for use in surveys and reporting systems. The definition is expressed as vision less than 3/60 (0.05) in the better eye with best possible optical correction. This corresponds to the inability to distinguish the fingers of a normal hand at a distance of 3 metres (10 feet), and it implies loss of “walk-about” vision. The 1972 recommendation has become universally accepted and is included in the ninth and tenth revisions of the International Classification of Diseases (ICD) (World Health Organization 1977, 1992). The WHO study group also defined other categories of visual loss, notably the concept of “low vision,” i.e. significant or severe visual impairment, but not quite blindness, in the individual. Low vision is thus defined as vision less than 6/18 (0.3) but equal to or better than 3/60 (0.05). It should be noted that the above recommended definitions are mainly for uniform reporting of data; each country may still have its own criteria for social and rehabilitative care for people with visual loss. The definition of trachoma, as given by the WHO Expert Committee on Trachoma in 1962, reads: “Trachoma is a specific, communicable kerato-conjunctivitis, usually of chronic evolution, …characterized by follicles, papillary hyperplasia, pannus, and it its later stages, cicatrization” (World Health Organization 1962). The definition for a case of trachomatous blindness commonly applied is the central opacification of the cornea (leucoma) in the presence of typical conjunctival scars and entropion/trichiasis. The classification of trachoma and its complications has evolved over a period of several decades, from the classification proposed by MacCallan in the early part of this century, up to the most recent simplified grading system proposed by the WHO (Thylefors et al. 1987). Differential diagnostic problems refer mainly to other forms of follicular conjunctivitis; criteria for the field diagnosis of trachoma were laid down by the WHO Expert Committee on Trachoma (World Health Organization 1962), e.g. the presence of typical follicles or infiltrate in the tarsal conjunctiva. Trachoma tends to be underestimated as a cause of blindness, especially in situations where complications of trachoma, 304 Global Epidemiology of Infectious Diseases such as suppurative keratitis and phthisis bulbi, are mistakenly recorded as the primary cause. Review of empirical data bases Data sources on trachoma are unfortunately very limited. The epidemiology of the disease is such that only population-based assessments are of value. Surveys of schoolchildren were carried out in many countries in the 1960s, but they are limited by the low rate of school attendance in some settings and the lack of data on trichiasis to confirm the blinding potential of trachoma in the particular population. Community-based surveys have not been undertaken in many countries in recent years, although the assessment of trachoma is part of the WHO-suggested procedure for an overall blindness survey (Thylefors 1987). The sampling poses a problem in such a situation, when in fact the epidemiological mapping of trachoma is the priority for the planning of intervention. Data for eye clinics are notoriously unreliable for trachoma, in view of the patchy distribution of the disease and its social profile, being most common in impoverished rural populations. Overall, trachoma studies tend to be biased to focus on areas where trachoma is strongly suspected or known to exist. Very few nationwide surveys have attempted to give a mapping of trachoma, which is obviously difficult, given the focal distribution of the disease in relation to socioeconomic, cultural and environmental circumstances. Available data on trachoma and related visual loss are reviewed in Table 10.1. Studies are included in this compilation only if they are population- Box 10.1 International Classification of diseases (ICD) codes related to trachoma ICD-9 ICD-10 076 Trachoma A.71 Trachoma 076.0 Initial stage Trachoma dubium A.71.0 Initial stage of trachoma Trachoma dubium 076.1 Active stage Granula conjunctivities (trachomatous) Trachomatous – follicular conjunctivitis – pannus A.71.1 Active stage of trachoma Granular conjunctivitis (trachomatous) Trachomatous – follicular conjunctivitis – pannus 076.9 Unspecified Trachoma NOS A.71.9 Trachoma, unspecified 374.0 Entropion and trichiasis of eyelid B.94.0 Sequelae of trachoma 139.1 Late effects of trachoma H.02.0 Entropion and trichiasis of eyelid 1987 1985 China China 1990 Australia 1973 1976 India India IND 1976–1977 Australia EME 1982–1985 Date of study 20 134 1 060 1 514 5 134 10 279 1 579 316 433 083 Population examined ≤ 3/60 ≤ 6/60 < 3/60 ≤ 6/60 — < 3/60 < 3/60 Community surveys in rural Uttar Pradesh.Villages closest to the health centre were chosen for the study. Only those reporting visual difficulty were examined. Community survey in rural Uttar Pradesh. The village surveyed is in the area of a rural health centre. Reports on blindness attributable to both trachoma and associated conjunctivitis. Survey of Aborigines in one area of southern Australia. Study population consisted of people who voluntarily presented for examination. Survey of Aborigines in west and central Australia. Method of sample selection is not clear. National random survey. Little information is provided on the sampling method. No definition of “rural” and “urban” is provided. Trachoma survey in one county. Does not present prevalence of trachomatous blindness Sample survey in suburban and rural areas. No information on sampling method is provided. WHO endorsed examination method was used. Blindness (low vision) definition Comments 353 1 038 1 321 711 NR 432 335 Blindness (< 3/60) 30 377 594 107 NR 47 32 continued Srivastava & Verma 1978 Chakrabarti, Gary & Siddhu 1974 Stocks, Newland & Hiller 1994 Taylor 1987 Hu et al. 1989 Zhang et al. 1992 World Health Organization 1985 Trachomatous Blindness (< 3/60) Reference Prevalence per 100 000 Review of available data on trachomatous blindness: population-based surveys after 1974 China CHI World Bank Region Country or area Table 10.1 Trachoma-Related Visual Loss 305 1986–1989 India West Bank & Gaza Strip MEC Brazil 1982–1983 1986 1981 India LAC Date of study 6 038 2 908 254 758 685 200 000 Population examined < 3/60 < 3/60 < 6/60 < 3/60 Random cluster sample in two geographically separate areas mainly populated by Palestinian Arabs. Most children under the age of 5 and a small number of older children and adults were unable to understand or cooperate in visual acuity testing. Authors state that the prevalence of trachomatous blindness in West Bank was twice that in Gaza strip. Random cluster sample in one municipality in southeastern Brazil. WHO Simplified Grading Scheme is used for assessing prevalence of trachoma infection. National random cluster survey. Sample sizes were determined based on the findings of the 1981 NSSO study. Clinical staging of trachoma not consistent with WHO classification. Results are inconsistent from one report to another. Only 11 cases of trachoma blindness (< 3/60) found. National census. Report includes no information on examination methods. Reports on blindness attributable to both trachoma and associated infection. Blindness (low vision) definition Comments 2 733 NR 697 250 Blindness (< 3/60) 364 0 4 50 Thomson & Chumbley 1984 Luna et al. 1992 Mohan 1989 NSSO survey 1981 Trachomatous Blindness (< 3/60) Reference Prevalence per 100 000 Review of available data on trachomatous blindness: population-based surveys after 1974 (continued) World Bank Region Country or area Table 10.1 306 Global Epidemiology of Infectious Diseases 1993 1989 1984 1981–1990 1991 1993 Morocco Pakistan Saudi Arabia Saudi Arabia Saudi Arabia Tunisia 1980–81 1990 Myanmar Nepal 1991–92 Mongolia OAI 1992 Morocco 39 887 21 103 4 345 3 547 2 882 4 268 14 76 6 690 4 797 8 878 < 3/60 < 3/60 < 3/60 < 3/60 < 3/60 < 3/60 < 3/60 < 3/60 — < 3/60 National random cluster survey. Most trachoma is located in the Far Western terai (western Nepal). Examination of nonrandomly selected villages belonging to 4 Regions of the Prevention of Blindness Programme. Random cluster sample in 3 of the 18 regions in Mongolia. Only people aged 40 years and older were examined. National, randomized cluster survey. Obvious bias (under representation of males 15–54 years of age). Random cluster sample in a single province. Definition of low vision used is unclear. Random cluster sample in a single province. This province was found to have higher levels of blindness than other regions of Saudi Arabia in the 1984 study. National random cluster sample representing the settled population of Saudi Arabia. Population-based survey in the North West Frontier Province. Obvious bias. Randomized cluster survey of a trachoma endemic area. Obvious bias. National, randomized, stratified, communitybased survey. 840 1,240 1,588 800 659 1 500 1 509 1 000 NR 760 23 62 0 34 0 141 160 90 NR 30 continued Brilliant et al. 1985 Than 1990 Baasanhu et al. 1994 Tunisia 1993 Al Faran et al. 1993 Badr et al. 1992 Tabbara & Ross-Degnam 1986 Pakistan Ministry of Health 1989 Morocco Ministry of Public Health 1992 Chami-Khazraji, Négrel & Ottmani 1994 Trachoma-Related Visual Loss 307 1991 1989 1990 Tonga Vanuatu Viet Nam 1984–1985 1988 1981 1986 1991 Congo Ethiopia The Gambia Ghana 962 874 1 383 6 185 5 002 1 841 1991 Burkina Faso Chad 7 047 1990 15 071 3 520 4 056 Population examined Benin SSA Date of study < 3/60 < 3/60 < 3/60 < 3/60 < 3/60 — Random cluster sample in one district. Conducted in a savannah area in central Ghana. Study did not include individuals younger than 30 years old. Only 67% of the sample population were examined. National random cluster sample. Well-conducted study with no obvious source of bias. Random cluster sample in one province. Conducted in a hot, dry, dusty province. The terms “urban” and “rural” are poorly defined. National random cluster sample. Little information regarding the sampling and examination methods is provided. National random cluster sample. Little information regarding the sampling and examination methods is provided. Randomized cluster survey of one region. National random cluster sample. Authors state that trachoma is only prevalent in the northern part of the country. Survey carried out in 4 northern and 4 southern provinces. No information in report as to methods used for sampling or examination. < 3/60 < 3/60 National random cluster sample. Only included persons aged 6 years or older. National random cluster survey. Study was restricted to persons aged 20 years and over. < 3/60 < 3/60 Blindness (low vision) definition Comments 2 495 700 940 307 2 310 NR 600 860 369 468 Blindness (< 3/60) 0 119 434 0 526 NR 0 27 0 25 Moll et al. 1994 Faal et al. 1989 Cerruti et al. 1981 World Health Organization 1990 World Health Organization 1987 Hutin et al. 1992 World Health Organization 1991 Nguyen et al. 1992 Newland et al. 1992 Newland et al. 1994 Trachomatous Blindness (< 3/60) Reference Prevalence per 100 000 Review of available data on trachomatous blindness: population-based surveys after 1974 (continued) World Bank Region Country or area Table 10.1 308 Global Epidemiology of Infectious Diseases 1990 1976 1976 1983 1985 1985 1990 1988–1989 Kenya Kenya Kenya Malawi Mali Mali Niger Nigeria 6 831 2 958 3 299 3 538 1 574 13 803 844 895 < 3/60 <3/60 < 3/60 < 3/60 < 3/60 < 3/60 < 3/60 < 3/60 Surveys of nonrandomly selected rural communities. Study conducted in an onchocerciasis mesoendemic area in northern Nigeria. Study did not include children younger than 5 years. Would be a good study for looking at sex ratio, but does not report total numbers of males and females examined. Community-based, randomized cluster survey. Obvious bias. Randomized, community based survey of a rural region. This thesis study was conducted in the Mopti Region. Randomized, community based survey of a rural region. This thesis study was conducted in the Kayes Region. Surveys of randomly selected villages in one region. Conducted in a region of central Malawi where blindness is particularly common. People younger than 6 years were not examined. Random cluster sample in 8 rural areas. Simple random sample at two nonrandomly selected sites. In central Kenya. Methods used for determining cause of blindness are not provided. Definitions of “active” and “inactive” disease are not provided. Study is too small to reliably assess prevalence of trachomatous blindness. Surveys conducted at nonrandomly selected sites in one region. Conducted in a semi-desert area in north-west Kenya. Study is too small to reliably assess prevalence of trachomatous blidness. 3 323 1 300 1 000 1 300 1 271 700 1 659 1 117 278 400 340 369 0 130 118 223 Abiose 1994 Kabo 1990 Boré 1985 continued Yattassaye 1985 Chirambo et al. 1986 Whitfield et al. 1990 Whitfield et al. 1983 Loewenthal & Peer 1990 Trachoma-Related Visual Loss 309 1979 1985 1986 South Africa South Africa Tanzania NR = not reported 1981–1986 1976 South Africa Togo Date of study 11 081 1 827 18 962 1 749 1 207 Population examined < 3/60 < 3/60 < 3/60 < 6/60 < 6/60 Random cluster surveys of 4 rural areas. WHO sampling procedures. Prevalence of trachomatous blindness is only available for 1 of 4 areas. Random cluster sample in an area known to be hyperendemic for trachoma.Visual acuity was only measured in individuals older than 7 years Examination of all residents in randomly selected villages in one district. District has an ophthalmic hospital. Far fewer males examined than females. Random cluster sample in a rural community. Conducted in an area known to be trachoma endemic. Sample not representative; working age males underepresented. Random cluster sample in a rural community. Conducted in an area known to be trachoma endemic. Sample not representative; working age males underepresented. Blindness (low vision) definition Comments NR 1 259 575 NR NR Blindness (< 3/60) NR 328 58 610 331 World Health Organization 1989 Rapoza et al. 1991 Bucher & Ijsslmuider 1988 Ballard et al. 1983 Ballard, Sutter & Fotheringhan 1978 Trachomatous Blindness (< 3/60) Reference Prevalence per 100 000 Review of available data on trachomatous blindness: population-based surveys after 1974 (continued) World Bank Region Country or area Table 10.1 310 Global Epidemiology of Infectious Diseases Trachoma-Related Visual Loss 311 based surveys in which the criteria for blindness are clearly defined. Only studies conducted after 1974 are included. In Table 10.1, blindness and trachomatous blindness refer to best corrected visual acuity less than 3/60. Values from studies that used different definitions of blindness have been adjusted based on the following assumptions: the prevalence of best corrected visual acuity less than 6/60 is 1.5 times the prevalence of best corrected visual acuity less than 3/60; the prevalence of best corrected visual acuity less than or equal to 6/60 is 2 times the prevalence of best corrected visual acuity less than 3/60 (Thylefors et al. 1995). Only 11 country-wide, population-based surveys (13 studies) were found to have looked at the prevalence of trachomatous blindness. Estimation of the burden of trachoma The calculation of disability adjusted life years (DALYs) lost as a result of trachomatous blindness and low vision requires the following information: age-specific and sex-specific values of the prevalence of trachomatous blindness and low vision for each of the eight GBD regions; the increased risk of mortality incurred by trachomatous blindness or visual loss; and disability weights for both blindness and low vision, which express the fraction of a year of life lost because of having to live for that year with disability (disability weights are discussed in the next section. In order to develop regional prevalence values for trachomatous blindness, countries were grouped by the authors according to the following scale: 0 = no known trachoma; 1 = trachoma infection, but no trachomatous blindness; 2 = trachoma infection, low prevalence of trachomatous blindness (≤ 100 per 100 000); 3 = trachoma infection, high prevalence of trachomatous blindness (> 100 per 100 000). For each region with endemic blinding trachoma, estimates based on the 11 available country studies and the country populations concerned were then applied for categories 2 and 3 above. Box 10.2 sets out these estimates. To consider both blindness and low vision caused by trachoma, the above-mentioned comprehensive review of available population-based surveys was used to determine the ratio of visual loss to blindness. It was found, based on 13 studies, that for each case of trachomatous blindness 312 Box 10.2 Global Epidemiology of Infectious Diseases Regional prevalence per 100 000 for trachomatous blindness according to prevalence level World Bank Region Low prevalence (category 2) High prevalence (category 3) SSA 75 203 MEC 31 160 OAI 26 — CHI 47 — IND 4 — there are 1.16 individuals with trachomatous visual loss (Table 10.2). This proportion has subsequently been applied in the assessment of low vision attributable to trachoma in each region and globally. Data analysis based on seven studies gave a sex ratio of 1.00:2.86 for male-to-female prevalence of trachomatous blindness. This ratio was applied uniformly to all countries with trachoma-related blindness and low vision (Table 10.3). Furthermore, the age distribution of the prevalence of blindness and low vision was calculated based on four studies, and the final result was applied to all endemic countries or areas (Table 10.4). Regional prevalence values for trachomatous blindness and low vision are shown in Table 10.5. Tables 10.6 to 10.11 provide age-specific and sex-specific values for the prevalence of trachomatous blindness in each of the regions and globally. Table 10.2 Ratio of people with low vision (< 6/18 – 3/60) to people with blindness (< 3/60) Population examined Number of Blind (< 3/60) Number with low vision (</6/18–3/60) Ratio of low visioned to blind Zgang et al. 1992 Thomson & Chumbley 1984 Chami-Khazraji et al. 1992 Pakistan Ministry of Health x1989 Badr et al. 1992 Tunisia 1993 Whitfield et al. 1983 Whitfield et al. 1990 Yattassaye 1985 Boré 1985 Abiose 1994 Rapoza et al. 1991 1 579 316 6 038 8 878 6 690 742 22 3 6 869 11 3 1 1.17:1.00 0.50:1.00 1.30:1.00 0.17:1.00 4 268 3 547 844 13 803 3 538 3 299 6 831 1 927 6 1 1 18 13 11 19 6 8 1 3 46 10 11 12 3 1.33:1.00 0.99:1.00 3.00:1.00 2.54:1.00 0.78:1.00 1.00:1.00 0.63:1.00 0.50:1.00 Totals 1 645 926 848 981 1.16:1.00 Reference Trachoma-Related Visual Loss Table 10.3 313 Ratio of female prevalence of trachomatous blindness to male prevalence Prevalence of trachomatous blindness per 100 000 Reference Taylor 1987 Tabbasa & Ross-Degnan 1986 Bucher & Ijsselmuiden 1958 Total Table 10.4 Age (years) Males Females Ratio of female prevalence to male prevalence 62 146 13 147 173 89 2.37 :1.00 1.19 :1.00 6.81:1.00 37 104 2.86 :1.00 Age distribution of the prevalence of trachomatous blindness Total population surveyed for visual impairmenta 0–4 5–14 15–44 45–59 60+ Total Trachomatous blindness Number blinda Prevalence per 100 000 Distribution of prevalence (%) 4 280 6 459 9 697 1 500 1 240 0 1 6 8 30 8 11 67 544 2 419 0 0 2 18 80 23 176 46 3 049 100 a. Numbers derived from Ballard et al. (1983), Tabbara & Ross-Degnan (1986), Rapoza et al. (1991), and Badr et al. (1992). Table 10.5 Prevalence of trachomatous blindness and low vision by region Trachomatous blindness Number Number 533 096 54 616 386 0 0 0 0 346 237 0 0 0 0 India Latin America & the xCaribbeanc Middle Eastern xCrescent 849 514 4 37 813 5 43 721 444 295 0 0 0 0 503 075 73 364 928 844 21 944 Other Asia & Islands 682 534 10 67 497 11 78 042 Sub-Saharan Africa 510 274 93 472 663 107 546 511 5 268 412 28 1 475 997 32 1 706 604 China Established Market xEconomiesa Formerly Socialist xEconomies of Europeb Total Prevalence per 100 000 1 134 693 47 797 790 Trachomatous low vision Prevalence per 100 000 World Bank Region 1990 Population (thousands) a. Blinding trachoma is known to exist only among Australian Aborigines, who are not considered representative of the region. b. Pockets of trachoma may exist but no information on trachomatous blindness could be identified. c. Trachoma is known to be endemic in parts of Brazil, Guatemala, Mexico and Peru but its blinding propensity is considered insignificant. 586 199 Total 548 494 57 946 90 401 284 078 64 403 51 666 Females Population (thousands) 1 134 693 118 189 187 398 591 383 137 077 100 646 Total 145 072 388 024 0 0 37 616 76 752 273 656 Females Total 533 096 0 0 53 883 112 406 366 807 Number Blinded by Trachoma 0 0 15 265 32 489 97 318 Males 167 738 0 0 17 650 37 565 112 523 Males 448 648 0 0 43 493 88 743 316 411 Females Males 59 789 101 752 200 525 47 567 29 768 439 401 Age (years) 0–4 5–14 15–44 45–59 60+ Total 410 113 56 679 95 263 183 242 46 005 28 924 Females Population (thousands) Total 849 514 116 468 197 015 383 767 93 572 58 692 10 309 0 0 1 136 2 426 6 747 Males 27 504 0 0 2 873 6 492 18 140 Females Number blinded by trachoma 37 813 0 0 4 061 8 911 24 841 Total 11 920 0 0 1 314 2 805 7 801 Males 31 802 0 0 3 322 7 506 20 74 Females 43 721 0 0 4 695 10 303 28 723 Total Number with low vision attributable to trachoma 616 386 0 0 6 301 129 968 424 116 Total Number with low vision attributable to trachoma Age and sex distribution of the people with blindness or low vision attributable to trachoma, India, 1990 60 243 96 997 307 305 72 674 48 980 0–4 5–14 15–44 45–59 60+ Table 10.7 Males Age and sex distribution of the people with blindness or low vision attributable to trachoma, China, 1990 Age (years) Table 10.6 314 Global Epidemiology of Infectious Diseases 256 389 Total 246 686 39 734 61 999 107 211 22 292 15 450 Females Population (thousands) 503 075 80 895 127 344 221 106 44 629 29 101 Total 97 304 267 624 0 0 30 991 57 994 178 40 Females Total 364 928 0 0 45 158 82 033 237 738 Number blinded by trachoma 0 0 12 872 22 720 61 712 Males 112 507 0 0 14 883 26 270 71 354 Males 309 437 0 0 35 832 67 055 206 550 Females 421 944 0 0 52 213 94 850 274 881 Total Number with low vision attributable to trachoma Males 43 763 84 032 160 825 34 138 20 208 342 966 Age (years) 0–4 5–14 15–44 45–59 60+ Total 339 568 41 988 80 217 159 611 35 090 22 662 Females Population (thousands) Total 682 534 85 751 164 249 320 436 69 228 42 870 17 622 0 0 2 220 4 242 11 160 Males 49 875 0 0 5 761 11 398 32 716 Females Number blinded by trachoma 67 497 0 0 8 136 15 820 43 540 Total 20 375 0 0 2 567 4 905 12 903 Males 57 667 0 0 6 661 13 179 37 828 Females 78 042 0 0 9 407 18 292 50 343 Total Number with low vision attributable to trachoma Age and sex distribution of the people with blindness or low vision attributable to trachoma, Other Asia and Islands, 1990 41 161 65 345 113 895 22 337 13 651 0–4 5–14 15–44 45–59 60+ Table 10.9 Males Age and sex distribution of the people with blindness or low vision attributable to trachoma, Middle Eastern Crescent, 1990 Age (years) Table 10.8 Trachoma-Related Visual Loss 315 47 484 70 258 103 764 20 308 10 508 252 322 0–4 5–14 15–44 45–59 60+ Total 257 952 47 030 69 818 106 257 22 117 12 730 Females Population (thousands) 510 274 94 514 140 076 210 021 42 425 23 238 Total 120 505 352 158 0 0 45 941 86 062 220 155 Females Total 472 663 0 0 65 250 118 627 288 786 Number blinded by trachoma 0 0 17 690 31 159 71 656 Males 139 333 0 0 20 454 36 027 82 852 Males 407 178 0 0 53 118 99 508 254 552 Females 546 511 0 0 75 445 137 161 333 906 Total Number with low vision attributable to trachoma 2 654 692 Total Males 321 304 551 205 1 250 941 312 385 218 857 0–4 5–14 15–44 45–59 60+ Age (years) 2 613 720 309 253 525 543 1 198 642 311 067 269 215 Females Population (thousands) Total 5 268 412 630 557 1 076 748 2 449 583 623 452 488 072 390 813 0 0 49 183 93 036 248 594 Males 1 085 184 0 0 123 181 239 697 723 306 Females Total 1 475 997 0 0 176 487 337 796 961 713 Number blinded by trachoma 451 872 0 0 56 867 107 572 287 433 Males 1 254 731 0 0 142 427 275 990 836 14 Females 1 706 604 0 0 204 061 390 573 1 111 969 Total Number with low vision attributable to trachoma Table 10.11 Age and sex distribution of the people with blindness or low vision attributable due to trachoma throughout the world, 1990 Males Age (years) Table 10.10 Age and sex distribution of the people with blindness or low vision attributable to trachoma, Sub-Saharan Africa, 1990 316 Global Epidemiology of Infectious Diseases Trachoma-Related Visual Loss 317 The increased risk of mortality associated with blindness and low vision is derived from a study conducted in Africa (Kirkwood et al. 1983), in the absence of data from other regions. This study found the standardized mortality rate to be 3.8 times higher among females blind as a result of all causes and 2.5 times higher among blind males than among normal sighted. The standardized mortality rate was found to be 1.5 times higher among females with low vision and 1.4 times higher among males with low vision than among normal sighted controls. Disability The rating of blindness in the grading of severity of disability needs to be adjusted. Blindness implies severe repercussions for daily living activities, as defined in category 6; all blind people should therefore be classified in category 6, and people with low vision in category 5. Other considerations The pathway to blindness from trachoma is not specific, in the sense that the corneal opacification induced by trachomatous trichiasis may be the result of secondary infections. Therefore, if there is not enough awareness of the endemicity of trachoma in an area, it is likely that there will be an underestimate as to the role of trachoma as a cause of blindness. Furthermore, in certain rural settings, particularly in north Africa, there are regular seasonal epidemics of conjunctivitis (Kupta, Nicetic & Reinhards 1968, Reinhards et al. 1968), which is transmitted by flies in great numbers. There is a known interaction between conjunctivitis epidemics and the increased prevalence and intensity of trachoma in such populations, but this fact may easily be overlooked by local health staff. A majority of victims of trachomatous blindness are women because of their prolonged and close contact with affected children, who act as a reservoir of Chlamydia. The impact of severe trachoma can therefore be considerable in making women of working age unable to fulfil their family and socioeconomic role. Disease as risk factor for other diseases Trachoma in its advanced stage, with corneal complications, increases considerably the risk of secondary corneal infections, which often lead to ulcers and severe loss of vision. Long-standing cases of trachoma are likely to develop the “dry eye” syndrome because of tear gland involvement; this, again, acts as a further risk factor for blindness from infections as a result of increased susceptibility of the eye. Furthermore, it should be noted that cases of advanced trachoma with corneal complications are difficult and high risk cases for eye surgery such as corneal transplants or cataract or glaucoma surgery. 318 Global Epidemiology of Infectious Diseases It has been demonstrated in Africa and North America that blindness is associated with increased mortality (Hirsh & Schwartz 1983, Kirkwood et al. 1983). This may be a result of increased vulnerability, as a consequence of neglect, absence of social care and increased risk of accidents; it may also be that blindness attributable to certain causes, such as cataract, is a marker of ageing and thus an increased mortality. Burden of trachoma and intervention Treatment of trachoma should be carried out on a priority basis in communities with blinding trachoma. The objective of treatment is normally limited to elimination of blindness from trachoma, as it is neither feasible to eradicate the disease nor to eliminate the chlamydiae in the populations concerned. Treatment of trachoma is based on antibiotic topical or systemic treatment or both, surgery for inturned eyelids, and health education pertaining to improved personal and environmental hygiene. Given that trachoma is often highly endemic in affected rural communities, the antibiotic treatment is mainly a suppressive scheme (Dawson, Jones & Tarizzo 1981) to reduce disease intensity; reinfections rapidly occur in the communities or families concerned unless other measures to improve living standards and behaviour occur (West et al. 1991). Thus, the effectiveness of trachoma antibiotic treatment is limited to disease suppression, not to elimination. Trichiasis surgery has recently been systematically evaluated (Reacher et al. 1992) and a uniform method proposed. Access to trachoma treatment for affected populations must be provided regularly and in a long-term (several years) perspective. The annual cost of preventing blindness from trachoma can schematically be calculated as the cost of one tube (5 gram) of tetracycline eye ointment per person per year; at present, this amounts to approximately 15 US cents in a bulk purchase. The preventive treatment should be provided throughout childhood if the level of intense inflammatory disease remains high in the community concerned. Distribution costs will depend on how community-based the control programme is in terms of primary health care, or volunteer or self-treatment options. A more comprehensive model for the cost of preventive topical or systemic treatment of trachoma has been put forward (Dawson & Schachter 1985). Trichiasis surgery can commonly be made available for a cost of approximately US$5–10 (consumables and partial cost of instruments). Cessation of the current trachoma control activities could potentially result in a doubling of the incidence of blindness from trachoma over a period of 10–20 years. In contrast, if currently available interventions were systematically applied in all endemic areas, it might be possible to eliminate trachoma as a cause of blindness over the next 10 years. This would necessitate strong governmental commitment with allocation of needed resources, as well as environmental improvements (water and sani- Trachoma-Related Visual Loss 319 tation) where needed, and behavioural changes for improved hygiene. A new and more effective antibiotic treatment could substantially improve the perspective of global trachoma control such as may be possible with azithromycin, a new and long-acting macrolide. Discussion Trachoma has been subject to numerous attempts of control in the past. Various medical and surgical treatment schemes have been used over the centuries to get rid of the characteristic follicles on the inside of the eyelids, from copper-sulphate etching in pharaonic Egypt, to surgical compression of follicles with special forceps well into the 20th century. On a public health level, trachoma was recognized as an important infectious eye disease and cause of blindness; it was a matter of concern in relation to population movements in Europe, and in the past immigration to the United States of America in the 19th and early 20th centuries. Effective public health measures specifically against trachoma were not taken until after the Second World War. By then antibiotics were becoming available, and when the WHO started its work in the late 1940s, trachoma was one of the first diseases subject to field research on more effective large-scale control measures. After a number of field studies during the 1950s, particularly in North Africa, a control strategy was designed consisting of: antibiotic treatment with 1 per cent tetracycline eye ointment, which could be applied on an intermittent basis; access to simplified and standardized trichiasis surgery; health education. This strategy was consistently implemented over the following two decades by the WHO, often in partnership with UNICEF , in most of the endemic countries. The results were often quite positive, particularly in areas of socioeconomic development, where there were improved living standards and opportunities for changes in lifestyle. This is not surprising; trachoma disappeared spontaneously in several European countries as a result of better hygiene and living standards well before the antibiotic era. Suppressive treatment of trachoma with antibiotics allows for rapid control of the blinding intensity of infection. That element of the strategy is very important in terms of preventing blindness, particularly in areas where there are no immediate prospects of improving living standards. The global picture of trachoma has changed considerably over the past few decades as a result of the above-mentioned disease control efforts and socio-economic developments. As the disease is so closely linked to the social and public health situation, with poverty, crowding, and lack of hygiene as major propagating factors, the epidemiological pattern varies accordingly. This leads in many settings to difficulties in properly 320 Global Epidemiology of Infectious Diseases assessing trachoma and its blinding propensity. The distribution of the disease is often quite focal, with variations in its intensity. Furthermore, significant differences in disease patterns can occur rapidly, as a result of sudden behavioural or environmental changes, for example urbanization or access to tap water. Today trachoma is largely confined to the poorest population groups in some 40 developing countries that are still endemic. The disease has been pushed back in many countries, and is now found mainly in underserved rural areas where there is little hope of socioeconomic development. This makes it imperative to mount interventions for the prevention of blindness from trachoma in these particularly vulnerable populations, as it may take many years before the benefits of improved living conditions wipe out the disease as a cause of visual loss. For the reasons given above, the present global estimate of trachoma given in this report can only be indicative. It is based on patchy epidemiological data, indicating only the likely magnitude of trachoma-related visual loss in known endemic countries. Still, in the absence of the large-scale population-based surveys that would be needed to derive a more reliable estimate, the present estimate is probably the most reasonable, recognizing the limitation of not having more accurate and updated information from several major populations where recent disease changes are likely to have occurred (e.g. China and India). It should also be noted that the present estimate is very conservative, as it is based on a strict definition of visual loss only in conjunction with trichiasis. This definition, although specific, does not take into account other pathways to blindness from trachoma, e.g. invasive pannus or corneal ulceration, often leading to a condition of phthisis bulbi, which can have multiple origins. It is important to recognize the association of blinding trachoma with low standards of living; this allows for the focusing of interventions on the poorest population groups, where trachoma is likely to be a very significant cause of blindness and an obstacle to community development. The SAFE strategy (surgery; antibiotics; facial cleanliness; environmental hygiene) includes all the essential elements for successful trachoma control. The surgery for trichiasis can prevent new cases of blindness; the antibiotics can control the infection; and the personal and environmental hygiene provide the answer to sustainability of achievements in the future. The recent availability of much-improved antibiotic treatment with azithromycin opens up new perspectives for rapid and effective control of trachomatous infection in the target populations, thus allowing for an accelerated implementation of the SAFE strategy where needed. Conclusions Trachoma is the most important cause of easily preventable visual loss in the world. It adds significantly to the disability burden amongst vulnerable poor population groups, where the disease is commonly found today. Trachoma-Related Visual Loss 321 Trachoma control efforts, together with socioeconomic progress, have pushed back the disease in many countries, but in remaining endemic areas there is still a great need for the prevention of blindness from trachoma. A comprehensive intervention strategy, including improved antibiotics, holds out the promise of eliminating trachoma as a public health problem. References Abiose A (1994) Distribution and aetiology of blindness and visual impairment in mesoendemic onchocercal communities, Kaduna State, Nigeria. British Journal of Ophthalmology, 78:8–13. Al Faran MF et al. (1993) Prevalence and causes of visual impairment and blindness in the south western region of Saudi Arabia. International Ophthalmology, 17:161–165. Badr IA et al. (1992) Changing patterns of visual loss in the Eastern Province, Kingdom of Saudi Arabia, Saudi Journal of Ophthalmology, 6:59–68. Baasanhu J et al. (1994) Prevalence and causes of blindness and visual impairment in Mongolia: a survey of populations aged 40 and older. Bulletin of the World Health Organization, 72:771–776. Ballard RC, Sutter EE, Fotheringham P (1978) Trachoma in a rural South African community. American Journal of Tropical Medicine and Hygiene, 27:113–120. Ballard RC et al. (1983) Trachoma in South Africa. Social Science and Medicine, 17:1755–1765. Bietti GB, Freyche MJ, Vozza R (1962) La diffusion actuelle du trachome dans le monde. Revue internationale du Trachome, No. 2-3:114–310. Boré, O. (1985) Thèse Doctorat en Médecine. École nationale de Médecine et de Pharmacie du Mali. [Doctoral thesis in Medicine. National school of Medicine and Pharmacology of Mali]. Brilliant LB et al. (1985) Epidemiology of blindness in Nepal. Bulletin of the World Health Organization, 63:375–386. Bucher PJM, Ijsselmuiden, CB (1988) Prevalence and causes of blindness in the Northern Transvaal. British Journal of Ophthalmology, 72:721–726. Cerruli L et al. (1981) Preliminary epidemiological study on the causes of blindness in one region of Ethiopia. Revue Internationale du Trachome et de Pathologie Oculaire Tropicale et Subtropicale, 3:117–129. Chakrabarti J, Garg AC, Siddhu CMS (1974) Prevalence and pattern of blindness in the field practice area of a rural health training center. Indian Journal of Ophthalmology, 22:4–8. Chami-Khazraji Y, Akalay O, Négrel AD, (1992) Prévalences et causes de la cécité et de la vision au Royaume de Maroc. Résultats de l’enguête nationale sur les causes de la prévalence des déficiences visuelles (ENCPDV). Rapport final. [Prevalence and causes of blindness and low vision in Morocco. Results of the national survey on the causes of vision loss. Final report] Morocco, Morocco Ministry of Public Health. 322 Global Epidemiology of Infectious Diseases Chami-Khazraji Y, Négrel AD, Ottmani S (1994) Prévalence et gravité du trachome dans les provinces de Tata, Figuig et Errachidia [Prevalence and severity of trachoma in the provinces of Tata, Figuig and Errachidia]. Ministère de la Santé publique, Programme OMS de Prévention de la Cécité. Chirambo MC et al.(1986) Blindness and visual impairment in southern Malawi. Bulletin of the World Health Organization, 64:567–572. Dawson CR, Schachter,J (1985) Strategies for treatment and control of blinding trachoma: cost-effectiveness of topical or systemic antibiotics. Review of Infectious Diseases, 7:768–773. Dawson CR, Jones BR, Tarizzo ML (1981) Guide to trachoma control. Geneva, World Health Organization. Faal H et al. (1989) National survey of blindness and low vision in The Gambia: results. British Journal of Ophthalmology, 73:82–87. Hirsh R, Schwartz B (1983) Increased mortality among elderly patients undergoing cataract extraction. Archives of Ophthalmology, 101:1034–1037. Hu et al. (1989) An epidemiologic survey of trachoma in Shunyi County, Beijing. Chinese Journal of Ophthalmology, 25:175–178. Hutin YJF et al. (1992) Le trachome dans la Région de Sabou (Burkina Faso). Une enquête épidémiologique [Trachoma in Sabou (Burkina Faso). An epidemiological investigation]. Bulletin de la Societé de Pathologie Exotique, 85:350–354. Jitta NM, Luharis A (1930) Rapport sur le trachome dans différents pays. Geneva, Société des Nations, Organisation d’Hygiène. Kabo (1990) Ministry of Public Health, Niger, (unpublished data). Kirkwood B et al. (1983) Relationship between mortality, visual acuity and microfilarial load in the area of the Onchocerciasis Control Programme. Transactions of the Royal Society of Tropical Medicine and Hygiene, 77:862–868. Kupka K, Nižeti B, Reinhards J (1968) Sampling studies on the epidemiology and control of trachoma in southern Morocco. Bulletin of the World Health Organization, 39:547–566. Loewenthal R, Peer J (1990) A prevalence survey of ophthalmic diseases among the Turkana tribe in north-west Kenya. British Journal of Ophthalmology, 74:84–88. Luna EJA et al. (1992) Epidemiology of trachoma in Bebedouro State of Sao Paulo, Brazil; prevalence and risk factors. International Journal of Epidemiology, 21:169–177. Mohan M (1989) Epidemiological studies on prevalence of trachoma and impact of trachoma control programme in India. (unpublished data). Moll, AC et al. (1994) Prevalence of blindness and low vision of people over 30 years in the Wenchi district, Ghana, in relation to eye care programmes. British Journal of Ophthalmology, 78:275–279. Newland HS et al. (1994) Epidemiology of blindness and visual impairment in the Kingdom of Tonga. British Journal of Ophthalmology, 78:344–348. Newland HS, et al. (1992) Epidemiology of blindness and visual impairment in Vanuatu. Bulletin of the World Health Organization, 70:369–372. Trachoma-Related Visual Loss 323 Nguyen et al. (1992) Cataract Intervention Programme in Vietnam. (unpublished data). National Sample Survey Organization (NSSO) survey (1981) (unpublished data). Pakistan Ministry of Health (1989) (unpublished data). Rapoza PA et al. (1991) Prevalence and causes of vision loss in central Tanzania. International Ophthalmology, 15:123–129. Reacher MH et al. (1992) A controlled trial of surgery for trachomatous trichiasis of the upper lid. Archives of Ophthalmology, 110:667–674. Reinhards J et al. (1968) Studies in the epidemiology and control of seasonal conjunctivitis and trachoma in southern Morocco. Bulletin of the World Health Organization, 39:497–545. Stocks NP, Newland H, Hiller J (1994) The epidemiology of blindness and trachoma in the Anangu Pitjantjatjara of South Australia. Medical Journal of Australia, 160:751–756. Srivastava RN, Verma BL (1978) An epidemiology study of blindness in an Indian rural community. Journal of Epidemiology and Community Health, 32:131–135. Tabbara KF, Ross-Degnan D (1986) Blindness in Saudi Arabia. JAMA, 255: 3378–3384. Than KM (1990) The model village eye health survey. (unpublished data). Taylor HR (1987) Prevalence and causes of blindness in Australian Aborigines. Medical Journal of Australia, 1:71–76. Thomson IM, Chumbley LC (1992) Eye disease in the West Bank and Gaza Strip. British Journal of Ophthalmology, 68:598–602. Thylefors B, Négrel AD, Pararajasegaram, R (1992) La surveillance épidémiologique de trachome; bilan et perspective [Epidemiological surveillance of trachoma: evaluation and perspective]. Revue internationale du Trachome et de Pathologie Oculaire, 69:107–114. Thylefors et al. (1987) A simple system for the assessment of trachoma and its complications. Bulletin of the World Health Organization, 65:477–483. Thylefors, B et al. (1995) Global data on blindness: an update. Bulletin of the World Health Organization, 73:115–121. Thylefors B (1987) A simplified methodology for the assessment of blindness and its main causes. World Health Statistics Quarterly, 40:129–141. Tunisia National study (1993) (unpublished data). West SK, et al. (1991) Facial cleanliness and risk of trachoma in families. Archives of Ophthalmology, 109:855–857. Whitfield R et al. (1983) Cataract and corneal opacity are the main causes of blindness in the Samburu Tribe of Kenya. Ophthalmic Surgery, 14:139–144. Whitfield R et al. (1990) Blindness and eye disease in Kenya: ocular status survey results from the Kenya Rural Blindness Prevention Project. British Journal of Ophthalmology, 74:333–340. 324 Global Epidemiology of Infectious Diseases World Health Organization (1959) Scientific Group on Trachoma Research, London, Paris, Geneva. World Health Organization. (1962) Expert Committee on Trachoma. Third report. Geneva, WHO (Technical Report Series No. 234). World Health Organization (1977) International Classification of Diseases, 1975 revision, Vol. 1. WHO, Geneva. World Health Organization (1984) Strategies for the prevention of blindness in national programmes. Geneva, World Health Organization. World Health Organization (1985) Prevention of blindness: China. WHO Weekly Epidemiological Record, 60:369–371. World Health Organization. (1987) Prevention of blindness. Weekly Epidemiological Record, 43:322–323. World Health Organization (1989) Prevalence surveys on the causes of blindness and visual loss: Togo. Weekly Epidemiological Record, 12:87–90. World Health Organization. (1990) Prevention of blindness: prevalence and causes of blindness and visual loss Congo. Weekly Epidemiological Record, 65:249–251. World Health Organization. 1991) Prevention of blindness: prevalence and causes of low vision: Benin. Weekly Epidemiological Record, 66:337–340. World Health Organization (1992) International Statistical Classification of Diseases and Related Health Problems, vol. 1, Geneva, WHO World Health Organization. Yattassaye M (1985) Thèse Doctorat en Médecine. École national de Médecine et de Pharmacie du Mali. [Doctoral thesis in Medicine. National school of Medicine and Pharmacology of Mali]. Zhang et al. (1992) National epidemiological survey of blindness and low vision in China. Chinese Medical Journal, 105:603–608. Chapter 11 Chagas Disease A Moncayo Introduction Chagas disease, named after the Brazilian physician Carlos Chagas who first described it in 1909, exists only on the American continent (Chagas 1909). It is caused by a flagellate protozoan parasite Trypanosoma cruzi transmitted to humans by blood-sucking triatomine bugs and by blood transfusion. Chagas disease has two successive phases: acute and chronic. The acute phase lasts six to eight weeks. Once the acute phase subsides, most infected patients recover an apparently healthy status, in which no organ damage can be demonstrated by the current standard methods of clinical diagnosis. The infection can be verified only by serological or parasitological tests. This form of the chronic phase of Chagas disease is called the indeterminate form. Most patients remain forever in this form of the disease. However, several years after of starting the chronic phase, and depending on the geographical area, 10–40 per cent of the infected individuals will develop lesions of different organs, mainly the heart and the digestive system. These conditions are called the cardiac or digestive forms of chronic Chagas disease. The chronic phase lasts the rest of the life of the infected individual. Chagas disease is the primary cause of cardiac lesions in young, economically productive adults in the endemic countries in Latin America. It has been targeted for elimination by the World Health Assembly in Resolution WHA51.14, approved in May 1998 (World Health Organization 1998b). Transmission by vectors Chagas disease is a zoonosis transmitted in natural foci or ecological units within a well-defined geographical environment. The ecological unit is composed of sylvan or domestic mammals and of sylvan triatoma bugs, 326 Global Epidemiology of Infectious Diseases both infected with T. cruzi. Continuous transmission is assured with or without the involvement of human beings. These conditions of transmission are present from latitude 42°N to latitude 40°S. T. cruzi infection extends from the south of the United States to the south of Argentina. There are two phases of the human disease: the acute phase, which appears shortly after the infection; and the chronic phase, which may last several years and irreversibly affect the autonomic nervous system and internal organs, in particular the heart, the oesophagus and the colon, and the peripheral nervous system. After an asymptomatic period of several years, 27 per cent of those infected develop cardiac symptoms that may lead to chronic heart failure and sudden death, 6 per cent develop digestive damage, mainly mega colon and mega oesophagous, and 3 per cent suffer peripheral nervous involvement (Pereira, Willcox & Coura 1985, Coura, Anunziato & Willcox 1983, Coura et al.1985). Transmission via blood transfusion The rural-to-urban migration movements that occurred in Latin America in the 1970s and 1980s changed the traditional epidemiological pattern of Chagas disease as a rural condition and transformed it into an urban infection that can be transmitted by blood transfusion. In most countries in Latin America it is now compulsory to screen for infected blood in blood banks, and systems have been established to do so. Definition Chagas disease and its clinical forms are classified in the tenth revision of the International Classification of Diseases (ICD-10) under code B57. There have been no changes in the classification of the disease in the last three revisions of the ICD. Methods and Measurement Transmission by vectors Data on the prevalence and distribution of Chagas disease improved in quality during the 1980s as a result of the demographically representative cross-sectional studies carried out in countries where accurate information was not available. A group of experts meeting in Brasilia in 1979 devised standard protocols to carry out countrywide prevalence studies on human T. cruzi infection and triatomine house infestation. These studies were carried out during the 1980s in collaboration with the Ministries of Health of Chile, Colombia, Ecuador, Honduras, Panama, Paraguay, and Uruguay. The accurate information obtained has made it easier for individual countries to plan and evaluate the effectiveness of national control programmes Chagas Disease 327 (Pan American Health Organization 1974, Cedillos 1975, Marinkelle 1976, Zeledon 1976, Cordova et al. 1980, Camargo et al. 1984, Ponce 1984, Reyes Lituma 1984, Lopez 1985, Matta 1985, Schenone et al. 1985, Sousa 1985, Salvatella et al. 1989, Acquatella 1987, Valencia 1990) (see Table 11.1). On the basis of these individual countrywide surveys it is estimated that the overall prevalence of human T. cruzi infection in the 18 endemic countries is 17 million cases. Some 100 million people (25 per cent of all the inhabitants of Latin America) are at risk of contracting T. cruzi infection (see Table 11.2). The incidence was estimated in 1991 at 700 000–800 000 new cases per year and the annual deaths attributable to the cardiac form of Chagas disease at 45 000 (UNDP/World Bank/WHO 1991). The endemic countries can be divided into three groups according to such indicators as: number of confirmed human cases; prevalence of seropositive tests in blood donors and population samples; and the presence of infected vectors and reservoirs. The main factors on which these criteria are based are the following: magnitude of the transmission; quantity and quality of the available epidemiological information; and existence or absence of coordinated actions towards the control of this disease (UNDP/World Bank/WHO 1987). The groups are identified as described below. Table 11.1 Prevalence studies of human T. cruzi infection in Latin America Sample size Argentina Bolivia Brazil Chile Colombia Costa Rica Ecuador El Salvador Guatemala Honduras Mexico Nicaragua Panama Paraguay Peru Uruguaya Venezuela Percentage infected Year Reference Remarks 10.0 1976 Estimation 56 000 a 1 352 917 a 13 514 a 20 000 1 420 532 524 3 952 a 3 802 a nd nd 1 770 4 037 a 92 24.0 4.2 16.9 30.0 11.7 10.7 20.0 16.6 15.2 nd nd 17.7 21.4 9.8 1980 1980 1985 1975 1976 1984 1975 1982 1984 nd nd 1984 1983 1980 Pan American Health Organization 1974 Valencia 1990 Camargo et al. 1984 Schenone et al. 1985 Marinkelle 1976 Zeledon 1976 Reyes Lituma 1984 Cedillos 1975 Matta 1985 Ponce 1984 Sousa 1985 Lopez 1985 Cordova et al. 1980 5 924 5 696 3.4 3.0 1985 1983 Endemic areas Countrywide Age group 0–6 years Countrywide Age group < 24 years nd . . . . . . Salvatella et al. 1989 Acquatella 1987 a Demographic sample statistically representative of the country‘s population. nd = no data Countrywide Countrywide Countrywide Endemic areas Endemic areas Endemic areas Rural areas Countrywide Countrywide 328 Table 11.2 Global Epidemiology of Infectious Diseases Prevalence of human T. cruzi infection in Latin America, 1975–1985 Population at risk (thousands) Percentage of total population Number of infected people (thousands) Argentina Bolivia Brazil Chile Colombia Costa Rica Ecuador El Salvador Guatemala Honduras Mexico Nicaragua Panama Paraguay Peru Uruguay Venezuela 6 900 1 800 41 054 11 600 3 000 1 112 3 823 2 146 4 022 1 824 nd nd 898 1 475 6 766 975 12 500 23 32 32 63 11 45 41 45 54 47 nd nd 47 31 39 33 72 2 640 1 300 6 180 1 460 900 130 30 900 1 100 300 nd nd 200 397 621 37 1 200 Total 99 895 25 17 395 nd = no data. Group 1 The countries of this group present high prevalence of human T. cruzi infection and high triatomine house infestation rates. These conditions have induced the national health authorities to establish control activities through vertical programmes or, more recently, within the context of primary health care strategies. Argentina, Bolivia, Brazil, Chile, Colombia, Honduras, Paraguay, Peru, Uruguay and Venezuela are included in this group. Group 2 All the countries included in this group show evidence of intra-domiciliary transmission, with a clear association between T. cruzi infection and electrocardiographic alterations as well as other pathologies attributable to Chagas’ disease. No formal control programmes have yet been established in these countries. Costa Rica, Ecuador, and Mexico belong to this group. Group 3 These countries show evidence of domiciliary transmission, but more accurate epidemiological data are needed to support a clear correlation between T. cruzi infections and clinical records. In all these countries, the acute phase of Chagas disease is frequently observed, and recent serologi- Chagas Disease 329 cal data indicate that the prevalence of seropositive reactions to T. cruzi antigens is high. El Salvador, Guatemala, Nicaragua, and Panama are included in this group. Transmission via blood transfusion Table 11.3 shows the extent of the problem of transmission via blood transfusion in some selected Latin American cities between 1980 and 1989 (Schmunis 1991). The prevalence of T. cruzi infection in blood varies between 1.3 and 51.0 per cent, and is generally much higher than that of hepatitis or HIV infection. The transmission of Chagas disease via blood transfusion is a real threat even in countries where the disease is not transmitted by vector, such as the United States of America and Canada, where cases of acute Chagas disease have been recently documented (Kirchoff et al. 1987, Grant, Gold & Wittener1989, Nickerson et al.1989). Course of human infection Morbidity Controlled prospective studies that describe the clinical evolution of the cardiac form in T. cruzi-infected persons have been carried out in different geographical areas of the continent. They provide accurate data on the evolution of infection and clinical pathology of Chagas disease. Overall, 58 per cent of the cases of myocardiopathy are observed in young, economically productive adults, aged 20– 40 years. Signs and symptoms of congestive heart failure are observed in 65 per cent of these cardiac patients. Furthermore, 52 per cent of digestive system lesions, such as mega colon and mega oesophagus, are observed in this same age group. There are no differences in clinical manifestations by sex (Coura et al. 1983). Sudden death is registered in 64 per cent of myocardiopathy patients with a median age of 47 years, whereas congestive heart failure is observed in 36 per cent of cardiac patients with a median age of 78 years (Pereira, Willcox & Coura 1985) After 7 years of study of a Brazilian community in a prospective electrocardiographic survey, it was observed that cardiac lesions developed soon after infection and their incidence was highest in younger age groups. Most of those affected had become infected before 20 years of age. As shown in Table 11.4, the incidence rates per 1000 person-years of right bundle branch block (RBBB)—a characteristic conduction alteration of chronic Chagasic myocardiopathy which leads to severe arrhythmia and heart failure—were 19.2 in the age group 10–14 years, 3.5 in the age group 20–39 years and 2.8 in the age group 40–59 years among seropositive people who had had a previously normal electrocardiogram. In contrast, 330 Table 11.3 Global Epidemiology of Infectious Diseases Prevalence of Trypanasoma cruzi infected blood in blood banks of selected countries Number of samples tested Percentage positive 58 284 2 003 2 441 4.9 17.6 8.4 Perez & Segura 1989 Perez & Segura 1989 Perez & Segura 1989 205 51.0 Carrasco 1990 2 413 3 000 56 902 14.6 4.8 2.9 Pereira 1984 Marzochi 1981 Dias & Brener 1984 214 62 3.7 14.5 Liendo et al.1985 Liendo et al. 1985 1 128 491 2.5 7.5 Guhl et al. 1987 Guhl et al. 1987 602 1.6 Urbina et al. 1991 1 054 3.2 Ecuador Ministry of Health 1961 1 225 11.6 Ponce & Ponce 1987 200 17.5 Velasco Castrejon & Guzman Bracho 1986 562 11.3 Paraguay Ministry of Health 1972 329 12.9 Peru Ministry of Health 1972 445 71 699 4.7 4.2 7.7 Franca 1986 Franca 1986 Franca 1986 195 476 1.3 Schmunis 1991 Reference Argentina Buenos Aires (1987) Santiago del Estero(1987) Cordoba (1982) Bolivia Santa Cruz (1990) Brazil Brasilia (1984) Parana (1987) Sao Paulo (1982) Chile Santiago (1983) Vicuna (1983) Colombia Bogotá (1990) Cucuta (1987) Costa Rica San Jose (1985) Ecuador Guayaquil (1961) Honduras Tegucigalpa (1987) Mexico Puebla (1986) Paraguay Asunción (1972) Peru Tacna (1972) Uruguay Paysandu (1983–84) Salto (1983–84) Tacuarembo (1983–84) Venezuela Various cities in the seronegative control groups there was no development of this characteristic circulatory lesion (Maguire, Hoff & Sherlock 1987). In another 9-year prospective study the incidence of ventricular circulatory defects (VCDs) including RBBB was 23.8 per 1000 person-years in the Chagas Disease Table 11.4 Age group (years) 10–14 20–39 40–59 10–19 20–39 30–59 > 30 331 Incidence of cardiac lesions and T. cruzi infection, rates per 1000 person-years Rates per 1000 person-years ECG lesion Seropositive RBBB RBBB RBBB VCDs VCDs RBBB RBBB 19.2 3.5 2.8 23.8 88.2 4.6 25.8% Seronegative 0.0 0.0 0.0 3.3a 6.8a 0.0 8.6% Reference Maguire, Hoff & Sherlock 1987 Maguire, Hoff & Sherlock 1987 Maguire, Hoff & Sherlock 1987 Mota et al. 1990 Mota et al. 1990 Mota et al. 1990 Arribada, Apt & Ugarte 1986 RBBB = right bundle branch block ECG = electrocardiogram VCDs = ventricular conduction defects a. p < 0.01 seropositive patients aged 10–19 years and 88.2 per 1000 person-years in the seropositive patients aged 20–39 years, whereas the incidence of these defects was 3.3. and 6.8, respectively in the seronegative groups of the same ages (Mota et al. 1990). Finally, in Chile after a 4-year follow-up study it was also found that RBBB was three times more frequent among people who were seropositive for T. cruzi (25.8 per cent) than among those who were seronegative (8.6 per cent) and the difference was statistically significant (p < 0.01) (Arribada, Apt & Ugarte 1986). It is assumed that the evolution of the human infection towards the development of chronic cardiac lesions is similar in all endemic areas throughout the continent. Mortality In a recent study in Minas Gerais, Brazil, it was found that among all causes of death in a group young individuals aged 20–50 years old infected with T. cruzi, 21 per cent was attributable to chronic Chagasic cardiomyopathy; in contrast, only 9 per cent of deaths in patients infected with T. cruzi and who were older than 60 years were attributable to this condition (Menezes 1989). Mortality attributable to cardiac lesions in infected people is twice as high in males 20–40 years old (25 per cent) as in females of the same age group (11 per cent) (p < 0.01) (Coura et al. 1985). In another study, excess mortality rates in seropositive patients aged 30 years or younger were 71.4 per 1000 person-years, whereas no deaths were registered in the seronegative group of the same age (Mota et al. 1990). As indicated in Table 11.5, the mortality rate per 1000 person-years for seropositive patients with RBBB was 33.5 and with ventricular extra systoles (VEs) was 39.2, whereas with the two conditions combined (RBBB and VEs) the mortality rate rose to 116.3 per 1000 person-years. 332 Table 11.5 Global Epidemiology of Infectious Diseases Excess mortality rates with cardiac lesions, rates per 1000 person-years ECG lesion RBBB VEs RBBB and VEs VEs (< 30 years) VEs (> 30 years) Seropositive Seronegative 33.5 39.2 116.30 71.4 67.6 0.0 0.0 0.0 0.0 0.0 Reference Maguire, Hoff & Sherlock 1987 Maguire, Hoff & Sherlock 1987 Maguire, Hoff & Sherlock 1987 Mota et al. 1990 Mota et al. 1990 RBBB = right bundle branch block. VEs = ventricular extra systoles. In the seronegative group with normal ECG the mortality was 3.9 per 1000 person-years (Maguire, Hoff & Sherlock 1987). The data reflected in Table 11.5 clearly indicate the much higher relative risk of developing myocardial lesions and dying as a consequence of these lesions in young individuals who are seropositive with regard to T. cruzi as compared with seronegative individuals of the same age group. It is assumed that the mortality due to cardiac lesions follows a similar pattern in all endemic areas throughout the continent. Burden and intervention Economic impact The economic impact of the disease during the chronic stage is very high, as shown by data from Brazil (Brazil Ministry of Health 1987). If we consider that about 30 per cent of infected people will develop severe cardiac and digestive lesions such as cardiac arrhythmia (75 000 cases), mega oesophagus (45 000 cases), and mega colon (30 000 cases) per year, the estimated costs for pacemaker implants and corrective surgery (average US$5000) would amount to approximately US$750 million per year, which would be enough for the improvement or construction of more than 700 000 rural dwellings at a minimum estimated cost of US$1000 each in Brazil. Between 1979 and 1981, 14 022 deaths were resulted from Chagas disease in Brazil, representing approximately 259 152 years of potential life lost (YPLL) before the age of retirement. Assuming that all the patients were unqualified rural workers only and that the minimum daily wage at the time was US$2.50, the total economic loss attributable to premature deaths would amount to US$237 million. Current control programmes The traditional vertical control programmes in the Latin American countries have focused on the spraying of insecticides on houses, household annexes and buildings. National control programmes aimed at the inter- Chagas Disease 333 ruption of the domestic and peridomestic cycles of transmission involving vectors, animal reservoirs and humans are feasible and have proven to be very effective. Reaching the goal of eliminating vector borne transmission is more feasible in areas where the vector is domiciliated like Triatoma infestans. Eight countries of the Americas have active control programmes that combine insecticide spraying with health education. The common pattern of the vertical, centralized control programmes comprises three operational phases: preparatory phase for mapping and general programming of activities and estimation of resources; attack phase during which a first massive insecticide spraying of houses takes place and is followed by a second spraying 3 to 6 months later, with further evaluations for selective re-spraying of re-infested houses; surveillance phase for the detection of residual foci of triatomines after the objective of the attack phase has been reached. Involvement of the community and the decentralization of residual control activities are essential elements of this last phase. A prime example of an active control programme is the programme that has been operating in Brazil since 1975 when 711 Brazilian municipalities had triatomine-infested dwellings. Ten years later, in 1986, only 186 municipalities remained infested. This represents a successful accomplishment of the programme objectives in 74 per cent of the originally infested municipalities (Dias 1987). In large parts of the Southern Cone countries, programmes have entered the surveillance phase characterized by monitoring of house infestation and, where necessary focal spraying. New tools for vector control Two new tools for triatomine control have been developed with the support of the WHO: fumigant canisters and insecticidal paints (UNDP/World Bank/WHO 1998b). After four years of implementation of a control programme in 600 scattered houses in Santiago del Estero, Argentina (covering some 3000 individuals), using fumigant canisters and a triatomine detection box for entomological surveillance, seropositivity in infants under 1 year of age dropped from 5.5 per cent to 0 per cent, while seropositivity in children under 4 years of age dropped from 12.4 to 0 per cent. In other words, the transmission of the disease through vectors was interrupted. The programme was implemented within a primary health care (PHC) context, and had a strong component of health education. The cost of the whole intervention, including canisters and triatomine detection boxes, PHC personnel, serological evaluation, and so on was US$4.70 per house per year, i.e. five times lower than the cost of the traditional vertical approach 334 Global Epidemiology of Infectious Diseases and insecticide application used in the government control programme (Chuit et al. 1992) In another field project in 4800 houses in central Brazil ( covering some 30 000 people), the insecticidal paints still had a high level of efficacy 24 months after application, keeping more than 85 per cent of the treated dwellings free of triatomines. This compares very favourably with the 60 per cent efficacy of traditional insecticide spraying. These insecticidal paints are manufactured in Sao Paulo, Brazil. The cost per treated house over the 24-month period was US$29 for the paints, compared to US$66.34 for Deltamethrin and US$73.40 for Benzene hexachloride. It was necessary to apply the insecticides twice during the 24-month period, while only one application of the paint was enough to attain the above-mentioned levels of efficacy (Oliveira Filho 1988, 1989). In an extended operation being carried out in selected areas of Argentina, Bolivia, Chile, Honduras, Paraguay, and Uruguay, these new control tools are being applied following a common protocol that will permit comparability of epidemiological and entomological efficacy, social acceptability and assessment of costs. A health education component will also be included in the common protocol. Results from Argentina, Chile, Honduras, and Paraguay are very encouraging, based on the entomological evaluations at 12 to 24 months post-intervention. The rates of house re-infestation when using the insecticide paints are two to three times lower than for houses sprayed with traditional insecticides (see Table 11.6). These innovations in triatomine control are being incorporated into the regular activities of the control programme by the Ministry of Health of Brazil in the State of Ceara, resulting in decreased operational costs and increased efficacy of vector control. Feasibility of interruption of transmission The tools for interrupting the domestic cycle of T. cruzi transmission, such as chemical control, housing improvement and health education, are available. In fact, the prevalence of the infection has decreased in countries that have consistently applied control measures. For example, after 20 years of control programmes in Argentina, positive serology in 18-year-old males has significantly decreased since 1980, and the number of reported new acute cases has decreased since the 1970s (Segura et al. 1985). In Brazil, transmission by vector has been interrupted throughout the State of Sao Paulo since the mid-1970s. Decreasing seropositivity in school children has paralleled the above control efficacy: in 1976, the incidence rate was 60 per cent, and in 1983 it dropped to 0 per cent (Souza et al. 1984). Transmission through transfusion could be prevented if blood was screened serologically and positive units were discarded. In some countries of the region, serological screening for T. cruzi is mandatory for blood donors. Chagas Disease Table 11.6 335 Efficacy of triatomine control tools: percentage of re-infested houses after different interventions, Argentina, Chile, Honduras and Paraguay, April 1994 Experimental group Percentage of houses re-infested Number of houses Chilea Hondurasb Argentinac Paraguayc 150 5.6d 1.6d 2.5d 5.6 150 nd 3.0 0.66d 8.0 150 2.1d 0.8d 2.3d 2.6d 300 8.9 2.7 3.6 4.3 I Peri-domicile: traditional insecticide Intra-domicile: paints II Peri-domicile: traditional insecticide Intra-domicile: canisters III Peri-domicile: paints Intra-domicile: paints IV (control) Peri-domicile: traditional insecticide Intra-domicile: traditional insecticide Source: Progress reports of field research projects, Geneva, World Health Organization, June 1994 (unpublished data). a. 24 months after interventions. b. 18 months after interventions. c. 12 months after interventions. d. Statistically significant difference compared with control group. Available knowledge therefore indicates that the most common ways of transmitting human T. cruzi infection could be interrupted by: implementing of vector control activities in houses in order first to reduce and then to eliminate the vector borne transmission of T. cruzi; strengthening the ability of blood banks to prevent transmission of Chagas disease and other diseases transmitted by blood transfusion, through the development and implementation of a policy for the use of human blood. Update Initiative of the Southern Cone Countries In Brasilia in June 1991, the Ministers of Health of Argentina, Brazil, Bolivia, Chile, Paraguay and Uruguay launched an Initiative for the elimination of Chagas disease during the 1990s (MERCOSUR 1991). Since Triatoma infestans, the vector of T.cruzi, is intra-domiciliary in these countries, sustained implementation of control measures has successfully 336 Global Epidemiology of Infectious Diseases interrupted transmission of Chagas, disease as indicated in Table 11.7 and Figure 11.1. In these countries there are 11 million infected people and 50 million are at risk. This represents 60 per cent of the prevalence of infected individuals for the whole continent. Chagas disease is recognized as an important public health problem and is being given increased priority for control, as demonstrated by the initiative taken by the governments of the Southern Cone countries. That initiative has been very successful. By cutting transmission in the Southern Cone countries, the incidence of Chagas disease will be reduced by over 65 per cent in the whole of Latin America. A total of US$450 million has Table 11.7 Human infection by Trypanosoma cruzi and reduction of incidence, Southern Cone initiative for the elimination of Chagas disease: 1983–2000 Country Age group (years) Prevalence of infection (percentage) in 1980–1985 Prevalence of infection (percentage) in 2001a Reduction of incidence in the age group (percentage) Argentina Brazil Chile Paraguay Uruguay 18 0–4 0–10 18 6–12 5.8 (Segura et al.1985) 5.0 (Camargo et al.1984) 5.4 (Schenone et al. 1985) 9.3c (Cerisola 1972) 2.5 (Salvatella et al. 1989) 1.2b 0.12 0.38 3.9 0.06 80.0 98.0 94.0 60.0 99.0 a. Reports to the 10th Meeting of the Intergovernment Commission of the Southern Cone Initiative, Montevideo, 2001. b. Data for 1993. c. Data for 1972. Figure 11.1 Southern Cone Initiative for the elimination of transmission of Chagas disease: incidence of infection 1980–2000 Incidence (%) 10 Paraguay (18 years) Argentina (18 years) Chile (<4 years) 1 Brazil (0–7 years) 0.1 0 1980 Uruguay (<12 years) 1985 1990 1995 Year Source: Reports by national Chagas disease control programmes, 1993–2000. 2000 Chagas Disease 337 been allotted by the six countries for control operations since the start of the initiative. Technical representatives of each ministry of the Southern Cone countries have been designated to form an Intergovernmental Commission in charge of implementation and evaluation of the control programmes. The Pan American Health Organization (PAHO) has been appointed to act as the Secretariat of this Commission. A programme guide, designed by the Commission and incorporating revisions submitted by the professional staff of the control programmes, has been used for the development of the country programmes. The proposed plans for Argentina, Bolivia, Brazil, Chile, Paraguay, and Uruguay are approved on a yearly basis by their respective governments. There have been 10 meetings of the Intergovernmental Commission, held in Buenos Aires, Argentina (1992), Santa Cruz, Bolivia (1993), Montevideo, Uruguay (1994), Asuncion, Paraguay (1995), Porto Alegre, Brazil (1996), Santiago, Chile (1997), Buenos Aires, Argentina (1998), Tarija, Bolivia (1999), Rio de Janeiro, Brazil (2000), and Montevideo, Uruguay (2001). At these meetings, the field activities of the country programmes were reviewed annually, using common indicators to assess the impact and cost-effectiveness of the national programmes. Current data (World Health Organization 1994, 1995, 1996, 1997, 1998a, 1999a, 2000a, 2000b) on ridding of houses of insects, screening blood banks, and the incidence of infection in the under-5 years age group indicate that the vector-borne and tranfusional transmission of Chagas disease were interrupted in Uruguay in 1997, Chile in 1999, and Brazil in 2000. Argentina, Bolivia, and Paraguay were expected to follow soon after. Argentina: The whole endemic area of the country has been placed under entomological surveillance, the most advanced phase of the elimination process. In 1980 the average house infestation rate for the country as a whole was 30 per cent; in 1998 it was 1.2 per cent. That change is equivalent to a 96 per cent reduction in house infestation by the vector. The seroprevalence rates for the whole country for the age group 0–4 years is 0.9 per cent which confirms the very low number of acute cases among children in this age group. In the age group 0–14 years, the rate is 1.9 per cent. These data show that the goal of interruption of vectorial transmission is being achieved. Regarding blood safety, the coverage of the blood donations screened against Chagas disease is 100 per cent in the blood banks of the public sector and 80 per cent in the blood banks of the private sector (World Health Organization 1996). Bolivia: Data on serological prevalence in the general population indicates a rate of 28.8 per cent while the serological prevalence rate in the age group of 0–4 years is 22.0 per cent in Cochabamba, it is 0.0 per cent in Potosi where there is an active vector control programme. In Tupiza, another department where there is an active control programme, the 338 Global Epidemiology of Infectious Diseases house infestation rate is 0.8 per cent (Pan American Health Organization 1998b). Brazil: The prevalence of human T. cruzi infection in the 7–14 year age group in 1999 was 0.04 per cent as compared with 18.5 per cent in 1980. This represents a 99.8 per cent reduction of incidence of infection in that age group. Results of serological tests in a limited number of samples in the population of the 0–4 year age group in 1999 indicate that the seroprevalence in that age group is 0.0 per cent which can be interpreted as a proof of the interruption of vectorial transmission of Chagas disease in Brazil. The number of domiciliated T. infestans insects captured by the control programme in the whole country in 1998 was only 485. This represents an average of 1 insect per 10 000 houses surveyed, i.e. an infestation rate far below the minimum required for effective transmission of the parasite to new patients. This information confirms the interruption of vectorial transmission of Chagas disease in Brazil. Chile: The overall infestation rate for the country was reduced from 3.2 per cent in 1994 to 0.14 per cent in 1999, a reduction of 99.8 per cent over three years. In 1999, just 26 T. infestans insects were captured in the interior of dwellings in the endemic areas of the whole country, representing 2.5 insects per 1000 houses, an infestation rate far below the threshold required for effective transmission of the parasite to new individuals. The infection rate in the age group 0–4 years in 1999 was 0.016 per cent which represents a reduction of 98.5 per cent as compared to the rate of 1.12 per cent that was found in the same age group in 1995. An independent commission visited the endemic areas of Chile in November 1999 and based on the above data certified the interruption of vectorial transmission (World Health Organization 1999a, 2000a). Paraguay: In a serological survey carried out in 1997 in a representative sample (940 individuals) of children less than 13 years old in marginal areas of the capital city of Asunción a significant decrease of prevalence rates was observed in all age groups when compared with data for 1982. Rural-urban migration to these marginal areas of Asunción comes mainly from Paraguari, Cordillera, and Central, which have the highest domiciliary infestation rates by triatomines. The zero prevalence rate in the age group of less than 4 years old indicates interruption of transmission by triatomines in the urban areas of the capital (Pan American Health Organization 1998a, 1998b). Uruguay: The interruption of vectorial and transfusional transmission was certified in 1997 and the whole country is under surveillance. The incidence of infections in the age group 0–12 years was 0 per cent which confirms the interruption of vectorial and transfusional transmission of Chagas disease in Uruguay since 1997 (World Health Organization 1998b). Chagas Disease 339 Initiative of the Andean Countries There are 5 million infected individuals in the Andean countries, and 25 million people those countries are at risk of contracting the infection. This represents 27 per cent of the prevalence of infected individuals for the whole continent. As the vectors of Chagas disease in these countries are not strictly domiciliated, it is necessary to adapt and test the vector control strategies with respect to the local entomological conditions. An initiative for the elimination of vectorial transmission in the Andean countries of Colombia, Ecuador, Peru, and Venezuela, was launched at an Intergovernmental meeting held in Bogotá, Colombia, in February 1997. At that meeting detailed country-by-country plans of action, including annual goals, budgetary needs, evaluation mechanisms and research needs, were prepared (OPS/OMS 1997). A second Intergovernmental meeting Commission was held in Maracay, Venezuela, in March 1999. Progress in control activities was reported as indicated below. Colombia: The preparatory phase of the national Chagas disease control programme has advanced. The prevalence survey has been completed and a map of the country featuring the risk areas has been prepared. There is a 100 per cent coverage of blood screening against T. cruzi infection in blood. Ecuador: The preparatory phase of the national Chagas disease control programme has advanced. A law reorganizing the control programme was issued on 8 December 1998, placing the programme under the Secretary of Tropical Medicine, with specific functions and budget. A law for compulsory blood screening against T. cruzi was issued in August 1998. The seroprevalence of infected blood in blood banks for the whole country is 0.2 per cent. Venezuela: The control programme was officially established in 1966. The objective of the programme was to interrupt intradomestic transmission through vector control by insecticide spraying. The programme for improvement of rural housing, originally initiated in the 1960s, assists rural inhabitants to substitute palm roofs, to plaster adobe walls and to cement earthen floors. In addition, routine screening for T. cruzi in blood banks began in 1988. In children under 10 years of age, the figures for seroprevalence rates of T. cruzi infection have decreased steadily in the past four decades: from 20.5 per cent in, 1958–1968 to 3.9 per cent in 1969–1979, and further down to 1.1 per cent in 1980-1989, to 0.8 per cent in 1990–1999. Between 1990 and 1999, incidence of infection in the age group 0–4 years old was reduced by 90 per cent to less than 1.0 per cent. The geographical distribution of T. cruzi transmission is restricted to the States of Portuguesa, Barinas and Lara. The prevalence of infected blood in blood banks was been reduced from 1.16 per cent in 1993 to 0.78 per cent in 1998 (World Health Organization 1999b). 340 Global Epidemiology of Infectious Diseases Initiative of the Central American Countries In Central American countries, there are 2 million infected individuals and 26 million are at risk of contracting the infection. This represents 11 per cent of the prevalence of infected individuals for the whole continent. As the vectors of Chagas disease in these countries are not strictly domiciliated, it is necessary to adapt and test the vector control strategies with respect to the local entomological conditions. In the Central American countries—Belize, Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, and Panama—progress in control of blood banks is proceeding well. All of these countries except one have issued legislation for compulsory blood screening against blood infected by T. cruzi. An initiative for elimination of vectorial transmission was launched at an intergovernmental meeting held in Tegucigalpa, Honduras in October 1997. At the meeting, detailed country-by-country plans of action, including annual goals, budgetary needs, evaluation mechanisms and research needs, were prepared (OPS/OMS 1997). Further follow-up intergovernmental meetings have been held in Guatemala City, Guatemala, in 1998, in Managua, Nicaragua, in 1999, and in San Salvador, El Salvador, in 2000. A summary of progress regarding the updating of epidemiological and entomological data, vector control activities and control of blood banks is given in Table 11.8. Conclusions Cost-effectiveness of Control Interventions The Ministry of Health of Brazil carried out a study to analyse the costeffectiveness and cost-benefit of the Chagas disease control programme in Brazil. Because of the chronic nature of the disease and the protracted period of evolution, a period of 21 years was chosen for the analysis: from 1975 to 1995. Data from different sources were used to carry out the evaluation (Akhavan 1997). Effectiveness was defined using various parameters; the main one was the burden of disease prevented, measured in DALYs (disability-adjusted life-years). From 1975 to 1995, the programme (excluding blood banks) prevented an estimated 89 per cent of potential disease transmission, preventing 2 339 000 new infections and 337 000 deaths. This translated into the prevented loss of 11 486 000 DALYs, 31 per cent from averted deaths and 69 per cent from averted disability. These results illustrate the large role of disability in the overall burden of disease caused by Chagas disease. The estimated benefits (expenditures prevented) of the programme (excluding blood banks) were US$7500 million, 63 percent of the savings being health care expenditures and 37 per cent social security expenditures (disability insurance and retirement benefits). The cost-effectiveness analysis demonstrated that for each US$39 spent on the programme, 1 DALY was gained. This places the programme and its Epidemiological information Not started Ongoing survey in age group 0–14 years Starting in 1999 Ongoing survey in age group 0–14 years Ongoing survey in age group 0–14 years Ongoing survey in age group 0–14 years Starting in 1999 Belize Costa Rica El Salvador Guatemala Honduras Nicaragua Panama Ongoing Ongoing Ongoing survey on mobility of T. dimidiata Ongoing survey on urban infestation. Ongoing survey on mobility of T. dimidiata Ongoing survey to map vectors at municipality level. Ongoing Ongoing Not started Entomological information Spraying activities started in 1998 in the central Departments, which are endemic for R. prolixus Starting in 1999 Rhodnius prolixus, Triatoma dimidiata Rhodnius prolixus, Triatoma dimidiata Surveillance activities as the vectors are sylvatic Spraying activities will start in 2000 in the Departments of Quiché, Chiquimula, Jalapa and Zacapa, which are endemic for R. prolixus Rhodnius prolixus, Triatoma dimidiata Rhodnius pallescens, Triatoma dimidiate Organization of control programme at the Ministry of Health Organization of control programme at the Ministry of Health Not started Vector control activities Triatoma dimidiata Triatoma dimidiata Unknown Main vector Summary of control activities in Central America, 1999 Country Table 11.8 Unknown 65% 100% 80% 100% Unknown Unknown Screening of blood banks (coverage) Chagas Disease 341 342 Global Epidemiology of Infectious Diseases activities in the category of interventions with a very high cost-effectiveness. The results of the cost-benefit analysis indicated savings of US$17 for each US dollar spent on prevention, also indicating that the programme is a health investment with a good rate of return. An analysis of other diseases with socio economic causes demonstrated that the decline in Chagas disease infection rates is attributable to the preventive activities, rather than to a general improvement in living conditions. Research Current and future research priorities will focus on operational field research for better planning of control activities, cost-effectiveness of interventions and assessment of the impact of control interventions. Regarding blood banks, research is centred on standardization and improvement of screening tests, evaluation of sensitivity and specificity of new screening tests, and cost analysis of multiple blood bank screening. A task force will focus its activities on the study of population dynamics of non-domiciliated triatomine vectors of Chagas disease present in the northern part of South America and in Central America. It is expected that the entomological information generated will assist national control programmes to adapt the vector control strategies that have proven very successful in interrupting the vectorial transmission of Chagas disease in the countries of the Southern Cone. Studies will be carried out on: triatomine distribution and house/peridomicile infestation by non- domiciliated species; genetic structure of populations of non-domiciliated triatomines; sylvatic/domestic mobility of vector populations; cost-effectiveness of different methods to ascertain house re-infestation; sensitivity of methods to detect re-infestation by non-domiciliated species; monitoring of insecticide efficacy; effect of bio-climatic changes on domiciliated and non-domiciliated vector populations. References Acquatella H (1987) Encuesta epidemiológica en sujetos con serología positiva para Enfermedad de Chagas [Epidemiological survey in individuals with positive serology for Chagas disease]. Ciencia y Tecnologia de Venezuela, 4:185–200. Akhavan D (1997) Analysis of cost-effectiveness of the Chagas disease control programme, Brasilia, Ministry of Health, National Health Foundation. Chagas Disease Table 11.9 1980–1985 343 The time table towards the interruption of transmission of Chagas disease Prevalence cross-sectional studies on human infection and house infestation in nine countries are carried out (Acquatella 1987, Camargo et al. 1984, Cedillos 1975, Cordova et al. 1980, Lopez 1985, Marinkelle 1976, Matta 1985, Pan American Health Organization 1974, Ponce 1984, Reyes Lituma 1984, Salvatella et al. 1989, Schenone et al. 1985, Sousa 1985,Valencia 1990, Zeledon 1976) Standardization of serological techniques and creation of a continental network of reference laboratories (Camargo, Sequra & Kagan 1986) 1985–1990 Follow-up prospective studies on the course of human infection (Arribada, Apt & Ugarte 1986, Coura, Anunziako & Willcox 1983, Maguire, Hoff & Sherlock 1987, Mota et al. 1990, Pereira, Willcox & Cava 1985) Cloning of parasite genome and production of defined antigens for improvement of diagnostic techniques (Peterson, Wrightsman & Manning 1986, Moncayo & Luquetti 1990) Industrial production of kits for control of blood banks in Argentina Development of new tools for vector control (Oliveira Filho 1988) 1990–1995 Multi-country field studies for evaluation of new vector control tools (Oliveira Filho 1997) Industrial production of paints, canisters and sensor boxes in Argentina and Brazil (Zerba 1988) Initiative of the Southern Cone countries is launched (Schmunis, Zicker & Moncayo 1996, Moncayo 1997) 1997 Initiatives of the Andean countries and the Central American countries are launched (OPS/OMS 1997) Uruguay declared free of vectorial and transfusional transmission of Chagas disease (World Health Organization 1998a, UNDP/World Bank/WHO 1998a) 1998 The 51st World Health Assembly endorses the goal for elimination of transmission of Chagas disease on the American continent in 2005 (World Health Organization 1998b) 1999 Treatment of the early chronic phase in children aged 7–11 years old, adopted as public health policy (PAHO/WHO 1999) Chile declared free of vectorial and transfusional transmission of Chagas disease (World Health Organization 200a) 2000 Brazil declared free of vectorial transmission of Chagas disease Arribada AC, Apt W Ugarte JM (1986) A four-year follow up survey of Chagasic cardiopathy in Chile. Bulletin of the Pan American Health Organization, 20: 245–266. Arturo AC, Werner AB, Ugarte JM (1986) A four-year follow up survey of Chagasic cardiopathy in Chile. Bulletin of the Pan American Health Association, 20:245–266. Brazil Ministry of Health (1987) Alguns aspectos economicos e de custo-beneficio em Doenca de Chagas no Brasil [Economic and cost-benefit aspects of Chagas disease in Brazil], Informacoes Epidemiologicas SUCAM, V.54. 344 Global Epidemiology of Infectious Diseases Camargo M et al. (1984) Inquerito sorológico da prevalencia de infeccao Chagásica no Brasil 1975/80 [Serological survey on prevalence of Chagasic infection in Brazil 1975/1980]. Revista do Instituto de Medicina Tropical de Sao Paulo, 26:192–204. Camargo M, Segura EL, Kagan IG (1986) Three years of collaboration on the standardization of Chagas disease serodiagnosis: an appraisal. PAHO Bulletin, 20:233–244. Carrasco R (1990) Prevalence of Trypanosoma cruzi infection in blood banks of seven departments of Bolivia. Memorias do Instituto Oswaldo Cruz, 85: 69–73. Cedillos RA (1975) Chagas disease in El Salvador. Bulletin of the Pan American Health Organization, 9:135–141. Cerisola A (1972) Muestreo serológico para la Enfermedad de Chagas en conscriptos de la clase 1951 [Serological sampling for Chagas disease in conscripts of the year 1951]. Revista Paraguaya de Microbiología, 7:90. Chagas C (1909) Nova tripanozomiase humana. Estudos sobre a morfolojia e o ciclo evolutivo do Schizotrypanum cruzi n.gen.,n,sp., ajente etiolojico de nova entidade morbida do homen [New human tripanosomiasis. Study of the morphology and evolutionary cycle of Schizotrypanum cruzi n.gen.,n,sp., etiological agent of the new human morbidity]. Memorias do Instituto Oswaldo Cruz, 1:159–218. Chuit R, et al. (1992) Result of a first step toward community-based surveillance of transmission of Chagas disease with appropriate technology in rural areas. American Journal of Tropical Medicine and Hygiene, 46:444–450. Cordova E et al. (1980) Enfermedad de Chagas en el Sur del Peru [Chagas disease in southern Peru]. Boletin Peruano de Parasitologia, 2:46–50. Coura JR et al. (1985) Morbidade da doenca de Chagas’: IV Estudo Longitudinal de dez anos em Pains e Iguatama, Minas Gerais, Brasil [Morbidity of Chagas’ disease: IV- Ten-year Longitudinal Study in Pains and Iguatama, Minas Gerais, Brasil]. Memorias do Instituto Oswaldo Cruz, 80:73–80. Coura JR, Anunziato N, Willcox HPF (1983) Morbidade da doenca de Chagas’: I- Estado de casos procendentes de varios estados de Brasil, observados no Rio de Janeiro [Morbidity of Chagas’ disease. I. Study of cases originating in various states of Brazil, observed in Rio de Janeiro]. Memorias do Instituto Oswaldo Cruz, 78:363–372. Dias JCP (1987) Control of Chagas disease in Brazil. Parasitology Today, 3:11, 336–341. Dias JCP, Brener Z (1984) Chagas disease and blood transfusion. Memorias do Instituto Oswaldo Cruz, 79(Suppl.):139–147. Ecuador Ministry of Health (1961). Report. Franca ME (1986) Chagas disease in Uruguay in the last twenty years. Revista Medica del Uruguay, 2:125–131. Grant IH, Gold JW, Wittner M (1989) Transfusion-associated acute Chagas disease acquired in the United States. Annals of Internal Medicine, 111:849–851. Chagas Disease 345 Guhl F et al. (1987) Rastreo seroepidemiológico de donantes de sangre chagásicos en una zona endemica (Norte de Santander, Colombia) [Seroepidemiologic tracing of chagasic blood donors in an endemic zone (Norte de Santander, Colombia)]. Revista Latinoamericana de Microbiologia, 29:63–66. Kirchoff L, Gam A, Gillian F (1987) American trypanosomiasis (Chagas disease) in Central American immigrants. American Journal of Medicine, 82:915. Liendo F et al. (1985) Enfermedad de Chagas en Chile, sectores urbanos: XIIIPrevalencia de infección por Trypanosoma cruzi en donantes de sangre del sector sur del area metropolitana de Santiago, 1985 [Chagas’ disease in Chile, urban sectors: XIII- Prevalence of infection by Trypanosoma cruzi in blood donors of the Southern sector of the metropolitan area of Santiago, 1985]. Boletin Chileno de Parasitologia, 40:82–84. Lopez L (1985) Encuesta serológica de prevalencia de la Enfermedad de Chagas en Paraguay [Serological survey on the prevalence of Chagas disease in Paraguay]. SENEPA project Report. Asuncion, Paraguay, Ministry of Health. Maguire JH, Hoff R, Sherlock I (1987) Cardiac morbidity and mortality due to Chagas disease: prospective electrocardiographic study of a Brazilian community. Circulation, 75:1140–1145. Marinkelle C (1976) Aspects of Chagas disease in Colombia. In: American Trypanosomiases Research, PAHO Scientific Publication, No. 318, Washington DC, 340–346. Marzochi MCA (1981) Chagas disease as an urban problem. Cadernos de Saude Publica, 2:7–12. Matta V (1985) Seroepidemiología de la Enfermedad de Chagas en Guatemala [Seroepidemiology of Chagas disease in Guatemala] Memorias VII Congreso Centroamericano de Microbiologia, San Jose, Costa Rica, p. 17. Menezes M (1989) Causas básicas de morte en Chagasicos Idosos [Basic causes of death in aged Chagas patients]. Arquivos Brasileiros de Cardiologia, 52:75–78. MERCOSUR (1991) Reunión de Ministros de Salud del MERCOSUR, Brasilia, Junio 1991[Meeting of the Ministers of Health of MERCOSUR, Brasilia, June 1991], Resolución No. 04-3-CS. Mota EA et al. (1990) A nine year prospective study of Chagas disease in a defined rural population in northeast Brazil. American Journal of Tropical Medicine and Hygiene, 42:29–440. Moncayo A, Luquetti A (1990) Multicentre double blind study for evaluation of Trypanosoma cruzi defined antigens as diagnostic reagents. Memorias do Instituto Oswaldo Cruz, 85:489–95. Moncayo A (1997) Progress towards elimination of transmission of Chagas disease in Latin America. WHO World Health Statistics Quarterly, 50:195–198. Nickerson P et al. (1989) Transfusion-associated Trypanosoma cruzi infection in a non-endemic area. Annals of Internal Medicine, 111:851–853. Oliveira Filho AM (1988) Development of insecticide formulations and determination of dosages and application schedules to fit specific situations. Revista Argentina de Microbiologia, 20(Suppl):39–48. 346 Global Epidemiology of Infectious Diseases Oliveira Filho AM (1989) Cost-effectiveness analysis in Chagas disease vectors control interventions. Memorias do Instituto Oswaldo Cruz, 84(Suppl.4): 109–418. Oliveira Filho AM (1997) Uso de nuevas herramientas para el control de triatominos en diferentes situaciones entomológicas en el Continente Americano. Revista da Sociedade Brasileira de Medicina Tropical, 30 :41–46. Organización Panamericana de la Salud/Organización Mundial de la Salud (1997) Informes de las Iniciativas de los Países Andinos, Bogotá Febrero 1997 y Tegucigalpa, Octubre 1997 [Report of the initiatives of the Andean countries, Bogotá February 1997 and Tegucigalpa October 1997]. Pan American Health Organization/World Health Organization (1999) Treatment of the early chronic phase of Chagas disease in children and adults. Report of an Expert Group. Washington, DC, October 1999. Pan American Health Organization (1974) Clinical aspects of Chagas disease. Bulletin of the Pan American Health Organization, 76:141–155. Pan American Health Organization (1998a) Report of the VI Meeting of the Intergovernment Commission of the Southern Cone Initiative, Santiago, Chile, March 1997. Washington, DC, (OPS/HPC/HCT/98.102). Pan American Health Organization (1998b) Report of the VII Meeting of the Intergovernment Commission of the Southern Cone Initiative, Buenos Aires, Argentina, 24–26 March 1998. Washington DC, (OPS/HPC/HCT/98.114). Paraguay Ministry of Health (1972) Report. Pereira JB, Willcox HP, Coura JR (1985) Morbidade da doenca de Chagas: IIIestudo longitudinal de seis anos, em Virgen da Lapa, MG, Brasil [Morbidity of Chagas disease: III- six-year longitudinal study in Vigem da Lapa, MG, Brazil]. Memorias do Instituto Oswaldo Cruz, 80:163–71. Pereira MG (1984) Characteristics of urban mortality from Chagas’ disease in Brazil’s Federal District. Bulletin of the Pan American Health Organization, 96:213–221. Perez A, Segura EL (1989) Blood transfusion and transmission of Chagas infection in Argentina. Revista Argentina de Transfusion, 15:127–132. Peru Ministry of Health (1972). Report. Peterson DS, Wrightsman RA, Manning JE (1986) Cloning of a major surfaceantigen gene of Trypanosoma cruzi and identification of a nonapeptide repeat. Nature, 322:566–568. Ponce C (1984) Prevalencia de la enfermedad de Chagas en Honduras [Prevalence of Chagas disease in Honduras]. Tegucigalpa, Honduras, Ministry of Health. Ponce C, Ponce E (1987) Infección por Trypanosma cruzi en donantes de sangre de diferentes hospitales de Honduras [Infection with Tripanosoma cruzi in blood donors in various hospitals in Honduras]. Presented at the 8th Latin American Congress of Parasitology, Guatemala, November 1987. Reyes Lituma V (1984) Estudio de prevalencia de enfermedad de Chagas en Ecuador [Prevalence study of Chagas disease in Ecuador]. Guayaquil, Ecuador, Ministry of Health, Vector Control Programme. Chagas Disease 347 Salvatella R et al. (1989) Seroprevalence of antibodies against Trypanososma cruzi in 13 Departments of Uruguay. Bulletin of the Pan American Health Organization, 107:108–117. Schenone H et al. (1985) Enfermedad de Chagas en Chile: sectores rurales y periurbanos del area de endemo-enzootia relaciones entre condiciones de la vivienda, infestación triatomidea domiciliaria e infección por Trypanosoma cruzi del vector, del humano y de mamíferos domésticos, 1982–1985 [Chagas disease in Chile: rural and periurban sectors of the endemo-enzootic area-relationship between housing conditions, domicilliary triatomid infestation and infection by Trypanosoma cruzi of the vector, humans and domestic mammals, 1982–1985]. Boletin Chileno de Parasitologia, 40:58–67. Schmunis GA (1991) Trypanosoma cruzi, the etiologic agent of Chagas’ disease: status in the blood supply in endemic and non-endemic countries. Transfusion, 31:547–557. Schmunis GA, Zicker F, Moncayo A (1996) Interruption of Chagas’ disease transmission through vector elimination. Lancet, 9035:171. Segura EL, et al. (1985) Decrease in the prevalence of infection by Trypanosoma cruzi (Chagas disease) in young men of Argentina. Bulletin of the Pan American Health Organization, 19:252–264. Sousa O (1985) Encuesta serológica de prevalencia de la enfermedad de Chagas en Pánama [Serological survey on prevalence of Chagas disease in Panama]. Project reportLaboratory of Biology, University of Panama, Informe de Proyecto (Project report), Ciudad de Panama. Souza AG et al. (1984) Consolidation of the control of Chagas disease vectors in the State of Sao Paulo. Memorias do Instituto Oswaldo Cruz, 79(Suppl.): 125–131. UNDP/World Bank/WHO (1987) Special Programme for Research and Training in Tropical Diseases. Eighth programme report. Geneva, p. 92. UNDP/World Bank/WHO (1991) Special Programme for Research and Training in Tropical Diseases. Eleventh programme report. Geneva, p. 67. UNDP/World Bank/WHO (1998a) Special Programme for Research and Training in Tropical Diseases. Geneva, (TDR Newsletter, 56:6). UNDP/World Bank/WHO (1998b) Special Programme for Research and Training in Tropical Diseases. Third external review, final report 27 and Reference Document No. 5. Urbina A et al. (1991) Serologic prevalence of infection with T. cruzi in blood donors in Costas Rica. Abstracts VII Congreso Centroamericano de Microbiología y Parasitología, No. 18, p. 204. Valencia A (1990) Investigación epidemiológica nacional de la enfermedad de Chagas [National epidemiological survey of Chagas disease]. La Paz, Bolivia, Ministerio de Prevision Social y Salud Publica (Ministry of Social Forecasting and Public Health) (PL-480 Title III). Velasco Castrejon O, Guzman Bracho C (1986) Importance of Chagas disease in Mexico. Revista Latino Americana de Microbiologia, 28:275–283. 348 Global Epidemiology of Infectious Diseases World Health Organization (1994) Chagas disease: elimination of transmission in Uruguay. Weekly Epidemiological Record, 6:38–40. World Health Organization (1995) Chagas disease: interruption of transmission in Chile. Weekly Epidemiological Record, 3:13–16. World Health Organization (1996) Chagas disease: progress towards elimination of transmission in Argentina. Weekly Epidemiological Record, 2:12–15. World Health Organization (1997) Chagas disease: interruption of transmission in Brazil. Weekly Epidemiological Record, 1/2:1–5. World Health Organization (1998a) Chagas disease: interruption of transmission in Uruguay. Weekly Epidemiological Record, 1/2:1–4. World Health Organization (1998b) Fifty-first World Health Assembly, resolution WHA51.14. World Health Organization (1999a) Chagas disease: interruption of transmission in Chile. Weekly Epidemiological Record, 2:9–11. World Health Organization (1999b) Chagas disease: progress towards interruption of transmission in Venezuela. Weekly Epidemiological Record, 35:289–292. World Health Organization (2000a) Chagas disease: certification of interruption of transmission in Chile. Weekly Epidemiological Record, 2:10–12. World Health Organization (2000b) Chagas disease: interruption of transmission in Brazil. Weekly Epidemiological Record, 19:153–155. Zeledon R (1976) Epidemiological pattern of Chagas disease in an endemic area of Costa Rica. American Journal of Tropical Medicine and Hygiene, 24: 214–225. Zerba EN (1988) Development of new insecticides and synergistic formulations of the Chagas disease vector control. Revista Argentina de Microbiología, 20 (Suppl):25–31. Chapter 12 Schistosomiasis KE Mott Introduction Schistosomiasis, sometimes called bilharziasis (after the German physician Theodor Bilharz), ranks high among parasitic diseases in terms of socioeconomic and public health importance in tropical and subtropical areas. One measure of that importance within the development strategies of endemic countries is reflected in executive level planning and inclusion of schistosomiasis control activities in national budgets in Brazil, China, Egypt, the Philippines, and Morocco (World Health Organization 1993). At least one form of schistosomiasis is now endemic in 74 tropical developing countries. It is estimated that over 200 million people residing in rural agricultural and periurban areas are infected, while 500–600 million more run the risk of becoming infected as a result of poverty, substandard hygiene in poor housing, and inadequate public infrastructure. Schistosomiasis is a major health risk in the rural areas of central China and Egypt and ranks high in other developing countries. Schistosomiasis is mainly a rural occupational disease that affects people engaged in agriculture or fishing. In many areas, a large proportion of children are infected by the age of 14 years; in other areas, women face the highest risk because of domestic and occupational contact with fresh water. Increased population movements help to propagate schistosomiasis, as evidenced by its introduction into an increasing number of periurban areas of north-eastern Brazil and Africa and among refugees in Somalia, Zimbabwe and Cambodia. Controlling the disease is of strategic relevance for development of tourism in endemic countries; it is recognized that tourists now increasingly acquire schistosomiasis as “off-track” tourism increases, sometimes with severe acute infection and unusual sequelae, including paralysis of the legs. 350 Global Epidemiology of Infectious Diseases Schistosome parasitic worms The major forms of human schistosomiasis are caused by 5 species of flatworm, or blood flukes, called schistosomes. Urinary schistosomiasis, caused by Schistosoma haematobium, is endemic in 54 countries in Africa and the Eastern Mediterranean. Intestinal schistosomiasis, caused by the mansoni worm, occurs in 53 countries in Africa, the Eastern Mediterranean, the Caribbean and South America. In 41 countries, both parasites are endemic. Another form of intestinal schistosomiasis caused by S. intercalatum has been reported in 7 central African countries. Oriental or Asiatic intestinal schistosomiasis, caused by the S. japonicum group of parasites (including S. mekongi in the Mekong river basin), is endemic in 7 countries in the South-East Asia and the Western Pacific region. Simple and inexpensive diagnosis Diagnostic methods for schistosomiasis have recently been reviewed (Feldmeier & Poggensee 1993) (Table 12.1). For diagnosis of urinary schistosomiasis, a simple sedimentation can be used efficiently. A syringe filtration technique using filter paper, or polycarbonate or nylon filters, makes it possible for a team of 5 to examine up to 200 children in an hour and a half. Children infected with S. haematobium nearly always have microscopic or visual blood in their urine (haematuria). Children needing treatment can be identified by merely looking at the urine specimen or checking for microscopic blood using chemical reagent strips. The eggs of intestinal schistosomiasis can be detected in faecal specimens by a technique using cellophane soaked in glycerine, the Kato or Kato-Katz technique, or Table 12.1 Diagnostic methods for schistosomiasis Technique Parasite Sensitivity of single examination Reference Kato-Katz S. mansoni 100% - infections > 100 epg Sleigh et al. (1982) Glass slide Questionnaire Reagent strips Reagent strips Reagent strips Filtration 45% - infection 1–50 epg 100% - infections > 50 epg Teesdale, Fahringer & Chitsulo (1985) S. haematobium 77–92% - infections > 64 eggs/10 ml Mott et al. (1985a) of urine over age 15 years Specificity = 62–82% S. haematobium Sensitivity = 97 per cent > 64 eggs/ Mott et al. (1985b) (haematuria) 10ml of urine Specificity = 85% S. haematobium Sensitivity = 88% Stephenson et al. (1984) (haematuria) Specificity = 97% S. mansoni S. haematobium Sensitivity = 67% overall = 83% (haematuria) = >50 eggs/10ml of urine Specificity = 80% S. haematobium 100% - infections > 20 eggs/10ml epg = eggs per gram Savioli et al. (1990) Warren et al. (1978) Schistosomiasis 351 between glass slides. The cost of the tests requiring a microscope is now US$0.01 or less. The chemical reagent strips cost about US$0.05. Safe and effective treatment Three safe, effective drugs are now available for schistosomiasis and all can be taken by mouth. Praziquantel, oxamniquine and metrifonate are all included in the WHO’s model list of essential drugs (World Health Organization 1990). Their discovery has revolutionized treatment of this disease. Praziquantel, effective against all forms of schistosomiasis, has proven to have few and only transient side-effects and millions of people have benefited from treatment. Oxamniquine is used exclusively to treat intestinal schistosomiasis in Africa and South America. Metrifonate has now proved to be safe and effective for the treatment of urinary schistosomiasis. Box 12.1 Schistosomiasis: life history of a worm People contaminate the environment by their unsanitary habits. They acquire schistoso- miasis infection through repeated daily contact with fresh water during fishing, farming, swimming, bathing, washing and recreational activities. Eggs, excreted in urine or faeces from an infected person, break open on reaching water, releasing a tiny parasite (a miracidium) which swims frantically through the water by means of the fine hairs (cilia) covering its body, in search of a freshwater snail in which it can develop further. The parasite must find a snail host within 8–12 hours before it perishes. Once it has penetrated the snail, the parasite divides many times until thousands of new forms (cercariae) break out of the snail into the water. This phase of development takes 4–7 weeks or longer, depending on the type of parasite. Outside the snail, the cercariae, which have long forked tails, can live for 48 hours at the longest. They can penetrate a person’s skin within a few seconds in order to continue their growth cycle. As the cercaria penetrates the skin using secretions from its special glands, its tail falls off. Within 48 hours it has passed completely through the skin into the blood vessels. Sometimes this process causes itching (swimmer’s itch or cercerial dermatitis), but otherwise most people never notice it. Within weeks, the young parasite transforms itself into a long worm—either a male or a female. The female can produce eggs only when a male worm is present. Male and female adult worms remain joined together for life (less than 5 years on average, though they can live for up to 40 years); the more slender female is held permanently in a groove in the front of the male’s body. Once eggs are produced—over 200 per day depending on the species—the cycle starts again. In intestinal schistosomiasis, the worms attach themselves to the walls of the blood ves- sels lining the intestines. In urinary schistosomiasis, they live in blood vessels of the bladder. Only about half of the eggs leave the body in the faeces (intestinal schistosomiasis) or in the urine (urinary schistosomiasis); the rest remain embedded in the body, damaging organs, including liver, spleen, heart and esophagus. Heavy infections with schistosome parasites, occurring mainly in children, cause the actual disease. The eggs laid by the female worm—not the worms themselves—damage the bladder, intestines or other organs. 352 Global Epidemiology of Infectious Diseases Even though re-infection may occur after treatment, the risk of developing severely diseased organs is diminished and even reversed in young children in the short term for all forms of schistosomiasis (Jordan, Webbe & Sturrock 1993). In most areas, a reduction in the overall number of cases has been obtained by mass treatment and maintained for 1.5–2 years and in other areas up to 5 years without further intervention. No long-term follow-up studies after treatment of S. haematobium or S. japonicum are available. Definition and measurement Historically, the term “bilharziasis” was commonly used for schistosomiasis. Following an expert group report on the epidemiology and control of schistosomiasis (World Health Organization 1967), “schistosomiasis” became the official terminology. Schistosomiasis is a group of diseases, rather than a single entity. The standard criteria on diagnosis of each type of schistosomiasis is the presence of eggs in excreta. The use of the term “schistosomiasis,” however, reduces the possibility of discriminating between the different diseases associated with each type of infection. Efforts were undertaken to refine the reporting of morbidity attributable to schistosomiasis in the development of the Ninth Revision of the International Classification of Diseases (ICD-9), whose preparation began in 1969 and concluded in 1975. The ICD-9 categories are as follows: 120.0 Schistosoma haematobium 120.1 Schistosoma mansoni 120.2 Schistosoma japonicum 120.3 Cutaneous 120.8 Other Infection by Schistosoma: Bovis Intercalatum Mattheei Spindale cherstermani 120.9 Schistosomiasis, unspecified In ICD-9, the adaptation of ICD-O for oncology permitted the designation of the cell type of carcinoma. This was relevant to bladder cancer associated with schistosomiasis (S. haematobium only) in which squamous cell types predominate (M805–M808) as compared to transitional cell types (M812–M813). Deficiencies in the description of morbidity attributable to schistosomiasis (in contrast to the detail of morbidity regarding amoebiasis and Chagas Schistosomiasis 353 disease) were addressed in preparations for the tenth revision of the International Classification of Diseases (ICD-10). The First Expert Committee on Control of Schistosomiasis proposed eight new definitions to be integrated into ICD-10 either under B65 or as a qualifier under the disease state (World Health Organization 1985). These were provided to the secretariat of ICD-10 as part of the background preparatory documents. The suggestions were partially accepted. In ICD-10, schistosomiasis is classified as B65 and each of the three main species infecting humans are classified separately. S. intercalatum and S. mekongi come under B65.8—other schistosomiases. The ICD-10 categories are as follows: B65.0 Schistosomiasis due to Schistosoma haematobium [urinary schistosomiasis] B65.1 Schistosomiasis due to Schistosoma mansoni [intestinal schistosomiasis] B65.2 Schistosomiasis due to Schistosoma japonicum [Asiatic schistosomiasis] B65.3 Cercarial dermatitis B65.8 Other schistosomiases Infections due to Schistosoma: Intercalatum Mattheei Mekongi B65.9 Schistosomiasis, unspecified The category for infection attributable to S. chestermani (the same as S. intercalatum) in ICD-9 was removed in ICD-10. A number of codes were introduced to classify clinical sequelae of schistosomiasis infection in ICD-10 (see Box 12.2). In ICD-10, haematuria attributable to schistosomiasis is classified under R31 rather than N02 since the origin is the bladder rather than the glomeruli. Unlike most other infectious diseases, infection with Schistosoma is epidemiologically and clinically different from disease attributable to Schistosoma. Schistosomiasis is a chronic infection whose onset is difficult to detect. Thus, incidence is generally not measured; if it is, it reflects only the specific geographical area studied and cannot be extrapolated. Prevalence of schistosomiasis is the usual epidemiological variable measured. Morbidity (disease or impairment of the normal state that affects performance of vital functions) is a result of prior or concurrent infection with Schistosoma. With currently available techniques it is, however, difficult to measure clinical impairment in the early stages of disease or in light infections. Schistosomiasis infection is thought to be associated with increased risk of other parasitic infections. Aside from the theoretical assumptions on 354 Box 12.2 Global Epidemiology of Infectious Diseases Tenth revision of the International Classification of diseases (ICD 10) codes for schistosomiasis sequelae ICD 10 code Sequela K77.0 Liver disorders in infectious and parasitic diseases classified elsewhere Hepatosplenic schistosomiasis J70.8 Cor pulmonale attributable to schistosomiasis would probably be designated under J70.8 N29.1 Disorders of the kidney and ureter in schistosomiasis (bilharziasis) Portal hypertension in schistosomiasis N33.8 Bladder disorder in schistosomiasis (bilharziasis) R31 Haematuria due to schistosomiasis is to be classified under R31 rather than N02 since the origin is the bladder rather than the glomeruli Source: World Health Organization (1987) the frequency of the association of schistosomiasis with other diseases at the population level (Bundy et al 1991), published reports on community studies in Brazil (Lehman et al. 1976), Chad (Buck et al. 1970), Egypt (El Malatawy et al. 1992), Kenya (Thiongo and Ouma 1987), Sierra Leone (White et al. 1982), and Zambia (Wenlock 1979), have shown that Schistosoma infection is not consistently associated with other parasitic infections. When present, the consequences of multiple infections on morbidity, disability and mortality have not been adequately studied. In summary, a “gold standard” for the measurement of schistosomiasis morbidity has not been developed. This chapter will attempt to clarify this matter so that, within national development priorities, the impact of schistosomiasis can be accurately identified and the ensuing changes induced by control and development can be reliably assessed. Potential confounding diseases Of all the types of schistosomiasis, disease attributable to Schistosoma haematobium is the most likely to be confounded with other diseases. When haematuria is present, the disease may be misdiagnosed as glomerulonephritis (ICD-10 code N02.9) or haematuria of unknown etiology (ICD-10 code R31). In a known endemic area, the direct correlation between gross and microscopic haematuria and infection attributable to S. haematobium permits the valid use of haematuria as a proxy for infection even after large scale treatment. The picture may be further complicated if renal biopsy is performed, since both S. mansoni and S. haematobium infection have been associated with morphological changes with clinical manifestations Schistosomiasis 355 ranging from nephrotic syndrome (ICD-10 code subdivisions .5 and .6, i.e. membranoproliferative glomerulonephritis types 1–3). In most endemic areas of schistosomiasis attributable to S. haematobium, the incidence of squamous cell bladder cancer is high. Epidemiological data do not usually cite the association with schistosomiasis as an etiological agent. Such detail is subject to the thoroughness of the surgical pathological studies (incomplete in endemic areas) or the training and personal interest of the surgeon (as in Egypt). Although most people in the endemic areas have light infections with no symptoms, the economic and health effects of schistosomiasis should not be underestimated. In the north-east of Brazil, in Egypt, and in Sudan, the work capacity of rural inhabitants is severely reduced because of the weakness and lethargy caused by the disease. The school performance and growth patterns of infected children are also retarded; after treatment, remarkable improvement occurs. A major constraint in the determination of the burden of disease of schistosomiasis is that it does not occur in isolation. It is one of the diseases of poverty and is linked to all other infectious and parasitic diseases of poverty. Within any community, it most affects the poorest segment of the population. Data sources available and their biases The standard references for the current distribution and prevalence of schistosomiasis are two WHO documents. In the CEGET/WHO Atlas of the global distribution of schistosomiasis (Doumenge et al. 1987), the geographical distribution of all forms of schistosomiasis is identified on the basis on an exhaustive compilation of published epidemiological studies as well as surveys carried out by ministries of health. The prevalence was estimated (extrapolated) using this geographical representation of data and the actual surveys or epidemiological data (Utroska et al. 1989). The heterogeneity of the spatial distribution of the prevalence of parasitic diseases is widely recognized (Ashford, Craig & Oppenheimer 1992, 1993). The clustering of heavy infections attributable to S. mansoni and S. japonicum is also heterogeneous (Mott 1982). With the exceptions of Algeria, Brazil, Botswana, China, Egypt, Mali, Malawi, Morocco, and Zimbabwe, however, the data available to Ministries of Health are not disaggregated to the village level; thus most national statistics are intrinsically inadequate. Review of the empirical databases by world bank regions There is no single standard database for schistosomiasis. Schistosomiasis is not usually reported as an infection nor are any of its disease manifestations reported separately (Iarotski & Davis 1981). The CEGET/WHO Atlas of the global distribution of schistosomiasis (Doumenge et al. 1987) 356 Global Epidemiology of Infectious Diseases comprises the most complete database to the village level of endemic countries. Systematic extrapolation of these data to obtain accurate national estimates of the prevalence or morbidity is not possible because of the heterogeneity of the data and the distribution of the infection and disease within any particular country. Despite these limitations, an attempt has been made to estimate prevalence in order to assess the potential need for antischistosomal drugs (Utroska et al. 1989). Unlike the analysis of information on acute infectious diseases, the interpretation of incidence data has many limitations in schistosomiasis (Jordan, Webbe & Sturrock 1993). To calculate incidence accurately requires greater sensitivity of the diagnostic techniques than those currently available or in use by health services (Sleigh & Mott 1986). Any age-specific data on prevalence of schistosomiasis attributable to S. mansoni and S. haematobium should be interpreted with caution and compared with the age distributions shown in Tables 12.2–12.8. There seems to be no standard age distribution for S. japonicum. This is probably because no endemic area has not been exposed to intervention since 1950 and in some areas, transmission has been eliminated. Extensive analytical reviews of the published literature on the morbidity of schistosomiasis have been conducted by Gryseels (1989) and Sleigh & Mott (1986). There is general agreement that heavy infections, as measured by faecal or urinary egg counts, are associated with disease. Most of the relevant studies are referred to in Tables 12.2–12.8 or discussed in the text of this review. The major information gap is the lack of statistics on schistosomiasis as a cause of outpatient visits, hospitalization and even as a cause of death. Established Market Economies Only Japan is endemic in this region. No new cases have been reported since 1978. Thus no mortality could be attributed below age 15 years. Formerly Socialist Economies Schistosomiasis is not endemic in these countries. India S. haematobium has been confirmed near Hyderabad. The extent of the problem requires further investigation and any estimates cannot be confirmed. China The current prevalence estimate of 1.5 million people infected, proposed by the Ministry of Public Health of the People’s Republic of China, is based on a national sample survey in 1989 (China Ministry of Public Health 1992). The sex ratio in the sample survey was 14.2:8.3 (male:female). Prevalence studies for China are summarized in Table 12.2. China 1990 Not stated Not stated Prevalence (%) Morbidity Surveys on nutrition and health in 49 counties Source S. Japonicum empirical databases, China 65 counties Location Date Table 12.2 M 0 0 1 2 5 8 17 30 42 70 95 122 197 181 247 209 231 29 1–4 5–9 10–14 15–19 20–24 25–29 30–34 35–39 40-44 45–49 50–54 55–59 60–64 65–69 70–74 75-79 80+ All 0 1 0 1 1 2 6 11 16 34 58 84 108 135 183 177 221 21 F Mortality Rate per million Age group (years) Results Junshi et al. (1990) Reference continued Note: rectal cancer and colon cancer (registered separately) had lower cumulative mortality rates for 0–64 years than schistosomiasis. All these were significantly associated with schistosomiasis japonica. Data from 49 counties used in calculations. Only 20 are endemic. Mortality estimates are probably low. Comments Schistosomiasis 357 China 1992 11.8 57 600 advanced cases in the entire country Prevalence (%) Morbidity Stratified group random sample = 1 per cent of the population of the endemic area: Hubei, Hunnan, Anhui, Jiangxi Source 0–4 5–9 10–14 15–19 20–29 30–39 40–49 50–59 60+ M F Age group (years) Results S. Japonicum empirical databases, China (continued) M = males, F = females Location Date Table 12.2 2.8 6.9 10.0 13.6 12.3 8.4 13.4 13.5 12.8 14.2 8.3 Prevalence (%) China Ministry of Public Health Reference Comments 358 Global Epidemiology of Infectious Diseases Schistosomiasis 359 Other Asia and Islands Schistosomiasis attributable to S. japonicum and S. mekongi are present in five countries of this region. Available prevalence data are summarized in Table 12.3. The prevalence of infection resulting from S. japonicum in Indonesia is less than 1 per cent among a population at risk of 10 000 persons. S. malayensis has only been reported in 10 persons of a single ethnic group in Malaysia. In Thailand, less than 5 cases have been reported. Thus schistosomiasis is prevalent only in Cambodia and the Lao People’s Democratic Republic (S. mekongi) and the Philippines (S. japonicum). There are no studies on comparative morbidity between S. japonicum and S. mekongi. The distribution in the Lao People’s Democratic Republic overlaps with a higher prevalence of opisthorchiasis, and increased morbidity may be present in these areas. In the Philippines, Tanaka et al. have attempted to measure incidence among schoolchildren (Tanaka et al. 1984) and to assess the impact of treatment on incidence (Tanaka et al 1985). Incidence at age 7 years was 18.6 per cent and at age 12 years was 33.9 per cent; no difference between sexes was observed. Sub-Saharan Africa Schistosomiasis attributable to S. mansoni or S. haematobium is endemic in 41 countries of this region. It is not endemic in Cape Verde, Comoros, Djibouti, and Lesotho. Representative population-based studies of the prevalence and morbidity of schistosomiasis are shown in Table 12.4. Brinkmann et al. (1988) have published data on age and sex distribution in 13 village clusters in Mali of infection with S. mansoni or S. haematobium. Schistosomiasis resulting from S. intercalatum has received more attention recently and is now reported from eight West and Central African states. The geographical extent of risk is unknown, but the morbidity seems low (see Table 12.5). One example of the limitations of estimating national prevalence of infection has been published by Ratard et al. (1992) concerning Cameroon. The data derived from a national school survey of 5th grade primary school students (average age 13 years: range 10–19 years) were extrapolated to show a maximum number of 719 000 cases (range 392 900–1 027 800) in contrast to an estimate of 2 239 591 cases (range 1 261 355–3 267 963), which was based on an assessment of published literature (Utroska et al. 1989). In general, national surveys have not been done in sub-Saharan African; where available, as in Zimbabwe (Taylor & Makura 1985), the limitations of these data are recognized and no attempt is made to estimate national prevalence. Latin America and the Caribbean Only schistosomiasis attributable to S. mansoni is present in this region. Several prevalence distributions are presented in Table 12.6. Mortality due to schistosomiasis in Brazil and Surinam is discussed below. Location Philippines Philippines 1979 1980 43 40.1 Prevalence (%) Hepatomegaly and splenomegaly assoication with infection 25% of those infected had hepatomegaly or splenomegaly Morbidity N = 755 (90% of the residents) Village N = 851 (45% of the residents) Village Source S. japonicum empirical databases by region, Other Asia and Islands Date Table 12.3 0 14 57 63 73 62 71 57 2 10 37 47 44 65 60 50 F M 0 12 40 68 81 81 74 77 65 63 63 Age group (years) <1 1–4 5–9 10–14 15–19 20–24 25–34 35–44 45–54 55–64 65+ 0 0 21 48 52 25 46 59 58 23 17 F Prevalence (%) M 1–4 5–9 10–14 15–19 20–29 30–39 40–49 50+ Prevalence (%) Age group (years) Results WHO workshop (1980) Lewert,Yogore & Blas (1979) References 360 Global Epidemiology of Infectious Diseases Philippines 1983 M = males F = females Philippines 1980 Village C: 44 Village B: 39 Village A: 26 35 Hepatomegaly and splenomegaly associated with infection Hepatomegaly associated with infection Village C N = 1113 Village B N = 824 Village A N = 289 N = 1010 Village M 10 13 37 19 39 0 47 33 25 20 Age group (years) A F 1–4 5–9 10–14 15–19 20–24 25–29 30–39 40–49 50–59 60+ 5 27 18 44 40 33 53 30 0 0 4 17 49 68 44 57 50 51 47 42 9 31 52 46 30 23 41 48 31 39 M B 43 44 35 40 58 35 35 48 34 39 F Village F 8 33 70 69 83 39 45 55 53 62 M 8 10 18 46 18 25 38 53 37 11 Prevalence (%) M 1–4 5–9 10–14 15–19 20–24 25–29 30–39 40–49 50–59 60+ Prevalence (%) Age group (years) C 6 24 46 52 29 50 47 55 49 49 F Domingo et al. (1980) Schistosomiasis 361 Location Uganda Kenya Date 1972 1976 Table 12.4 82% 89% Prevalence 0 6 3 9 9 Haematemesis (ever) (%) Splenomegaly 3% vs 0 in uninfected 0–4 5–9 10–19 20–39 40+ Age group (years) Morbidity Village N=416 Village N = 231 Source S. mansoni empirical databases, sub-Saharan Africa 1–4 5–9 10–19 20–29 30–39 40–49 50–59 60–69 70+ Age group (years) <1 1–4 5–9 10–14 15–19 20–29 30–39 40–49 50–59 60+ Age group (years) Results 71 91 1000 97 80 73 67 1000 20 M 47 95 97 80 67 70 58 67 86 F Prevalence (%) 30 60 80 95 100 100 100 100 100 100 Prevalence (%) (estimated from graph) Siongok et al. (1976) Ongom & Bradley (1972) Reference Kato technique used Raw data not presented Comments 362 Global Epidemiology of Infectious Diseases Ethiopia Kenya 1976 1979 47% Village B = 10.7% Village A = 43.3% Abdominal pain Low 30% household sample Village B N = 155 Village A N = 352 50% household sample: 12 52 93 85 81 48 36 22 1–4 5–9 10–14 15–19 20–29 30–39 40–49 50+ 0–4 5–9 10–14 15–19 20–24 25–29 30–39 40–49 50–59 60+ Age group (years) M 0 6 18 5 8 0 0 0 M 15 23 57 73 85 78 58 1000 71 90 M 17 32 55 50 38 67 57 47 67 50 F Prevalence (%) 8 49 60 25 27 25 9 36 F 20 29 0 0 0 0 7 F Village B Prevalence (%) Village A Age group (years) Smith, Warren & Mahmoud (1979) Hiatt (1976) continued Kato technique used Kato technique used Schistosomiasis 363 Burundi Uganda 1988 1992 M = malesF = females Location Date Table 12.4 F = 81% M = 82% 33% Prevalence High intensity of infection Hepatomegaly and splenomegaly associated with infection Morbidity 20% population sample n = 6 203 5% population sample Source M M 51 83 93 90 90 91 75 0–4 5–9 10–14 15–19 20–29 30–39 40+ 73 77 94 91 81 76 84 F Prevalence (%) (estimated from graph) F Prevalence (%) 11 27 43 46 44 37 34 35 36 Age group (years) 0–4 5–9 10–14 15–19 20–29 30–39 40–49 50–59 60+ Age group (years) Results S. mansoni empirical databases, sub-Saharan Africa (continued) Kabaterine et al. (1992) Gryseels (1988) Reference Kato technique used Comments 364 Global Epidemiology of Infectious Diseases Equatorial Guinea Gabon 1990 1992 M = males, F = females Location 29 21 Splenomegaly caused by infection Diarrhoea and blood in the stool associated with infection Prevalence (%) Morbidity Village N = 354 Periurban N = 1221 Source 12 38 32 18 7 3 0–4 5–9 10–14 15–24 25–44 45+ 10 35 48 26 9 9 F M 14 48 52 60 50 25 0 0 7 Age group (years) 0–4 5–9 10–14 15–19 20–29 30–39 40–49 50–59 60+ 8 46 65 53 35 22 18 0 4 F Prevalence (%) M Age group (years) Prevalence (%) Results S. intercalatum empirical databases by region, sub-Saharan Africa Date Table 12.5 Martin-Prevel et al. (1992) Simarro, Sima & Mir (1990) Reference Sedimentation used first, Kato for quantification. Kato technique used. Note infection almost absent after age 45 years. Comments Schistosomiasis 365 Brazil Brazil 1962 1976 80 83 Prevalence (%) 52% spleen Morbidity Groups N = 363 Town Source <1 1–4 5–9 10–14 15–19 20–24 25–34 35–64 65+ Age group (years) 10 39 65 69 89 1000 85 80 73 M 50 33 77 86 86 1000 96 90 79 F Prevalence (%) Age 10–14 years = 50% of spleen and heaviest infections Results Lehman et al. (1976) Kloetzel (1962) Reference S. mansoni empirical databases by region, Latin America and the Caribbean M = males, F = females Location Date Table 12.6 Kato technique used First study of its kind Limited to children Comments 366 Global Epidemiology of Infectious Diseases Schistosomiasis 367 Middle Eastern Crescent S. haematobium is endemic in 15 countries in this region and S. mansoni is present in 7 of these countries. Some distributions of prevalence by age and sex are shown in Tables 12.7 and 12.8. Mortality from schistosomiasis Mortality from schistosomiasis has been poorly documented in most endemic countries. Death certificates and patients’ records rarely identify schistosomiasis as the underlying cause of death. Well-designed prospective autopsy studies have nevertheless been published (Cheever et al. 1978). In that study, schistosomiasis was the direct cause of death in 9.2 per cent of the infected individuals and was the attributable cause of death in 6.2 per cent of all deaths. The breakdown of these causes among those infected is shown in Table 12.9. Table 12.10 summarizes the results of available mortality studies by World Bank region; these results are discussed below. Sub-Saharan Africa Annual mortality attributable to S. haematobium infection in East Africa has been estimated at 1–2 per 1000 infected adults (Forsyth 1969). In Chad, in 1966, the mortality rate of schistosomiasis attributable to S. haematobium infection was 1 per 1000 infected persons (Buck et al. 1970). Latin American and the Caribbean In 1984, the annual mortality attributable to schistosomiasis caused by S. mansoni in Brazil was estimated at 0.5 per 100 000 total population; at the same time in Suriname the figure was estimated to be 2.4 per 100 000 inhabitants. The control of schistosomiasis through large-scale chemotherapy in Brazil was associated with a decline in annual mortality between 1977 and 1988, from 0.67 to 0.44 deaths per 100 000 inhabitants. S. intercalatum infection has never been reported as a cause of death in the region. China Before the introduction of praziquantel in China, severe acute schistosomiasis attributable to S. japonicum had a 2.5–20.7 per cent mortality rate, and in Leyte, the Philippines, the annual mortality among 135 untreated patients with severe schistosomiasis was 1.8 per cent (Blas et al. 1986). It is expected that the more widespread use of current antischistosomal drugs for morbidity control in highly endemic areas will also reduce mortality. The national survey by the Ministry of Health of China appears to confirm the total annual mortality rate. A total of 982 severe cases were detected among 165 384 people examined; (16 953 people in the sample cohort were infected). If it is assumed that only infected people had severe disease, the rate of severe cases was 5.79 per cent. This is probably not reasonable since many advanced cases do not excrete eggs. The peak mor- Egypt Egypt 1981 1982 37 29 Prevalence (%) Morbidity 6 villages N = 5 998 2 villages N = 2403 Source <1 1–4 5–9 10–14 15–19 20–24 25–29 30–39 40–49 50–59 60+ F 20 18 36 43 18 15 9 5 7 5 12 M 20 57 75 66 41 42 43 42 40 32 Age group (years) 1–4 5–9 10–14 15–19 20–24 25–29 30–39 40–49 50–59 60+ 21 41 49 35 21 17 11 12 13 13 F Prevalence (%) M 18 19 55 75 54 30 35 27 30 26 24 Age group (years) Prevalence (%) (estimated) Results S. haematobium empirical databases by region, Middle Eastern Crescent M = males, F = females Location Date Table 12.7 King et al. (1982) Mansour et al. (1981) Reference Comments 368 Global Epidemiology of Infectious Diseases S. mansoni: 40.5 4 10 23 6 17 6 N = 8712 25% sample in 8 villages Village N = 1747 2–5 6–10 11–15 16–20 21–30 31+ Age group (years) 5–9 10–19 20–29 30–39 40–49 50+ Age group (years) Egypt 9 22 29 20 17 8 Infected (%) Not infected (%) 1977 1–4 5–9 10–14 15–19 20–39 40+ Age group (years) Spleen Source Sudan 1976 Morbidity Prevalence Location Date 48 S.mansoni empirical databases by region, Middle Eastern Crescent Table 12.8 F 24 80 50 45 22 15 15 33 50 40 25 18 Prevalence (%) (estimated from graph) 24 80 60 55 40 50 M Prevalence (%) Results El-Alamy & Cline (1977) Omer et al. (1976) Reference continued Comments Schistosomiasis 369 M = males, F = females 74 Egypt M 134 204 331 165 233 194 194 83 Age group (years) 1–4 5–9 10–19 20–29 30–39 40–49 50–59 60+ 95 67 213 199 151 109 112 100 F Intensity (geometric mean no. eggs/gram) Village N = 537 M 25 71 89 80 86 80 75 50 Age group (years) 1–4 5–9 10–19 20–29 30–39 40–49 50–59 60+ 33 57 81 71 78 56 55 50 F Prevalence (%) Results 1980 Source Prevalence Location Date Morbidity S.mansoni empirical databases by region, Middle Eastern Crescent (continued) Table 12.8 Abdel-Wahab et al. (1980) Reference Comments 370 Global Epidemiology of Infectious Diseases Schistosomiasis Table 12.9 371 Causes of death related to schistosomiasis Attributed cause of death Percentage of deaths Mean age (years) Obstructive uropathy Bladder cancer Symmer’s fibrosis Colonic polyposis Salmonellosis 2.5 2.4 3.1 0.8 0.4 42 38 30 35 — Total deaths in infected persons 9.2 Source: Cheever et al. (1978). tality was in the 45–59 year age group. Overall mortality rates were 29 per million population for males and 21 per million population for females. The most recent data on mortality from to schistosomiasis have been produced by a study of 65 Chinese counties (Junshi et al. 1990). Schistosomiasis ranked as the most significant cause of death resulting from colorectal, rectal and colon cancer. Only 20 of these counties were currently endemic for schistosomiasis. The cumulative mortality rate attributable to schistosomiasis for both sexes was as high as 38.99 per 1000 deaths. S. japonicum is possibly carcinogenic in humans, causing colorectal carcinoma. The cumulative data from China support this conclusion; see also the comprehensive review by the International Agency for Research on Cancer (1994). Other Asia and Islands One in 20 deaths in the Khong district of the Lao People’s Democratic Republic were attributable to schistosomiasis (S. mekongi) according to survey data obtained in preparation for the national control programme on opisthorchiasis and schistosomiasis (A. Sleigh, personal communication 1992). The case fatality rate of schistosomiasis attributable to Schistosoma japonicum infection in the Philippines is estimated to be 1.78 per cent per year (Blas. et al 1986). The recognized number of cases in Thailand and Malaysia is so low that neither the populations nor the populations at risk should be included in the totals. Middle Eastern Crescent Mortality from schistosomiasis is the cause of 6 per cent of all deaths in Egypt. The annual crude mortality rate is 1 per 1000 infected individuals, and is higher in men than in women. The male-to-female sex ratio is 47.2:23.0, based on the combined raw data of King et al. (1982) and Mansour et al. (1981). Reference China Ministry of Public Health (1992) El Malatawy et al. (1992) Utroska et al. 1989. Lopes (1984) Buck et al. (1970) Blas et al. 1986 A. Sleigh (personal communication 1992) Region and type China — S. japonicum only Middle East Crescent — S. mansoni Middle East Crescent — S. haematobium Latin America & Caribbean — S. mansoni only Sub-Saharan Africa — S. haematobium Other Asian Countries & Islands — S. japonicum (Philippines & Indonesia) S. mekongi (Lao PDR & Cambodia) Females Males Females Males Females Males Females Males Females Males Females Males Females Males Sex 0-14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ 0–14 15–65+ Age group (years) 117 742 000 134 580 000 116 848 000 141 104 000 80 845 000 140 766 000 78 420 000 144 264 000 106 506 000 149 883 000 101 733 000 144 953 000 106 506 000 149 883 000 101 733 000 144 953 000 157 240 000 427 959 000 148 347 000 400 147 000 Population (number) Table 12.10 Studies of mortality rates associated with schistosomiasis, World Bank regions 50 000 75 000 50 000 75 000 151 252 208 748 99 248 140 752 27 318 640 22 082 160 21 607 502 13 872 797 1 755 725 3 244 275 1 986 486 2 913 514 2 127 152 1 719 416 1 036 544 837 858 1 550 687 3 419 463 1 347 793 2 972 057 205 427 741 240 120 073 433 260 Number of people infected 5 25 5 25 3 55 2 39 27 2 208 21 1 000 3 64 3 62 21 1 719 10 600 1.5 34 3 30 50 530 40 380 Number of deaths per year 372 Global Epidemiology of Infectious Diseases Schistosomiasis 373 Disability In a critical review, Prescott (1979) pointed out that, at that time, no clear empirical basis had been found for the assumption that schistosomiasis seriously impairs labour productivity. He attributed this to possible sampling bias resulting in the exclusion of workers with severe disease. Subsequent studies have lent more support to the thesis that not only does severe infection cause disability but infection alone limits productivity and well-being. In the workplace, decreased productivity has infrequently been shown to be the result of to schistosomiasis. In contrast, increased absenteeism has consistently been shown to be related to schistosomiasis. In irrigation workers, who are at greatest risk, schistosomiasis infection was associated with a 3.67 per cent greater absenteeism rate than among those not infected; in cane-cutters, schistosomiasis infection was associated with 1.64 per cent greater absenteeism (Foster 1967). These figures are based on observation of over 28 000 consecutive work shifts and the difference related to the irrigation workers was considered statistically significant. As Prescott points out, the determinants of absenteeism were not analysed. Since Prescott’s review, a substantial amount of data has been published supporting the hypothesis that infection without clinical manifestations causes work days lost: 26.2 to 41.6 work days lost per year as a result of S. japonicum (Blas 1989); 4.4 work days per year as a result of S. haematobium (Ghana Health Assessment Project Team 1981, Morrow et al. 1984). Absenteeism was noted to be twice as high among labourers infected with S. mansoni as among those uninfected (Ndamba et al. 1991); 6 working days per year were estimated to be lost due to S. mansoni infection (Chowdhry & Levy 1988); in Madagascar 50 per cent of people reported unable to work had evidence of urinary schistosomiasis and remained out of work for at least 10 days (Breuil, Moyroud & Coulanges 1983). There is a latency period of 5–15 years between infection and development of severe disease due to S. mansoni, thus hepatosplenomegaly with portal hypertension attributable to S. mansoni occurs after age 20 years. Approximately 10 per cent of those infected will develop severe disease. Since 1980 a number of studies have shown that schistosomiasis resulting from S. mansoni has a negative effect on productivity. A reduction of 18 per cent in the work output of those individuals with heavy infection or hepatosplenomegaly can be expected (Awad El Karim et al. 1980). According to a recent study, physical performance increased 4.3 per cent after treatment, and work output (as measured by the amount of sugar cane cut in a given time) increased 16.6 per cent (Ndamba et al. 1993). Infected individuals without organomegaly were observed to have no reduction of work output (Van Ee & Polderman 1984). A minimum of 12 per cent lower exercise capacity was found in children with S. haematobium infection in Zimbabwe (Ndamba 1986). This deficit was recovered within one month after treatment. Similarly, in Kenya, a 374 Global Epidemiology of Infectious Diseases 7–10 per cent improvement in exercise capacity was found one month after treatment in children with S. haematobium infection (Latham et al. 1990). Of those with hepatosplenomegaly attributable to S. mansoni, 33 to 67 per cent have oesophageal varices (De Cock et al. 1982, Saad et al. 1991). Of those with oesophageal varices, 3 to 4 per cent have recurrent upper gastrointestinal bleeding. After oesophageal varices have bled once, there is a tendency for the bleeding to recur, and 3–4 months per year will be lost either in or out of hospital. Disease of the central nervous system, affecting the spinal cord, is more frequent and causes more disability than is currently recognized, especially among migrants into endemic areas of S. mansoni transmission (World Health Organization 1989, Joubert et al. 1990). In the Philippines, the minimum contribution of schistosomiasis attributable to S. japonicum to the etiology of epilepsy is 6 per 1000 of the total rate of 11 per 1000 (Hayashi 1979). Spinal cord involvement attributable to schistosomiasis (S. S. mansoni mostly) is the cause of 1 in 4 of all hospital admissions to the neurological nervice of a general hospital in Durban, South Africa (Haribhai et al. 1991). Average disability weights for cases of schistosomiasis infection were estimated using expert panels, person-trade off methods for 22 indicator conditions, and rankings of other conditions (including schistosomiasis) against the indicator conditions as described by Murray & Lopez (1996a). The resulting average disability weights were 0.005 for children aged 0–14 years and 0.006 for adults aged 15 years and over. Estimation of disability-adjusted life years General methods used for the calculation of disability-adjusted life years (DALYs) lost due to a disease are given by Murray & Lopez (1996a). DALYs are the sum of years of life lost due to mortality (YLLs) and years lived with disability (YLDs). The prevalence of schistosomiasis and deaths attributable to schistosomiasis were calculated for the eight World Bank regions, based on the data reviewed above. The resulting estimates are shown in Tables 12.11 and 12.12 and have also been published by Murray & Lopez (1996b). YLLs are calculated directly from the deaths shown in Tables 12.11 and 12.12 Total global deaths in 1990 resulting from schistosomiasis were estimated to be around 8000. It must be emphasised that these are direct schistosomiasis deaths and do not include the cancer and other disease deaths attributable to schistosomiasis (as discussed below). YLDs are usually calculated from regional estimates of incident cases and average durations. However, because of the very limited information available to estimate incidence rates for schistosomiasis from prevalence and case fatality data, YLDs for schistosomiasis were calculated directly from the prevalence estimates (in effect assuming a one year duration and prevalence equal to incidence). Schistosomiasis 375 Tables 12.11 and 12.12 also show the estimated DALYs for the eight World Bank regions and the regional variation in DALYs per 100 000 population resulting from schistosomiasis. Schistosomiasis as a risk factor for other diseases Cancer Squamous cell bladder carcinoma is the leading cause of cancer among 20–44 year old men in Egypt and is a leading cause of death in this age group (Koroltchouk et al. 1987). El-Bolkainy & Chu (1981) estimated that the rate of squamous cell carcinoma among people with S. haematobium infection in Egypt is 2 per 1000 and 4 per 1000 among rural residents. In an older study, squamous cell bladder cancer was estimated to occur in 10 of every 1000 people infected with S. haematobium (Halawani & Tamami 1955); 26 per cent of deaths attributable to schistosomiasis were associated with bladder cancer in Egypt. The latency between initial infection and onset was estimated to be 20–30 years, and the male-to-female ratio was 9.5:1. In Angola the peak age of incidence of squamous cell carcinoma of the bladder is 44 years of age. An incidence rate of 8 per 10 000 rural inhabitants is suggested on the basis of epidemiological and clinical data (Lopes 1984). In Malawi, Mozambique and Zambia, the incidence rates of squamous cell bladder cancer are eight times as high as in the United States of America or the United Kingdom (Lucas 1982). It has been estimated that primary prevention to control urinary schistosomiasis would reduce the global rate of carcinoma of the bladder by 5000–10 000 cases per year (Koroltchouk et al. 1987). Other diseases Intestinal schistosomiasis caused by S. mansoni and S. japonicum may complicate a variety of other diseases that increase disability. Typhoid fever. If a person with schistosomiasis acquires typhoid fever, the duration of the typhoid fever will increase 2–3 times the average duration and may lead to total incapacity or death in 30 per cent of untreated persons (Salih et al. 1977). Hepatitis B. The effect of schistosomiasis infection on concurrent hepatitis B has been recently reviewed, and the limitations of the interpretation of clinical and epidemiological data have been analysed (Chen et al. 1993). There is a consensus among clinicians in the highly endemic areas of China and Egypt that in patients with schistosomiasis the duration of hepatitis increases up to 5 times the average, and the risk of chronic liver disease is greater and may result in total incapacity (Ghaffar et al. 1990). Rate per 100 000 Prevalence Number (thousands) 0 0 0 0 0 0 0–4 5–14 15–44 45–59 60+ All Ages India 0–4 5–14 15–44 45–59 60+ All Ages 0 0 0 0 0 0 0 0 0 0 0 0 Formerly Socialist Economies (FSE) 0–4 5–14 15–44 45–59 60+ All Ages — — — — — — — — — — — — — — — — — — Established Market Economies (EME) Age group (years) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Number (thousands) Deaths — — — — — — — — — — — — — — — — — — Rate per 100 000 — — — — — — — — — — — — — — — — — — Number (thousands) YLLs Table 12.11 Epidemiological estimates for schistosomiasis,1990, males 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Rate per 100 000 YLDs — — — — — — — — — — — — — — — — — — Number (thousands) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Rate per 100 000 DALYs 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Number (thousands) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Rate per 100 000 376 Global Epidemiology of Infectious Diseases 35 170 501 145 96 947 35 167 192 55 37 485 11 148 45 013 39 991 7652 3442 107 246 0–4 5–14 15–44 45–59 60+ All Ages 301 1 455 2 549 496 199 5 000 79 198 119 162 182 141 59 175 164 199 196 162 23 478 64 068 38 540 37 682 32 758 42 504 1 047 2 791 2 444 2 230 1 398 2 256 Latin America and the Carribean 0–4 5–14 15–44 45–59 60+ All Ages Sub–Saharan Africa 0–4 5–14 15–44 45–59 60+ All Ages Other Asia and Islands 0–4 5–14 15–44 45–59 60+ All Ages China 0 0 1 1 0 2 0 0 0 0 0 0 0 0 0 0 0 1 — 0 1 3 5 1 — — — 0 1 0 — — 0 0 1 0 — — 3 1 — 5 1 12 22 8 3 47 — — 1 2 1 4 — — 4 4 2 11 2 3 4 2 17 21 39 29 19 0 0 1 6 5 1 0 0 1 6 4 2 1 8 22 3 1 35 24 261 348 47 14 695 — 1 2 — — 3 — 1 4 1 — 7 3 15 21 13 7 16 51 371 335 231 133 275 0 1 1 0 0 1 0 1 1 1 0 1 1 9 25 4 1 40 25 273 371 56 18 742 0 1 3 2 1 7 0 1 9 5 3 17 3 17 24 18 7 18 continued 53 389 358 276 171 294 0 1 2 6 5 2 0 1 3 7 6 3 Schistosomiasis 377 12 149 49 852 47 271 9 135 4 089 122 495 630 3 048 4 038 786 315 8 817 3 781 9 044 3 782 2 924 1 868 4 616 1 531 4 664 3 545 3 519 2 308 3 439 Deaths 1 2 1 1 5 1 1 Number (thousands) 1 2 2 1 Rate per 100 000 YLLs 1 19 46 21 10 97 — 7 15 6 2 30 Number (thousands) YLLs, years of life lost;YLDs, years lived with disability; DALYs, disability-adjusted life years. 0–4 5–14 15–44 45–59 60+ All Ages World 0–4 5–14 15–44 45–59 60+ All Ages Rate per 100 000 Prevalence Number (thousands) Middle Eastern Crescent Age group (years) 3 4 7 5 4 11 13 27 15 12 Rate per 100 000 Table 12.11 Epidemiological estimates for schistosomiasis,1990, males (continued) YLDs 26 289 412 56 17 800 1 18 35 5 1 60 Number (thousands) 8 52 33 18 8 30 2 28 31 22 7 23 Rate per 100 000 DALYs 28 308 458 77 27 898 2 25 50 10 4 91 Number (thousands) 9 56 37 25 12 34 5 38 44 45 29 35 Rate per 100 000 378 Global Epidemiology of Infectious Diseases Schistosomiasis 379 Furthermore, there is a decreased response to hepatitis B vaccine (Bassily et al. 1987, Ghaffar et al. 1990). Rift Valley Fever. During epidemics of Rift Valley Fever in Egypt and East Africa, it has been hypothesized that persons with intestinal schistosomiasis had a high mortality rate due to liver failure (WHO Regional Office for the Eastern Mediterranean 1983). The haemorrhagic form may occur in up to 6 per cent of patients and the mortality rate of this form of RVF is about 30 per cent. Economic Burden and intervention The economic constraints in developing countries where schistosomiasis is endemic require careful analysis and review to assess the feasibility of control with limited resources. The cost of disease and the price of health have been a focus of attention of the World Health Organization since its inception (Winslow 1951). Economic costs The economic costs of schistosomiasis are substantial. In 1953 schistosomiasis was estimated to cost Egypt about U$57 million dollars and to decrease productivity by 33 per cent (Wright & Dobrovolny 1953). The projected savings to be made by eradicating schistosomiasis from Japan was estimated in 1952 to be US$3 million per year (Hunter et al. 1952). In Fukuoka and Saga prefectures alone there were 28 617 infected people representing 24 213 520 man-hours lost per year and wages lost per year equivalent to US$2 522 200, while the estimated cost of treatment of infected invividuals per year was U$177 938 (U$6.22 per person). Audibert (1986) estimated that, in the rice irrigation projects of north Cameroon, a 10 per cent increase in the prevalence of schistosomiasis resulted in a 4.9 per cent decrease in rice output. Estimation of costs of interventions An estimation of the cost of different interventions is the first step in a cost-effectiveness analysis of control strategy options. In a strategy aimed primarily at morbidity control, the main elements are chemotherapy and health education. The simplest item to estimate is the actual cost of the antischistosomal drugs, but the cost of central and peripheral storage, transport to distribution centres, and administrative support must also be considered. After choosing the most appropriate drug and establishing a dosage schedule, the next cost is that of delivery. This will reflect the choice of drug and the dosage schedule. Cost calculations must also include the requirements for diagnosis. Selective diagnosis of S. haematobium infection using questionnaires or chemical-reagent strips costs relatively little compared with the high cost of stool examination to detect S. mansoni Number (thousands) Rate per 100 000 0 0 0 0 0 0 0–4 5–14 15–44 45–59 60+ All ages India 0–4 5–14 15–44 45–59 60+ All Aaes 0 0 0 0 0 0 0 0 0 0 0 0 Formerly Socialist Economies 0–4 5–14 15–44 45–59 60+ All ages — — — — — — — — — — — — — — — — — — Established Market Economies (EME) Age group (years) Prevalence 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Number (thousands) Deaths — — — — — — — — — — — — — — — — — — Rate per 100 000 — — — — — — — — — — — — — — — — — — Number (thousands) YLLs Table 12.12 Epidemiological estimates for schistosomiasis,1990, females 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Rate per 100 000 — — — — — — — — — — — — — — — — — — Number (thousands) YLDs 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Rate per 100 000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Number (thousands) DALYs 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Rate per 100 000 380 Global Epidemiology of Infectious Diseases 21 99 293 84 56 553 26 124 146 42 28 365 7 512 30 330 28 124 5 382 2 421 73 769 0–4 5–14 15–44 45–59 60+ All ages 340 1 646 2 289 446 179 4 900 61 154 91 120 12 107 36 110 103 131 108 101 15 972 43 442 26 468 24 333 19 017 28 598 1 230 3 244 2 199 1 908 1 062 2 200 Latin America and the Caribbean 0–4 5–14 15–44 45–59 60+ All ages Sub–Saharan Africa 0–4 5–14 15–44 45–59 60+ All ages Other Asia and Islands 0–4 5–14 15–44 45–59 60+ All ages China 0 0 0 0 0 1 0 0 1 0 0 1 0 0 0 0 0 1 0 0 0 0 0 1 — — 0 0 0 0 — 0 1 1 2 1 — — — 0 1 0 — — — 0 1 0 — — 3 1 — 5 — 8 19 4 1 33 — — — 1 — 1 — — 1 2 2 5 0 0 3 4 0 2 0 12 18 18 8 13 0 0 0 3 0 0 0 0 0 3 4 1 1 10 20 3 1 34 16 176 245 33 10 480 — 1 1 — — 2 — 1 3 1 — 4 4 20 19 13 6 15 34 252 231 149 79 186 0 1 1 0 0 1 0 1 1 2 0 1 1 10 22 4 1 38 17 184 264 37 11 513 0 1 2 1 1 5 0 1 4 3 2 9 4 20 21 17 6 17 continued 36 264 248 167 86 199 0 1 1 3 4 1 0 1 1 5 4 2 Schistosomiasis 381 Number (thousands) 8 307 34 175 33 846 6 537 2 917 85 781 408 1 976 2 994 584 234 6 194 2 686 6 503 2 824 2 101 1 084 3 282 1 028 3 187 2 792 2 614 1 512 2 511 Rate per 100 000 0 0 1 1 1 3 0 0 0 0 0 1 Number (thousands) Deaths 0 0 0 0 0 0 — — 0 1 1 0 Rate per 100 000 YLLs 0 9 33 11 6 58 — — 10 2 1 13 Number (thousands) YLLs, years of life lost;YLDs, years lived with disability; DALYs, disability-adjusted life years. 0–4 5–14 15–44 45–59 60+ All ages World 0–4 5–14 15–44 45–59 60+ All ages Middle Eastern Crescent Age group (years) Prevalence 0 2 3 4 2 2 0 0 9 9 7 5 Rate per 100 000 Table 12.12 Epidemiological estimates for schistosomiasis,1990, females (continued) 18 198 295 40 12 563 1 11 26 4 1 43 Number (thousands) YLDs 6 38 25 13 4 22 3 18 24 18 6 17 Rate per 100 000 19 207 328 51 17 622 1 11 36 6 2 56 Number (thousands) DALYs 6 39 27 16 6 24 3 18 34 27 13 23 Rate per 100 000 382 Global Epidemiology of Infectious Diseases Schistosomiasis 383 infection. The cost of large-scale chemotherapy will be affected by the method of identifying communities for treatment. Cost calculations frequently omit the cost of employing the necessary personnel. The cost of equipment, particularly vehicles, usually appears in the accounting, but the cost of offices and laboratories is rarely considered, nor is the depreciation of these capital assets. In a strategy that aims at morbidity control and the reduction of transmission, the additional costs of, for example, mollusciciding and environmental management must be considered. The cost of mollusciciding includes the purchase of molluscicides and the cost of their application (e.g. transportation, storage, administration, equipment and personnel). The cost of environmental management includes capital investments in engineering and equipment and recurrent operational costs (e.g. personnel and materials). Since the cost estimates above are intended to allow comparison, a uniform method of cost calculation must be used so that costs can be adjusted to reflect local variation. The costs of control Studies on the costs of control programmes based on chemotherapy were evaluated by the World Health Organization Expert Committee (WHO 1993), which noted the inadequacy of the available data. Costs estimated in studies ranged from US$0.70 to US$3.10 per person, but development costs, salaries, and the cost of failure were often excluded. The government health allocations for schistosomiasis control were last reviewed in 1976 (Iarotski & Davis 1981). Nineteen endemic countries responded to a World Health Organization questionnaire regarding: (1) the total health budget in US dollars and its proportion of the total national budget and (2) the allocation for schistosomiasis control in US dollars and as a proportion of the health budget. In those countries responding, between 2 per cent (Saudi Arabia) and 18 per cent (Dominican Republic) of the national budget was allocated to health and between 0.002 per cent (Indonesia) to 2.8 per cent (Saint Lucia) of the health budget was directly allocated to schistosomiasis control. Current expenditures for schistosomiasis control are not available. Some hypothetical projections have been made based on pilot projects. The cheapest regimen using praziquantel involved existing health services in areas of low endemicity. When specialized programmes were required, costs varied from US$1.50 to US$6.53 per person per year. Costs per infected person treated were inevitably higher, with a minimum cost of US$3.00 for the specialized programmes. When the annual per capita expenditure on all forms of primary health care is only between US$1 and US$4 in many endemic countries, most cannot afford vertical programmes, except perhaps in small areas. Drugs account for a relatively low proportion of total delivery costs in vertical programmes. On the other hand, their cost is more prominent and may be the major expenditure in programmes integrated into primary health 384 Global Epidemiology of Infectious Diseases care. A reduction in the price of praziquantel could significantly reduce the cost of such programmes. The cost of identifying communities with a high prevalence of S. haematobium infection can be substantially reduced by using questionnaires and reagent strips. Chemotherapy targeted at schoolchildren will also reduce the costs of control, and in most communities will reach the groups with the highest prevalence and severity of infection. Increasing the interval between treatments will also reduce costs, but this has to be balanced against a reduced impact on morbidity. Further studies are needed on ways of reducing the cost of identifying communities for treatment (especially for S. mansoni infection) and the cost of drug delivery. Estimation of effectiveness Effectiveness has been estimated in terms of coverage (number of people receiving treatment per number of people infected or at risk) or cure (coverage per drug efficacy). This allows estimation of the cheapest approach to delivering treatment, which is not necessarily the same as the most cost-effective approach to reducing disease. For a better assessment of effectiveness, more information is required on the morbidity attributable to schistosomiasis at the community level, the relationship between this morbidity and the prevalence and intensity of infection, and the relationship between morbidity and disease predisposition. This approach may require modelling the dynamics of transmission and morbidity, and the impact of control, in similar ways to methods developed for other helminthic infections. Indirect costs In assessing indirect costs of the disease, a distinction is traditionally made between the cost of health care for infected people and the cost of reduced economic productivity as a result of the disease. Little is known about the health-seeking behaviour of people suffering from schistosomiasis and the direct cost incurred by the health system; more research is needed in this area. Given levels of prevalence and intensity of infection can be associated with very different levels of clinical disease both between and within countries. Economic estimates show similar variations in the indirect cost of the disease. These inconsistencies do not imply that schistosomiasis has no demonstrable economic impact, but rather that the indirect costs might be considerable in certain circumstances, as in areas where the prevalence of clinical disease is high or in communities that have recognized schistosomiasis as a significant health problem. In the assessment of indirect costs, rural household income should be the focus of analysis rather than the wages of individual workers. The long-term economic consequences of the disease should also be explored. S. haematobium infection can be associated with malnutrition and stunting of growth in children. There is a positive correlation between Schistosomiasis 385 education and productivity, and if schistosomiasis inhibits educational achievement it could also eventually affect production. Discussion and conclusions This chapter has documented the data and methods used to estimate the global burden of schistosomiasis in 1990. The paucity of epidemiological data, particularly on the sequelae and consequences of schistosomiasis infection, have inevitably required that a very simplified disease model be used for the estimations. Ideally, a number of other factors should be considered in the calculation of the burden of disease attributable to schistosomiasis. Multiple schistosoma infections In 35 sub-Saharan countries and in 6 countries of the Middle Eastern Crescent, both S. mansoni and S. haematobium are present and overlap in some areas. The age and sex distribution is similar to that of each infection alone (Saladin et al. 1983, Granier et al. 1985, Kazura et al. 1985, El Malatawy et al. 1992). The only study to analyse the interaction between the two infections in the same individuals nevertheless showed that the intensity of S. haematobium infection is greater in those with concomitant S. mansoni infection than in those with S. haematobium alone; the converse was not observed (Robert, Bouvier & Rougemont 1989). Multiple parasitic infections Schistosomiasis may be only one of multiple parasitic infections in the same individual. Various intestinal helminthic infections have been associated with S. mansoni infection (Lehman et al. 1976) Polyparasitism is the rule rather than the exception in most developing countries. Its occurrence is, however, focal and wide variations are present within any country (Bundy et al. 1991). Estimating the proportional contribution of each of multiple infections to the burden of disease has not been attempted. Reinfection In endemic areas re-exposure to infection is constant or seasonal and is proportional to the overall prevalence, i.e. high prevalence is associated with high rates of reinfection. In assessing the burden of schistosomiasis, the risk of reinfection should be considered. It is hoped that future revised estimates of the global burden of disease for schistosomiasis will be able to take into account improved epidemiological information on the incidence, prevalence, sequelae and case fatality of the various types of schistosomiasis infection, and that control efforts will have resulted in reductions in the prevalence of this disease in most regions. 386 Global Epidemiology of Infectious Diseases References Abdel-Wahab MF et al. (1980) Schistosomiasis mansoni in an Egyptian village in the Nile Delta. American Journal of Tropical Medicine and Hygiene, 29:868–874. Ashford RW, Craig PS, Oppenheimer SJ (1992) Polyparasitism on the Kenya Coast. 1 – Prevalence and the association between parasitic infections. Annals of Tropical Medicine and Parasitology, 86:671–679. Ashford RW, Craig PS, Oppenheimer SJ (1993) Polyparasitism on the Kenya Coast: 2 – Spatial heterogeneity in parasite distributions. Annals of Tropical Medicine and Parasitology, 87:283–293. Audibert M (1986) Agricultural non-wage production and health status: A case study in a tropical environment. Journal of Development Economics, 24:275–291. Awad El Karim MA et al. (1980) Quantitative egg excretion and work capacity in a Gezira population infected with S. mansoni. American Journal of Tropical Medicine and Hygiene, 29:54–61. Bassily S et al. (1987) Immunogenicity of hepatitis B vaccine in patients infected with Schistosoma mansoni. American Journal of Tropical Medicine and Hygiene, 36:549–553. Blas BL et al. (1986) An attempt to study the case fatality rate in Schistosoma japonicum infection in the Philippines. Southeast Asian Journal of Tropical Medicine and Public Health, 17:67–70. Blas B (1989) The Schistosomiasis Control Programme in the Philippines. Meeting on the Strategy of Control of Morbidity due to Schistosomiasis, Geneva, 10–13 October 1989. World Health Organization, Geneva, (unpublished abstracts). Breuil J, Moyroud J, Coulanges P (1983) Tentatives d’appreciation du retentissement socio-economique des schistosomiases a Madagascar [Attempts to evaluate the socioeconomic repercussions of schistosomiasis in Madagascar]. Archives de l’Institut Pasteur de Madagascar, 50:97–111. Brinkmann UK et al. (1988) Experiences with mass chemotherapy in the control of schistosomiasis in Mali. Tropical Medicine and Parasitology, 39:167–174. Buck AA et al. (1970) Health and disease in Chad: epidemiology, culture and environment in five villages. Baltimore, Johns Hopkins University Press. Bundy DAP et al. (1990) The epidemiological implications of a multiple-infection approach to the control of human helminth infections. Transactions of the Royal Society of Tropical Medicine and Hygiene, 85:274–276. Chen MG et al. (1993) Hepatitis B and schistosomiasis: interaction or no interaction. Tropical Diseases Bulletin, 90:R97–R115. Cheever AW et al. (1978) Schistosoma mansoni and S. hematobium infections in Egypt: III Extrahepatic pathology. American Journal of Tropical Medicine and Hygiene, 27:55–75. China Ministry of Public Health (1992) Abstracts. International Symposium on Schistosomiasis. November 24–26 1992, Beijing. pp. 1–12. Chowdhury AW, Levy BS (1988) Morbidity estimates of occupational illnesses and injuries in Kenya: human and economic costs. Paper presented at the Annual Schistosomiasis 387 Medical Scientific Conference, Kenya Medical Research Institute – Kenya Tropical Research Institute, Nairobi, 2 February (unpublished). De Cock KM et al. (1982) Esophageal varices in Nairobi, Kenya: a study of 68 cases. American Journal of Tropical Medicine and Hygiene, 31:579–588. Domingo EO et al. (1980) Morbidity in schistosomiasis japonica in relation to intensity of infection: study of a community in Leyte, Philippines. American Journal of Tropical Medicine and Hygiene, 29:858–867. Doumenge J-P et al. (1987) CEGET/WHO atlas of the global distribution of schistosomiasis. Presses Universitaires de Bordeaux. El-Alamy MA, Cline BL (1977) Prevalence and intensity of Schistosoma haematobium and S. mansoni infection in Qalyub, Egypt. American Journal of Tropical Medicine and Hygiene, 26:470–472. El-Bolkainy MN, Chu EW (1981) Detection of bladder cancer associated with schistosomiasis. Cairo, National Cancer Institute. El Malatawy A et al. (1992) Selective population chemotherapy among schoolchildren in Beheira governorate: the UNICEF/Arab Republic of Egypt/WHO Schistosomiasis Control Project. Bulletin of World Health Organization, 70: 47–566. Feldmeier H , Poggensee G (1993) Diagnostic techniques in schistosomiasis control: a review. Acta Tropica, 52:205–220. Forsyth DM (1969) A longitudinal study of endemic urinary schistosomiasis in a small East African Community. Bulletin of World Health Organization, 40:771–783. Foster R (1967) Schistosomiasis on an irrigated estate in East Africa: Ill-Effects of asymptomatic infection on health and industrial efficiency. Journal of Tropical Medicine and Hygiene, 70:185–195. Ghaffar YA et al. (1990) Hepatitis B vaccination in children infected with Schistosoma mansoni: correlation with ultrasonographic data. American Journal of Tropical Medicine and Hygiene, 43:516–519. Ghana Health Assessment Project Team (1981) A quantitative method of assessing the health impact of different diseases in less developed countries. International Journal of Epidemiology, 10:73–80. Granier H et al. (1985) Etude épidémiologique des bilharzioses intestinale et urinaire dans la région de Tala Mokolo (Monts Mandara-Nord Cameroun): [Epidemiologic study of intestinal and urinary bilharziasis in the Tala Mokolo region (Mandara Mountains, North Cameroon]. Médecine Tropicale, 45:27–33. Gryseels B (1988) The morbidity of schistosomiasis mansoni in the Rusizi Plain (Burundi). Transactions of the Royal Society of Tropical Medicine and Hygiene, 82:582–587. Gryseels B (1989) The relevance of schistosomiasis for public health. Tropical Medicine and Parasitology, 40:134–142. Halawani A, Tamami M (1955) Preliminary report on the cytological diagnosis and incidence of bilharzia cancer of the bladder in Egypt. Journal of the Egyptian Medical Association, 38:455–465. 388 Global Epidemiology of Infectious Diseases Haribhai HC et al. (1991) Spinal cord schistosomiasis: a clinical, laboratory and radiological study, with a note on therapeutic aspects. Brain, 114:709–726. Hayashi M (1979) Clinical studies on cerebral schistosomiasis japonica in the Philippines. Bulletin of the Tokyo Medical and Dental University, 26:287–297. Hiatt RA (1976) Morbidity from Schistosoma mansoni infections: an epidemiologic study based on quantitative analysis of egg excretion in two highland Ethiopian villages. American Journal of Tropical Medicine and Hygiene, 25:808–817. Hunter GW et al. (1952) Studies on schistosomiasis: VI – Control of the snail host of schistosomiasis in Japan with sodium pentachlorophenate (Santobrite). American Journal of Tropical Medicine and Hygiene, 1:831–847. Iarotski, LS, Davis A (1981) The schistosomiasis problem in the world: results of a WHO questionnaire survey. Bulletin of the World Health Organization, 59:115–127. Inaba Y et al. (1984) A case-control study on liver cancer with special emphasis on the possible aetiological role of schistosomiasis. International Journal of Epidemiology, 13:408–412. International Agency for Research on Cancer (1994) Schistosomes, liver flukes and helicobacter pylori. Lyon, International Agency for Research on Cancer (IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Volume 61). Jordan P, Webbe G, Sturrock RF, eds. (1993) Human Schistosomiasis. Wallingford, CAB International. Joubert J et al. (1990) Schistosomiasis of the spinal cord: underdiagnosed in South Africa? South African Medical Journal, 77:297–299. Junshi C et al. (1990) Diet, life-style and mortality in China: a study of the characteristics of 65 Chinese counties. Oxford, Oxford University Press. Kabatereine NB, Ariho C, Christensen NO (1992) Schistosoma mansoni in Pachweach, Nebbi District, Uganda, 40 years after Nelson. Tropical Medicine and Parasitology, 43:162–166. Kazura JW et al. (1985) Swamp rice farming: possible effects on endemicity of schistosomiasis mansoni and haematobia in a population in Liberia. American Journal of Tropical Medicine and Hygiene, 34:107–111. King CL et al. (1982) Prevalence and intensity of Schistosoma haematobium infection in six villages of Upper Egypt. American Journal of Tropical Medicine and Hygiene, 31:320–327. Kloetzel K (1962) Splenomegaly in schistosomiasis mansoni. American Journal of Tropical Medicine and Hygiene, 11:472–476. Koroltchouk V et al. (1987) Bladder cancer: approaches to prevention and control. Bulletin of the World Health Organization, 65:513–520. Latham MC et al. (1990) Metrifonate or praziquantel treatment improves physical fitness and appetite of Kenyan schoolboys with Schistosoma haematobium and hookworm infections. American Journal of Tropical Medicine and Hygiene, 43: 170–179. Lehman JS et al. (1976) The intensity and effects of infection with Schistosoma mansoni in a rural community in northeast Brazil. American Journal of Tropical Medicine and Hygiene, 25:285–294. Schistosomiasis 389 Lewert RM, Yogore MG Jr., Blas BL (1979). Schistosomiasis japonica in Barrio San Antonio, Basey, Samar, The Philippines. I. Epidemiology and morbidity. American Journal of Tropical Medicine and Hygiene, 28:1010–25. Lopes CAS (1984) Cancerizaçao vesical e schistosomiase [Bladder cancer and schistosomiasis]. Faculdade de Medicina da Universidade de Porto. Lucas SB (1982) Squamous cell carcinoma of the bladder and schistosomiasis. East African Medical Journal, 59:345–351. Mansour NS et al. (1981) A longitudinal study of Schistosoma haematobium infection in Qena governorate, Upper Egypt: 1-Initial epidemiological findings. American Journal of Tropical Medicine and Hygiene, 30:795–805. Martin-Prevel Y et al. (1992) An epidemiological study of a Schistosoma intercalatum focus in south-east Gabon. Transactions of the Royal Society of Tropical Medicine and Hygiene, 86:401–405. Morrow RH (1984) The application of a quantitative approach to the assessment of the relative importance of vector and soil transmitted diseases in Ghana. Social Science and Medicine, 19:1039–1049. Mott KE (1982) Epidemiological considerations for parasite vaccine development. Pontificiae Academiae Scientarium Scripta Varia, 47:5–23. Mott KE et al. (1985a) Evaluation of reagent strips in urine tests for detection of Schistosoma haematobium infection: a comparative study in Ghana and Zambia. Bulletin of the World Health Organization, 63:125–133. Mott KE et al. (1985b) Indirect screening for Schistosoma haematobium infection: a comparative study in Ghana and Zambia. Bulletin of the World Health Organization, 63:135–142. Murray CJL, Lopez AD (1996a) The global burden of disease. Global Burden of Disease and Injury Series, Volume I. Cambridge, Harvard University Press. Murray CJL, Lopez AD (1996b) Global health statistics: a compendium of incidence, prevalence and mortality estimates for over 200 conditions. Global Burden of Disease and Injury Series, Volume II. Cambridge, Harvard University Press. Ndamba J (1986) Schistosomiasis: its effects on the physical performance of school children in Zimbabwe. Central African Journal of Medicine, 32:289–293. Ndamba J et al. (1991) Morbidity due to Schistosoma mansoni among sugar-cane cutters in Zimbabwe. International Journal of Epidemiology, 20:787–795. Ndamba J et al. (1993) The physical fitness and work performance of agricultural workers infected with Schistosoma mansoni in Zimbabwe. Annals of Tropical Medicine and Parasitology, 87:553–561. Omer AHS et al. (1976) Infection with Schistosoma mansoni in the Gezira area of the Sudan. Journal of Tropical Medicine and Hygiene, 79:151–157. Ongom VL, Bradley DJ (1972) The epidemiology and consequences of Schistosoma mansoni infection in West Nile, Uganda: 1–Field studies of a community at Panyagoro. Transactions of the Royal Society of Tropical Medicine and Hygiene, 66:835–851. Prescott NM (1979) Schistosomiasis and development. World Development, 7:1–14. 390 Global Epidemiology of Infectious Diseases Ratard RC et al. (1992) Estimation of the number of cases of schistosomiasis in a country: the example of Cameroon. Transactions of the Royal Society of Tropical Medicine and Hygiene, 86:274–276. Robert CF, Bouvier S, Rougemont A (1989) Epidemiology of schistosomiasis in the riverine population of Lagdo Lake, northern Cameroon: mixed infections and ethnic factors. Tropical Medicine and Parasitiology, 40:153–158. Saad AM et al. (1991) Oesophageal varices in a region of the Sudan endemic for Schistosoma mansoni. British Journal of Surgery, 78:1252–1253. Saladin B et al. (1983) A pilot control trial of schistosomiasis in central Liberia by mass chemotherapy of target populations, combined with focal application of molluscicide. Acta Tropica, 40:271–295. Salih SY et al. (1977) Salmonellosis complicating schistosomiasis in the Sudan. Journal of Tropical Medicine and Hygiene, 80:14–18. Savioli L et al. (1990) Control of morbidity due to Schistosoma haematobium on Pemba Island: Egg excretion and haematuria as indicators of infection. American Journal of Tropical Medicine and Hygiene, 43:289–295. Simarro P, Sima FO, Mir M (1990) Urban epidemiology of Schistosoma intercalatum in the city of Bata, Equitorial Guinea. Tropical Medicine and Parasitology, 41:254–256. Siongok TKA et al. (1976) Morbidity in schistosomiasis mansoni in relation to intensity of infection: study of a community in Machakos, Kenya. American Journal of Tropical Medicine and Hygiene, 25:273–284. Sleigh AC and Mott KE (1986) Schistosomiasis. Clinics in tropical medicine and communicable diseases. London, Saunders, 1:643–670. Sleigh A et al. (1982) Comparison of filtration staining (Bell) and thick smear (Kato) for the detection and quantitation of Schistosoma mansoni eggs in faeces. Transactions of the Royal Society of Tropical Medicine and Hygiene, 76:403–406. Smith DH, Warren KS, Mahmoud AAF (1979) Morbidity in schistosomiasis mansoni in relation to intensity of infection: study of a community in Kisumu, Kenya. American Journal of Tropical Medicine and Hygiene, 28:220–229. Stephenson LS et al. (1984) Sensitivity and specificity of reagent strips in screening of Kenyan children for Schistosoma haematobium infection. American Journal of Tropical Medicine and Hygiene, 33:862–871. Tanaka H et al. (1984) Epidemiology and transmission dynamics: evaluation of control measures by means of annual incidence of infection with Schistosoma japonicum among school children at Dagami, Leyte, Philippines. Southeast Asian Journal of Tropical Medicine and Public Health, 15:480–481. Tanaka H et al. (1985) Longitudinal study on Schistosoma japonicum infections in the Philippines: 4 – Effect of mass-chemotherapy with praziquantel in incidence at Dagami, Leyte. Japanese Journal of Experimental Medicine, 55:161–165. Taylor P, Makura O (1985) Prevalence and distribution of schistosomiasis in Zimbabwe. Annals of Tropical Medicine and Parasitology, 79:287–299. Teesdale CH, Fahringer K, Chitsulo L (1985) Egg count variability and sensitivity of a thin smear technique for the diagnosis of Schistosoma mansoni. Transactions of the Royal Society of Tropical Medicine and Hygiene, 79:369–373. Schistosomiasis 391 Thiongo FW, Ouma J (1987) Prevalence of schistosomes and other parasites in Taito Division of Taita-Taveta District. East African Medical Journal, 64:665–671. Utroska JA et al. (1989) An estimate of global needs for praziquantel within schistosomiasis control programmes. Geneva, World Health Organization (WHO/ SCHISTO/89.102). Van Ee JH, Polderman AM (1984) Physiological performance and work capacity of tin mine labourers infested with schistosomiasis in Zaire. Tropical and Geographical Medicine, 36:259–266. Warren KS et al. (1978) Quantification of infection with Schistosoma haematobium in relation to epidemiology and selective population chemotherapy. 1 – Minimal number of daily egg counts in urine necessary to establish intensity of infection. Journal of Infectious Diseases, 138:849–855. Wenlock RW. The epidemiology of tropical parasitic diseases in rural Zambia and the consequences for public health. Journal of Tropical Medicine & Hygiene, 82: 90–98. White PT, Coleman M, Jupp BP (1982) Swamp rice development, schistosomiasis and onchocerciasis in Southeast Sierra Leone. American Journal of Tropical Medicine and Hygiene, 31:490–498. Winslow, CEA (1951) The cost of sickness and the price of health. Geneva, World Health Organization (WHO Monograph Series, N. 7). World Health Organization (1967) World Health Organization Scientific Group. Measurement of the public health importance of bilharziasis. Geneva, World Health Organization (WHO Technical Report Series No. 349). World Health Organization Regional Office for the Eastern Mediterranean (1983) Technical guide for the diagnosis, prevention and control of Rift Valley Fever in man and animals. WHO/EMRO Technical Publication No. 8. World Health Organization (1985) WHO Expert Committee on Control of Schistosomiasis. The control of schistosomiasis. Geneva, World Health Organization (WHO Technical Report Series, No. 728). World Health Organization (1990) World Health Organization model prescribing information: drugs used in parasitic diseases. Geneva, World Health Organization. World Health Organization (1993) WHO Expert Committee on Control of Schistosomiasis. The control of schistosomiasis. Second report. Geneva, World Health Organization, (WHO Technical Report Series, No. 830). World Health Organization (1989) Progress in assessment of morbidity due to schistosomiasis: reviews of recent literature. Schistosoma haematobium, Schistosoma intercalatum, Schistosoma japonicum, Schistosoma mansoni. Reprints from the Tropical Diseases Bulletin 85 Nos. 6 & 10 (1988); 86 Nos. 4 & 8 (1989). Wright WH, Dobrovolny CG (1953) Experiments in the control of schistosomiasis in Brazil. Public Health Reports, 68:1156–1160.
