Download integrated community case management of malaria and pneumonia

From Department of Public Health Sciences (Global Health), Karolinska
Institutet, Stockholm, Sweden
and
the College of Health Sciences, Makerere University, Kampala, Uganda
INTEGRATED COMMUNITY CASE
MANAGEMENT OF MALARIA AND PNEUMONIA
IN EASTERN UGANDA
CARE-SEEKING, ADHERENCE, AND COMMUNITY HEALTH WORKER
PERFORMANCE
Joan Nakayaga Kalyango
MAKERERE UNIVERSITY
Kampala and Stockholm 2013
All previously published papers were reproduced with permission from the publishers.
Published by Karolinska Institutet and Makerere University.
Printed by Universitetsservice-AB
Nanna Svartz vag 4,
SE-17177, Stockholm, Sweden
© Joan Nakayaga Kalyango, 2013
ISBN 978-91-7549-303-9
ii
ABSTRACT
Background: Despite being easily preventable and treatable, malaria and pneumonia are major
killers of children aged less than five years. Integrated community based interventions through which
lay persons called community health workers (CHWs) can manage malaria, pneumonia, diarrhea and
neonatal conditions are recommended by WHO and UNICEF. However, there is limited information
on care-seeking and performance of CHWs in the context of integrated illness management.
Main aim: To assess care-seeking and quality of care in integrated community case management of
malaria and pneumonia in children aged less than five years in Uganda so as to inform the
implementation of integrated community case management of childhood illness strategy (ICCM).
Methods: Four studies (I-IV) were nested in a cluster randomized trial in Iganga-Mayuge
demographic surveillance site in eastern Uganda. In this trial CHWs treated malaria and pneumonia
(intervention arm) or malaria alone (control arm) in children aged 4-59 months. Performance of
CHWs (I) was assessed using: questionnaires (with knowledge tests, case scenarios) and record
reviews for 125 CHWs; observations among 57 CHWs in the intervention arm; and four focus group
discussions with CHWs. Adherence to treatment was assessed using pill counts and caregiver reports
among 1256 children treated by CHWs (II). Receipt of prompt and appropriate antibiotics for
pneumonia symptoms and treatment outcomes were assessed among 1276 children treated by CHWs
(III). Care-seeking and management of malaria and pneumonia were assessed among 1095 children
and from 13 key informant interviews (IV).
Results: Care-seeking from CHWs was higher in the intervention than the control arm (31% vs 22%,
p=0.01) (IV). CHWs’ performance on malaria symptoms was similar in the intervention and control
arms on: overall knowledge, eliciting signs and symptoms, and prescribing (I). More children treated
by CHWs received prompt and appropriate malaria treatment compared to other health providers
(37% vs 9%, p<0.001) (IV). CHWs had high scores in prescribing for pneumonia but had lower:
overall knowledge of pneumonia (40%), and scores on eliciting pneumonia signs and symptoms
(25%). Only 35% of CHWs counted respiratory rates within two breaths of rates counted by the
physician, and 12% of children without fast breathing received antibiotics while 82% with fast
breathing received antibiotics (I). Children treated by CHWs in the intervention arm were more likely
to receive prompt and appropriate antibiotics for pneumonia symptoms compared to the control arm
(RR=3.51, 95% CI = 1.75-7.03) (III). There was also a higher reduction in the proportion of children
with fast breathing from day one to day four in the intervention compared to the control arm (9.2% vs
4.2%, p=0.01); and a lower proportion of febrile children on day four (1% vs 4%; RR=0.29, 95% CI
= 0.11-0.78) (III). Adherence to combined antimalarials and antibiotics was similar to adherence to
antimalarials alone in the intervention arm (mean 99% both groups) (II).
Conclusions: CHWs’ performance on malaria was not affected by additional roles of pneumonia
management, but they had challenges in assessment of pneumonia symptoms. CHWs should be
supported with continued training, adequate supervision and provision of drugs, diagnostics and other
supplies.
Key words: malaria, pneumonia, community health worker, community case management, Uganda
iii
LIST OF PUBLICATIONS
I.
Joan N. Kalyango, Elizeus Rutebemberwa, Tobias Alfven, Sarah Ssali,
Stefan Peterson, Charles Karamagi, 2012, Performance of community
health workers under integrated community case management of
childhood illnesses in Eastern Uganda. Malaria Journal, 11, 282.
II.
Joan N. Kalyango, Elizeus Rutebemberwa, Charles Karamagi, Edison
Mworozi, Sarah Ssali, Tobias Alfven, Stefan Peterson, 2013, High
adherence to antimalarials and antibiotics under integrated
community case management of childhood illness in IgangaMayuge DSS, Uganda. PLoS One, 8(3), e60481.
III.
Joan N Kalyango, Tobias Alfven, Stefan Peterson, Kevin Mugenyi,
Charles Karamagi, Elizeus Rutebemberwa, 2013: Integrated
community case management of malaria and pneumonia increases
prompt and appropriate treatment for pneumonia symptoms in
children under five years in Eastern Uganda. Malaria Journal, 12,
340.
IV.
Kalyango JN, Lindstrand A, Rutebemberwa E, Ssali S, Kadobera D,
Karamagi C, Peterson S, Alfven T, 2012, Increased use of community
medicines distributors and rational use of drugs in children less
than five years of age in Uganda caused by integrated community
case management of fever. American Journal of Tropical Medicine
and Hygiene, 87, 36-45.
The papers will be referred to by their roman numerals I-IV.
iv
CONTENTS
1
2
3
4
5
INTRODUCTION ...................................................................................1
1.1 Global health.......................................................................................1
1.2 Universal health coverage ..................................................................1
1.3 Health systems ....................................................................................1
1.4 Quality of health care..........................................................................3
1.5 Drug use ..............................................................................................3
1.6 Drug resistance....................................................................................4
1.7 Child mortality....................................................................................5
1.8 Causes of child deaths ........................................................................5
1.8.1 Malaria ....................................................................................6
1.8.2 Pneumonia ..............................................................................6
1.9 Care-seeking for childhood illness .....................................................7
1.10 Global targets of child mortality.......................................................8
1.10.1 Progress towards MDG4 ........................................................8
1.10.2 Global strategies to accelerate achievement of MDG4 .........8
1.11 Strategies to address child mortality.................................................8
1.11.1 Primary health care.................................................................9
1.11.2 Child survival and development revolution...........................9
1.11.3 Integrated management of childhood illnesses......................9
1.12 Community case management..........................................................9
1.12.1 Community health workers ..................................................10
1.12.2 Home management of malaria .............................................10
1.12.3 Integrated community case management of common
childhood illnesses............................................................................11
1.13 Uganda country profile ...................................................................11
1.13.1 Uganda health indicators......................................................12
1.13.2 Uganda health system...........................................................12
1.13.3 The role of the private sector................................................13
1.13.4 Community health workers in Uganda ................................13
RATIONALE FOR THE STUDIES.....................................................16
CONCEPTUAL FRAMEWORK .........................................................17
AIM AND OBJECTIVES.....................................................................19
4.1 General aim .......................................................................................19
4.1.1 Specific objectives................................................................19
METHODS ............................................................................................20
5.1 Study area and population ................................................................20
5.2 Description of the cluster randomized trial......................................21
5.2.1 Selection of CHWs...............................................................21
5.2.2 Training of CHWs ................................................................21
5.2.3 Randomization......................................................................22
5.2.4 Supervision of CHWs...........................................................22
5.2.5 Patient management in the cRCT ........................................22
5.3 Study design and data collection methods .......................................24
5.3.1 Study design .........................................................................24
5.3.2 Overview of the data collection methods ............................24
v
5.3.3 Sampling, sample size and data collection methods............26
5.3.4 Data management .................................................................29
5.3.5 Data analysis .........................................................................30
5.3.6 Summary of methods............................................................31
5.3.7 Ethical Issues ........................................................................32
6 RESULTS.............................................................................................. 33
6.1 Utilization of services (study IV) .....................................................33
6.2 Quality of care...................................................................................34
6.3 Health status (study III) ....................................................................39
6.4 Patient satisfaction (studies I, III and IV).........................................40
6.5 Equity (study IV)...............................................................................40
6.6 Influence of livelihood assets on utilization of services
and quality of care.......................................................................................41
6.7 Other factors associated with utilization of CHWs
and quality of care.......................................................................................42
7 DISCUSSION ....................................................................................... 43
7.1 Main findings ....................................................................................43
7.2 Considerations for the health systems ..............................................49
7.3 Methodological considerations.........................................................50
8 Conclusions and policy implications.................................................... 54
9 Acknowledgements............................................................................... 55
10 References ............................................................................................. 58
vi
LIST OF ABBREVIATIONS
ACTs
AL
ARI
CCM
CDD
CHWs
CI
C-IMCI
CMDs
CQ
cRCT
DHS
FGDs
GAVI
GIVS
GOBI
HBMF
HDSS
HMM
HSSIP
ICCM
IMCI
ITNs
KIIs
KIs
LMIC
MDGs
MoH
n.d
OR
ORS
PCA
PFP
PNFP
PPPH
RBM
RDT
RR
SP
U5MR
UN
UNICEF
VHT
WHO
Artemisinin-based Combination Therapies
Artemether-Lumefantrine
Acute Respiratory Infection
Community Case Management
Community Drug Distributor
Community Health Workers
Confidence Interval
Community-Integrated Management of Childhood Illnesses
Community Medicine Distributors
Chloroquine
Cluster Randomized Controlled Trial
Demographic Health Surveys
Focus Group Discussions
Global Alliance for Vaccines and Immunization
Global Immunization Vision and Strategy
Growth monitoring, Oral rehydration, Breastfeeding and
Immunization
Home Based Management of Fever
Health and Demographic Surveillance Site
Home Management of Malaria
Health Sector Strategic and Investment Plan
Integrated Community Case Management
Integrated Management of Childhood Illnesses
Insecticide Treated Nets
Key Informant Interviews
Key Informants
Low and Middle Income Countries
Millennium Development Goals
Ministry of Health
no date
Odds Ratio
Oral Rehydration Salts
Principal Components Analysis
Private For Profit
Private Not For Profit
Public-Private Partnerships in Health
Roll Back Malaria
Rapid Diagnostic Test
Relative Risk
Sulphadoxine-Pyrimethamine
Under Five Mortality Rate
United Nations
United Nations Children’s Fund
Village Health Team
World Health Organization
vii
OPERATIONAL DEFINITIONS
Adherence to medicines: The extent to which a person’s behavior in taking medicines corresponds
with agreed recommendations from a health care provider (WHO, 2003; Osterberg and Blaschke,
2005).
Level of adherence: The percentage of prescribed medicines taken correctly over a threeday period based on caregiver reports and pill counts.
Non-adherence to medicines: Missing some of the prescribed medicines or not taking the
medicines appropriately (i.e. had less than 100% level of adherence) (Kachur et al., 2004). A
further classification of non-adherence where patients who had taken all medicines as
prescribed but who vomited within 30 minutes and did not have a replacement dose
administered was made, as recommended by a previous study (Beer et al., 2009).
Optimal adherence: This considered how antimalarials were taken and what they were
taken with. Optimal adherence was taking all the medicines as prescribed and with a fatty
meal as recommended for artemether-lumefantrine by the manufacturer; good adherence was
taking all the antimalarials as prescribed but the medicines were either not taken consistently
with a fatty meal or the medicines were not taken with a fatty meal at all; and non-adherence
was defined as missing some of the medicines or not taking medicines in the correct schedule
regardless of what they were taken with (Achan et al., 2009).
Appropriate antibiotic: Receiving a recommended antibiotic for pneumonia according to national
(Ministry of Health (Uganda), 2010b) or community health workers’ guidelines (Department of
Health Policy Planning and Management, 2010), or the British National Formularly (British Medical
Association and Royal Pharmaceutical Society of Great Britain, 2012), a widely used reference in
Uganda.
Appropriate treatment: Receiving the recommended drug, in the recommended dose, and for the
recommended frequency, and duration (Nsungwa-Sabiiti et al., 2005).
Community Health Workers (CHWs): These are persons selected from the communities in which
they live and work; they are selected by and are answerable to the communities, are supported by the
health system, and undergo shorter training than professional health workers (WHO, 2007a). They
have also been referred to as community drug distributors, community medicine distributors, village
health team, and village malaria workers, among others.
CHW performance: Ability to elicit signs and symptoms and classify illness, identify and respond
to danger signs, prescribe medicines (dosing, medicines administration instructions), and store
medicines appropriately.
Fever: Febrile to touch or history of ‘hot body’ in the last 24 hours as reported by the caregiver, or
axillary temperature 37.5oC.
Malaria: Fever or history of fever in the last 24 hours in a high-risk malaria area (WHO and
UNICEF, 2008).
viii
Pneumonia: Acute infections of the lungs caused by viruses, bacteria or fungi presenting as cough
and/or difficult breathing, with or without fever, with either fast breathing or lower chest in-drawing
or noisy breathing in a calm child (WHO, 2013e; WHO and UNICEF, 2008).
CHW-classification of pneumonia: Non-severe pneumonia was the presence of cough or
difficult breathing and fast breathing similar to definition used in IMCI guidelines (WHO and
UNICEF, 2008).
Pneumonia symptoms: Caregiver reports of cough accompanied by fast and/or difficult
breathing similar to definition used in multiple indicator cluster surveys and demographic
health surveys (UNICEF, 2013b).
Prompt treatment: Receiving treatment within 24 hours from onset of symptoms (Roll Back
Malaria / World Health Organization, 2003). In order to overcome the difficulty of counting 24
hours, this definition was modified to receiving treatment on the day of onset of symptoms or the
next day if the onset of illness was at night, as has been used in other studies (Ajayi et al., 2008b;
Rutebemberwa et al., 2009a).
Quality of care: The extent to which health services for individuals and populations increase the
likelihood of desired health outcomes and are consistent with current professional knowledge
(Institute of Medicine, 1990). In this thesis, quality of care focused on the caregivers’, community
health workers’ and programme performance.
Rapid/fast breathing: Having 50 breaths per minute in children aged 4-12 months and 40 breaths
per minute in children aged 12-59 months (WHO and UNICEF, 2008).
Rational use of drugs: Patients receive the appropriate medicine, in the proper dose, for an adequate
period of time, and at the lowest cost to them and their community (WHO, 1985).
Recommended drug: Medicines recommended for illness based on national (Ministry of Health
(Uganda), 2010b) and community health workers’ treatment guidelines (Department of Health Policy
Planning and Management, 2010) as well as treatment recommendations of the British National
Formulary (Joint Formulary Committee, 2007), a widely used reference in Uganda.
ix
PREAMBLE
Every minute, six children below the age of five years in sub Saharan Africa die, mainly from
preventable causes (WHO, 2012d). These are not mere statistics but children with faces and names,
and for whom their families have great dreams. Every minute, six families in this region start on the
painful journey of losing a child, pain which could have been prevented. The children are also robbed
of the opportunity to contribute to the progress of their families, communities, and nations.
Commitment of nations to the Millennium Development Goals has reduced under-five mortality but
the reduction is still insufficient to reach the set targets of a two-thirds reduction from the rate in 1990
by 2015. The countries that need to make the greatest reductions in mortality have not done so
(WHO, 2012f). Many of the deaths could be prevented through full implementation of a few
effective interventions, but coverage of these interventions is still low (Bryce et al., 2003; BoschiPinto et al., 2009).
Efforts to improve access to effective interventions for children initially targeted health facilities, but
have recently expanded to involve communities in their health care. This two pronged approach is
necessary because many children are either not taken to health facilities or are taken late resulting in
increased deaths. Community interventions such as home management of malaria (HMM),
community management of pneumonia, and more recently the integrated community case
management of childhood illnesses (ICCM) are major steps towards improvement of access to health
care in resource limited settings. ICCM was deemed necessary because many childhood diseases
have overlapping symptoms and sometimes present together (English et al., 1996; Källander et al.,
2004). ICCM implies more roles for the community health workers (CHWs), the first-level health
workers with no formal medical training.
While this strategy seems ideal in solving the problems of poor access to health care for common
childhood illnesses, a number of questions arise: (1) Will having CHWs that can treat multiple
illnesses really improve access to effective care? Will it then lead to reduced mortality? (2) Can
CHWs handle the complexity of multiple illnesses? Can they make proper diagnoses? Can they treat
the children with appropriate medicines or refer them as required? Can they be good stewards of the
antimicrobials that they are entrusted with, so that they are not misused? Is there a specific type of
person that can perform as a CHW more effectively than another? How will the motivation and
retention of CHWs (who are usually not paid salaries) be maintained? (3) Will the establishment of
the CHW cadre of health worker be another self-defeating effort where drugs, diagnostics and other
essential supplies will be out of stock, just as with formal health services? (4) What about the
caregivers? Will they have sufficient trust in the care provided by the CHWs and seek care from
them? Will they give the medicines as instructed? Will they comply with referral?
This thesis attempts to address some of these questions, and builds on previous work (Källander,
2006; Rutebemberwa, 2009; Hildenwall, 2009; Mukanga, 2012; Nsungwa-Sabiiti, 2009; YeboahAntwi et al., 2010; Degefie et al., 2009; Kelly et al., 2001); in an attempt to contribute to the health of
children.
x
1 INTRODUCTION
1.1 GLOBAL HEALTH
Despite improvements in health over the last century that have seen people living longer and
healthier lives, millions of people still die prematurely of diseases that are preventable or curable.
Many of the deaths could be prevented at relatively little more expense (Carr, 2004). However, about
one billion people lack access to health care, more so the poor for whom health services are often out
of reach (Action for Global Health, 2010). A considerable number of people die from infectious
diseases especially in resource-poor settings where access to care is inadequate. Infectious diseases
caused about 34% of the deaths in Africa in 2011 compared with 2.6% in Europe (WHO, 2012b).
More than two million people, mostly in low and middle income (LMIC) countries, die annually
from diseases which are preventable by vaccines recommended by the World Health Organization
(WHO) (WHO, 2006b). An even greater number (about 36 million) die from non-communicable
diseases (WHO, 2012i). These illnesses take their toll on the life expectancy of nations and great
differences are seen between nations where people have better access to health care compared with
those where access to health care is poor. In 2012, the life expectancy at birth in high-income
countries was about 80 years compared to 60 years in low-income countries (WHO, 2012e). Life
expectancy at birth is strongly dependent on infant and child mortality and therefore differences in
life expectancy may reflect differences in infant and child mortality in these regions. All human
beings have a fundamental right to enjoy the highest attainable standard of health and it should
therefore be promoted and protected. Poor health also has a negative impact on economic and social
development.
1.2 UNIVERSAL HEALTH COVERAGE
A key aspect that is necessary to the promotion and protection of health is timely access to health
services, which needs a well-functioning health financing system in order to ensure that the services
which people need exist, and that they can afford them when they need them (WHO, 2012h). In 2005
member states of the WHO committed to development of financing systems to ensure that all people
have access to health services and do not suffer financial hardships due to payment for health
services, a goal which was termed as “universal health coverage”. Universal health coverage is
therefore defined as ensuring that all people have access to needed promotive, preventive, curative
and rehabilitative health services, of sufficient quality to be effective, while also ensuring that people
do not suffer financial hardship when paying for these services (WHO, 2005c).
Universal health coverage has become a major goal for health reform in many countries and a
priority objective of the WHO (Rodin and de Ferranti, 2012; Holmes, 2012). However, the world is
still a long way from universal health coverage, more especially in resource-limited settings. Service
coverage is still low, with inequitable distribution of the available services. For example, global
measles immunization coverage was 85% in 2010, with low-income countries having coverage of
78%. Children from poor families have a higher risk of death than those from richer families. In
addition, up to 11% of people suffer financial hardships and about 5% of the population is forced into
poverty due to payment for health services (WHO, 2012i).
1.3 HEALTH SYSTEMS
Progressing towards universal health coverage requires strong and efficient health systems (WHO,
2012g). WHO has defined health systems as consisting of all organizations, people and actions
whose primary aim is to promote, restore or maintain health. The goals of the health system are to
1
improve the health and health equity of persons that are served, in ways that are responsive to
people’s expectations, financially fair, and make the best and most efficient use of available resources
(WHO, 2007b).
In order to achieve these goals, six interconnected pillars, referred to as building blocks, have been
identified, including: service delivery from well-maintained health facilities; well trained and
adequately paid health work force; reliable health information on which to base decisions and
policies; medical technologies, vaccines and drugs adequate to deliver quality health care; robust
health financing mechanisms; and leadership and governance (Figure 1) (de Savigny and Adam,
2009). The people participate in the health systems as beneficiaries and as actors in driving the health
system.
Figure 1. Building blocks of a health system
Source: (de Savigny and Adam, 2009).
Health systems especially those in resource-limited settings, are faced with several challenges. There
are reports of large gaps in coverage of crucial components of health services, low-quality services
that do not contribute optimally to improved health outcomes, and inequity in health outcomes. In
addition, many countries lack the human resources necessary to deliver essential health interventions.
Extreme shortages of health workers exist in 57 countries, 36 of which are in Africa. These shortages
are due to limited production capacity, migration of health workers within and across countries, a
poor mix of skills, and demographic imbalances (WHO, 2007a). In response, countries have turned to
shifting of some basic health care tasks to low-level or informal health workers providing
community-based interventions.
Funds to deliver health care are also still insufficient. Many people cannot access the health services
because they have to pay for them at the time of need, and a large number of those that are able to
access the services become impoverished. Each year an estimated 100 million people suffer financial
catastrophes because of payment for health care (WHO, 2012h). The financing for health in resourcelimited settings comes mainly from donor aid with little or no health insurance.
As a result of poor financial systems and poor leadership and governance, there is lack of essential
medicines and technologies to deliver the necessary health care. A survey done in 39 countries found
low availability of essential medicines, and this varied in the public (20%) and the private sector
2
(56%). There is also misuse of available resources (WHO, 2007b) and planning and resource
allocation are not guided by appropriate health information, because many countries lack up-to-date
information for these functions or the available information is not used appropriately. In over 60
countries, less than a quarter of deaths are recorded by vital registration systems (WHO, 2007a).
Given the prevailing situation, the health stewards need to find innovative ways of health care
provision. A number of interventions have been set up to target specific conditions or age groups.
However, the success of these interventions needs to be looked at in light of existing health systems.
In addition, strengthening of health systems should be done in light of the context in which they exist
(Commission on social determinants of health, 2008).
1.4 QUALITY OF HEALTH CARE
Quality of care is one of the intermediate outcomes of health systems. The inadequacies in many
health systems have led to poor quality in health care. Quality of care has been defined as the extent
to which health services for individuals and populations increase the likelihood of desired health
outcomes and are consistent with current professional knowledge (Institute of Medicine, 1990). The
care should be: effective (adherent to an evidence base and results in improved health outcomes for
individuals and communities); efficient (maximizes resource use and avoids waste); accessible
(timely, geographically reasonable, and provided in a setting where skills and resources are
appropriate to medical need); acceptable/patient-centred (takes into account the preferences and
aspirations of individual service users and the cultures of the community); equitable (does not vary in
quality because of personal characteristics such as gender, race, ethnicity, geographical location or
socio-economic status); and safe (minimizes risks and harm to service users). Although new
interventions aimed at improving health will usually have high quality of care as an intermediate
objective, it is not always certain to what extent this is achieved.
1.5 DRUG USE
Drugs are a major part of health care, with most diseases requiring drug treatment. It is important that
the drugs are used appropriately in order to achieve adequate treatment while minimizing the
development of resistance, and avoiding wastage of scarce resources as well as potential health
hazards. The need for appropriate drug use has led to focus on the rational drug use concept. Rational
use of drugs is defined as patients receiving the appropriate medicines, in doses that meet their own
individual requirements, for an adequate period of time and at the lowest cost to both them and the
community (WHO, 2002; WHO, 1985). Despite various efforts made to promote rational use of
drugs, drug misuse is still a major problem worldwide. The problem is worse in the private compared
with the public sector and in LMIC compared to the high-income countries (WHO, 2011c).
In many LMIC less than 50% of patients are treated according to standard treatment guidelines
(WHO, 2009), and more than 80% of prescriptions are filled by unqualified health providers with
poor dispensing practices (WHO, 2011c). In addition, the problem of self-medication often with
inappropriate medicines is common (Geissler et al., 2000; Shankar et al., 2002; Awad and Eltayeb,
2007; Mbonye et al., 2008). There is misuse of medicines including antibiotics (Isturiz and Claude,
2000; Chandy et al., 2012; Byarugaba, 2004), and yet a large proportion of the people that need
medicines do not receive them (WHO, 2011c; Pecoul et al., 1999; Kristiansson et al., 2009). Coupled
with this is the problem of non-adherence; about half of patients do not take medicines as prescribed
(WHO, 2011c).
Irrational use of medicines for malaria management has been reported in some studies with children
receiving wrong drugs, wrong dosages or medicines for the wrong duration (Abuya et al., 2007;
3
Kemble et al., 2006; Oreagba et al., 2007; Osterholt et al., 2006; Nshakira et al., 2002; Baume et al.,
2000; Nsungwa-Sabiiti et al., 2007; Nsungwa-Sabiiti et al., 2005). The proportion of children that
receive prompt, effective treatment is much lower than the 80% target (Roll Back Malaria, 2005).
Factors that influence drug use include: socio-demographic characteristics such as age, sex, education
level, and mother’s occupation; illness perceptions; financial constraints leading to lack of access to
medicines; inadequate knowledge of correct dosage; time of consultation; illness presentation; and
in-service training of health workers coupled with availability of wall charts and guidelines in health
facilities as well as more frequent support supervision (Chibwana et al., 2009; Osterholt et al., 2006;
Santos et al., 2009; Baume et al., 2000; Nsungwa-Sabiiti et al., 2007; Sirima et al., 2003; Zurovac et
al., 2004).
There have also been several reports of inappropriate management of respiratory infections in
children. Some children receive antibiotics for respiratory illnesses in which antibiotics are not
required while others do not receive antibiotics when they should. In LMIC less than one-third of
children receive antibiotics for suspected pneumonia (WHO, 2013a; UNICEF, 2012). In addition,
wrong doses, and drugs given in the wrong frequency and for the wrong duration, have been
observed (Hui et al., 1997; Kristiansson et al., 2009; Del Fiol et al., 2012; Chuc et al., 2001; Bojalil et
al., 1998; Gwimile et al., 2012). Inappropriate use of antimicrobials like antibiotics and antimalarials
may lead to poor treatment outcomes and increase the risk of drug resistance.
1.6 DRUG RESISTANCE
Although increased access to medicines is expected to improve treatment outcomes, the increased use
of antimicrobials also increases the risk of drug resistance due to increase in drug pressure, which
leads to selection of resistant strains (McGarock, 2002; Ilic et al., 2012; Cars and Nordberg, 2005;
Laxminarayan et al., 2006). The problem is further worsened by misuse of drugs with inadequate
dosages and non-adherence which may lead to inadequate drug exposure and therefore increase the
risk of resistance (Olofsson and Cars, 2007; Cars and Nordberg, 2005; Yeung and White, 2005).
Antimalarials have previously been hit with resistance (White, 1999). In 2000, reports of widespread
resistance to chloroquine (CQ), the then first-line drug for uncomplicated malaria, led to its
combination with sulphadoxine-pyrimethamine (SP) (Nanyunja et al., 2011). However, because there
was already some resistance to SP, reports of clinical failures to the new combination soon became
apparent. WHO recommended artemisinin-based combination therapies (ACTs) in 2001 (WHO,
2001) and many countries adopted this policy (WHO, 2012c). Uganda chose a combination of
artemether-lumefantrine (AL) in 2005 (Ministry of Health (Uganda), 2005).
There is a need to protect these drugs from resistance because of the limitations in options for malaria
treatment. Quinine, the alternative to ACTs has been associated with several adverse effects, and
these noxious outcomes together with the long duration of administration make it less favourable to
patients (Baird, 2005). Studies have shown varying levels of adherence to antimalarials (39-97%)
with some very low levels likely to increase the risk of resistance (Onyango et al., 2012; Kachur et
al., 2004; Ajayi et al., 2008a; Lawford et al., 2011; Lemma et al., 2011; Yeung and White, 2005).
There are reports indicating that resistance to ACTs is emerging in south-east Asia and efforts to
tackle the problem are under way through the global plan for artemisinin resistance containment
(Breman, 2012).
The first-line treatment for non-severe pneumonia in Uganda and a few other resource-limited
countries has been cotrimoxazole (Ministry of Health (Uganda), 2010b; Grant et al., 2009).
Cotrimoxazole suffers from gross misuse through self-medication (Enato and Uwaga, 2011; Sturm et
4
al., 1997; Oshikoya et al., 2007), and inappropriate prescribing and dispensing (Trap and Hansen,
2002). In addition, it is widely used for prophylaxis against opportunistic infections in HIV following
recommendations by WHO and other bodies (WHO et al., 2005). It is thus not surprising that wide
spread resistance to cotrimoxazole has been reported (Joloba et al., 2001; Hoa et al., 2010).
Recommendations regarding change of the first-line treatment for non-severe pneumonia to
amoxicillin were made following the changing sensitivity patterns to cotrimoxazole (Grant et al.,
2009). The effectiveness of these medicines should be protected in order to avoid the need to move to
more expensive medicines which may also have more adverse effects.
1.7 CHILD MORTALITY
The number of children worldwide that died before their fifth birthday was estimated at 6.6 million in
2012 (UNICEF, 2013a). Although this mortality is a considerable reduction from the 12.6 million
deaths in 1990, it is still high and some areas have had much slower reductions in under-five
mortality than others. Sub-Saharan Africa and south Asia have registered the lowest reductions in
under-five mortality rates (U5MR) since 1990 at 45% and 54% respectively, and these regions bear
the greatest proportion of under-five deaths. In 2012 these regions contributed about 82% of the
global under-five mortality (UNICEF, 2013a). In addition, inequalities in mortality exist even within
regions. The under-five mortality is higher in poor households, rural areas, and where the mother has
not had basic education. Children born in the poorest households have twice the risk of death as those
from the wealthiest households (United Nations, 2011).
1.8 CAUSES OF CHILD DEATHS
Worldwide, a few conditions account for a large proportion of deaths. In 2012 the five major causes
of death were: pneumonia (17%), pre-term birth complications (15%), intrapartum-related
complications (10%), diarrhoea (9%), and malaria (7%), with malnutrition a major underlying factor
contributing to 45% of all under-five deaths (UN Inter-agency group for child mortality estimation,
2013; UNICEF, 2013a). Many of these causes of death could be prevented or treated with low-cost
interventions. Although some reductions have been noted, mostly with the infectious causes of death,
the conditions still cause considerable mortality.
In sub-Saharan Africa the major causes of death in 2012 were: neonatal conditions (30%),
pneumonia (17%), malaria (14%), and diarrhoea (11%) (Figure 2) (UNICEF, 2013a). This thesis
addresses pneumonia and malaria, some of the leading causes of mortality in sub-Saharan Africa.
5
Pneumonia,
17%
Others, 21%
Measles , 1%
Malaria, 14%
AIDS, 3%
Injuries , 4%
Diarrhoea,
11%
Neonatal
causes, 30%
Figure 2. Causes of child mortality in sub-Saharan Africa
Source: (UNICEF, 2013a)
1.8.1 Malaria
Malaria is a disease caused by the plasmodium parasite which is transmitted through the bites of
infected female Anopheles mosquitoes. There are four types of plasmodium species that cause
malaria in humans including: Plasmodium falciparum, P. vivax, P. ovale and P. malariae. P.
falciparum malaria is the most severe, causing most of the deaths (WHO, 2013c). Signs and
symptoms of malaria include: fever, headache, and vomiting. Children with severe malaria may
present with severe anaemia, respiratory distress in response to metabolic acidosis, or repeated
convulsions which may be a sign of cerebral malaria.
Key interventions to control malaria include prevention through the use of insecticide-treated nets
(ITNs) and indoor residual spraying, and early diagnosis and treatment. The recommended first-line
treatment for uncomplicated malaria is currently ACTs (WHO, 2010a). Since 2010 WHO
recommends that all cases of suspected malaria are confirmed by parasite-based diagnosis using
either microscopy or rapid diagnostic tests (RDTs), and symptom-based diagnosis should only be
used where the parasite-based diagnosis is inaccessible (WHO, 2010a). Prior to that, presentation
with fever or a history of fever in malaria-endemic areas was presumed to be malaria and therefore
treated with antimalarials (WHO and UNICEF, 2008).
1.8.2 Pneumonia
Pneumonia is a form of acute respiratory illness in which primarily the air sacs in the lungs (alveoli)
are affected. The alveoli usually fill with air when a person without pneumonia breathes, but in a
person with pneumonia they are filled with pus and fluid and this makes breathing difficult and
painful, and limits the oxygen intake. Pneumonia can be caused by various infectious agents
including viruses, bacteria, and rarely fungi. The most common causes of pneumonia in children are:
Streptococcus pneumoniae, Haemophilus influenzae, respiratory syncytial virus (the most common
viral cause) and, in the case of HIV-infected children, Pneumocystis jiroveci (WHO, 2013e). Viral
6
and bacterial infections in children present in a similar way. The signs and symptoms of pneumonia
commonly include: rapid breathing, fever, cough, nasal flaring, lower chest-in-drawing and low
oxygen saturation.
The diagnosis of pneumonia may involve the use of chest x-rays, blood tests, sputum tests, and pulse
oximetry to measure the oxygen levels. However, these methods may not be available in many
resource-limited settings, and WHO has proposed a symptom-based classification of pneumonia. In
children under five years who have cough and/or difficult breathing, with or without fever, presence
of fast breathing or lower chest-in-drawing is classified as pneumonia (WHO and UNICEF, 2008).
These suggestions for disease classification were made in the 1980s before the introduction of newer
vaccines that target the common causes of pneumonia and before the effects of the HIV pandemic.
The presentation of pneumonia could have changed during this time (English and Scott, 2008). The
use of the suggested symptoms for pneumonia classification also creates challenges because they
overlap with those of malaria and are therefore not specific (Graham et al., 2008).
Pneumonia can be prevented through: immunization with Haemophilus influenzae type b (Hib),
pneumococcal, measles, and whooping cough vaccines; adequate nutrition with exclusive breast
feeding of infants for the first six months and giving vitamin A supplements; addressing
environmental factors such as the minimization of indoor air pollution; and encouraging good
hygiene, especially in crowded homes. H. influenzae vaccine currently reaches about half of the
world’s children while the pneumococcal vaccine is mostly available in high-income countries.
Measles vaccine is more widely used with coverage of about 84% in 2011 (WHO, 2013b).
1.9 CARE-SEEKING FOR CHILDHOOD ILLNESS
Prompt treatment (within 24 hours of onset of symptoms) of common causes of child mortality with
effective medicines is critical for improvement of treatment outcomes (UNICEF, 2013c; UNICEF
and WHO, 2006; WHO, 2005a). An important starting point is that caregivers seek appropriate
health care. However, caregivers delay or do not take children to health providers. Home care of
childhood illnesses is common, with most households taking children to health providers only after
home care with medicines bought from a nearby drug shop or pharmacy, or left over from a previous
illness or herbs has failed (Diaz et al., 2013; Smith et al., 2010; Nsungwa-Sabiiti et al., 2005; Ukwaja
and Olufemi, 2010). The sources of care outside the home are varied and may include: private drug
shops and clinics, public health facilities, non-governmental health organization facilities and
traditional health providers (Rutebemberwa et al., 2009b).
Care-seeking is influenced by caregiver recognition of symptoms, perceptions of illness severity and
community-preferred practices towards illness as well as long distances to health facilities, financial
constraints such as lack of transport or money for health care, availability of alternative sources of
treatment (e.g. traditional healers), symptoms of the child’s illness, and perceptions and attitudes
towards health providers (Getahun et al., 2010; Diaz et al., 2013; Najnin et al., 2011; Chibwana et al.,
2009; Hildenwall et al., 2008; Tinuade et al., 2010; Das and Ravindran, 2010; Ahorlu et al., 2006;
Maslove et al., 2009; Warsame et al., 2007; Taffa and Chepngeno, 2005; Ndugwa and Zulu, 2008;
Luque et al., 2008; Colvin et al., 2013; Comoro et al., 2003; Okeke and Okeibunor, 2010; Källander
et al., 2008; Rutebemberwa et al., 2009a; Ukwaja and Olufemi, 2010).
Socio-economic status of households has a strong influence on when and where people seek care.
Wealthier households are more likely to seek care for their children outside the home and are
unlikely to be impeded by delays which arise from lack of money for transport or health care
(Kristiansson et al., 2009; Colvin et al., 2013). Therefore strategies that increase availability of
7
affordable health care where people live are likely to influence the care-seeking practices for children
especially for poor children.
1.10 GLOBAL TARGETS OF CHILD MORTALITY
In 2000 all UN member states made a commitment to the United Nations Millennium Declaration
which had eight goals referred to as the Millennium Development Goals (MDGs). One of the goals
(MDG4) was reduction of child mortality by two thirds between 1990 and 2015 (United Nations,
2000).
1.10.1 Progress towards MDG4
Some progress has been made regarding reduction of under-five mortality, but it is still insufficient to
reach the MDG4 targets. Reduction in under-five mortality grew from 1.8% between 1990 and 2000
to 3.2% between 2000 and 2011 (United Nations, 2011). Sub-Saharan Africa, the region with the
highest U5MR, also doubled its mortality reduction rate, from 1.5% in the period between 1990 and
2000 to 3.1% in the period 2000-2011. However, with its mortality still at 98 deaths per 1,000 live
births in 2012, this reduction rate is insufficient to reach the target of 58 deaths per 1,000 live births
(United Nations, 2013; UN Inter-agency group for child mortality estimation, 2013).
1.10.2 Global strategies to accelerate achievement of MDG4
In order to reduce child mortality there is need to ensure universal coverage of key effective and
affordable interventions including appropriate care for newborns and their mothers; vaccines; infant
and young child feeding; prevention and case management of pneumonia, diarrhea and sepsis;
malaria control; and prevention and care of HIV/AIDS (WHO, 2013d). The WHO therefore
promotes strategies on appropriate and timely treatment of complications for newborns, integrated
management of common illnesses in under-fives, expanded programmes on immunization, and infant
and young child feeding. Globally several other strategies and initiatives have been implemented to
address the MDG4 and these include: the Global Strategy on Measles Initiative, Global
Immunization Vision and Strategy 2006-2015 (GIVS) (WHO, 2011b), the Global Alliance for
Vaccines and Immunization (GAVI) (GAVI Alliance, n.d.), and the Global Strategy on Women’s
and Children’s Health (The Partnership for Maternal Newborn and Child Health, n.d.). There are
other strategies that affect MDG4 but which were implemented to address MDG6 such as the Global
Fund initiative on malaria, TB and HIV (The Global Fund to fight AIDS tuberculosis and malaria,
2005); the Stop TB strategy (WHO, 2006a); the Roll Back Malaria Initiative (WHO, 2005b) and
President Obama’s Global Health Initiative (Global Health Initiative, 2009).
With two years left to 2015, the world is now looking to the post-2015 era. A number of areas have
already been identified where emphasis is to be placed, including: providing a more holistic
framework to address diseases and also putting into perspective the increasing burden of noncommunicable diseases, going beyond focus on mortality but also addressing morbidity and quality
of life issues; and strengthening health systems to address inequalities and ensure sustainable
progress. Universal health coverage will be key in the post-2015 agenda (Save the Children, 2012;
Task Team for the Global Thematic Consultation on Health, 2013).
1.11 STRATEGIES TO ADDRESS CHILD MORTALITY
Over the years a number of strategies have been developed to address, among other things, under-five
mortality. These strategies include Primary Health Care (1978), the Child Survival and Development
Revolution (1980), Integrated Management of Childhood Illnesses (IMCI) (1996), Home
8
Management of Malaria (HMM) (2002) and more recently the Integrated Community Case
Management of Childhood Illnesses (ICCM) (2004).
1.11.1 Primary health care
Primary health care was adopted with the Alma-Ata Declaration of 1978 and its ultimate goal was
Health for All. It made some achievements that reduced under-five mortality including increase in
childhood immunization coverage and access to safe water and sanitation. However, it suffered
setbacks to providing equitable access to essential health care due to financial barriers, health worker
shortages and the HIV pandemic (WHO, 2008).
1.11.2 Child survival and development revolution
In the 1980s the global recession had severe negative impact on especially the poor. Infant and
childhood mortality were selected as the major indices to address since they were used to measure the
development of countries. Interventions with simple technologies were thus adopted in what was
termed as GOBI including growth monitoring to keep regular check on child nutritional status, oral
rehydration for diarrheal diseases, breast feeding as the perfect nutrition start in life and
immunization. This strategy had saved the lives of about 12 million children by the 1990s (UNICEF,
1996). However, all these strategies were set up as vertical programmes, and in some countries
immunization took the largest proportion of the budget.
1.11.3 Integrated management of childhood illnesses
IMCI was started in 1996 as a strategy that integrates key available measures for disease prevention
and health problems during childhood, for their early detection and effective treatment, and for
promoting healthy habits within the family and community. Its aims were meant to be achieved
through: improving the skills of health workers for the prevention and treatment of childhood
diseases, strengthening the health system, and improving family and community care practices
(community IMCI, c-IMCI). Its aims were to reduce infant mortality and the incidence and
seriousness of illnesses and health problems, as well as to improve growth and development of
children during the first years of life (Pan American Health Organization and WHO, n.d.). IMCI has
been shown to improve health worker performance, and to lead to better quality of care and rational
drug use at costs that are lower or similar to investments in routine child health services in areas
where it has been implemented (Nguyen et al., 2013; Gouws et al., 2004; Mason et al., 2009).
However, it did not achieve the expected impact on mortality, mainly because of the poor healthseeking behaviour that made caregivers delay or not take sick children to health facilities (Chopra et
al., 2012).
1.12 COMMUNITY CASE MANAGEMENT
The need to address the poor treatment seeking behaviors for children under IMCI led to the
recommendation of community case management (CCM). There was low use of lifesaving
interventions in areas with high child mortality resulting from: long distances to health facilities,
inadequate knowledge of illness and sources of health care, and mistrust of the quality of care offered
by the health facilities. CCM was therefore meant to complement the health-facility based services in
expanding capacity to treat children with poor access to health facilities (CORE Group et al., 2012).
It strengthens the treatment aspects of c-IMCI and not only includes the provision of treatment in
communities but also promotes timely seeking of care, appropriate home care, and referrals to health
facilities as well as supervision by the health facilities. CCM targets the conditions that cause most
child deaths and has been implemented mostly for malaria and diarrheoa, but also for pneumonia,
malnutrition and neonatal conditions (de Sousa et al., 2011). A key feature of CCM is the use of
9
trained and supervised community members, commonly referred to as ‘community health workers’ to
deliver care with linkage to the health facilities. The health providers under CCM classify the
children’s illness using symptom-based algorithms adopted from the IMCI guidelines. The WHO
recommendation of universal parasite-based diagnosis in 2010 (WHO, 2010a) has necessitated the
incorporation of rapid diagnostic tests (RDTs) for malaria into CCM. There is limited experience
with integrated community care for the five target conditions.
1.12.1 Community health workers
CHWs have participated in the provision of primary health care in various parts of the world for
several decades. There is a growing body of evidence supporting their contribution to the
improvement of the health of communities, especially in areas with limited access to professional
health providers (WHO, 2010b). Literature shows that CHWs can successfully provide care for
childhood illnesses (Mubi et al., 2011; Ajayi et al., 2008b; Sazawal and Black, 2003; Theodoratou et
al., 2010; Hopkins et al., 2007; Sirima et al., 2003; Baqui et al., 2009; Perry and Zulliger, 2012).
However, many of these studies have involved CHWs who are managing single illnesses.
1.12.1.1 Definition of Community Health Workers
The term ‘community health worker’ comprises a variety of community health aides who are selected
and trained to provide basic health care in their communities. The CHWs are referred to by a wide
range of names such as community drug distributors (CDDs), village malaria workers, lay health
workers, community medicine distributors (CMDs), village health teams (VHTs), to mention a few,
but the basic defining characteristics are that the persons are members of the community where they
work; selected by the communities; answerable to the communities for their activities; supported by
the health system but not necessarily part of that health system; and trained for shorter durations than
the professional health workers (WHO, 2007a).
The profiles of CHWs vary from setting to setting and are therefore difficult to generalize
internationally. The CHWs may be men or women, young or old, literate or illiterate, paid workers or
volunteers. The most important point is that the definition of the CHW is responsive to the societal
norms to ensure community acceptance and ownership (WHO, 2007a; Prassad and Muraleedharan,
2007; Bhutta et al., 2010).
The roles of CHWs are diverse within and across countries and programs. In some cases they may
perform a wide range of tasks that could be preventive, curative or developmental, while in other
cases they are appointed for very specific tasks (WHO, 2007a). The training of the CHWs may also
vary from a few days to several months (O'Brien et al., 2009; Prassad and Muraleedharan, 2007; de
Sousa et al., 2011).
1.12.2 Home management of malaria
The home management of malaria strategy (HMM) was recommended by the WHO in order to
improve access to care for malaria. Uganda was one of the first countries to implement this strategy,
locally referred to as the home-based management of fever (HBMF), in 2002 (Uganda Ministry of
Health, n.d.). The strategy aimed at availing antimalarials for prompt community treatment of fever
in children. Some studies found that community management of fever with antimalarials significantly
increased the proportion of children that received prompt treatment (Staedke et al., 2009; NsungwaSabiiti et al., 2007); and reduced under-five morbidity (Das et al., 2008; Sirima et al., 2003) and
mortality (Kidane and Morrow, 2000; Hopkins et al., 2007). However, significant symptom overlap
between malaria and pneumonia has been noted (English et al., 1996; Källander et al., 2004; Ukwaja
10
and Olufemi, 2010), which could have resulted in many patients with pneumonia perhaps being
inappropriately managed with antimalarials only (Källander et al., 2008). The integrated management
of malaria and pneumonia has been advocated for because of this symptom overlap.
1.12.3 Integrated community case management of common
childhood illnesses
WHO and UNICEF have recommended the integrated community case management of common
childhood illnesses (ICCM) in response to findings of multiple illnesses in children or symptom
overlap of common illnesses (WHO and UNICEF, 2004). In this strategy CHWs assess children for
symptoms of malaria, pneumonia, and diarrhea and give antimalarials, antibiotics or oral rehydration
salts (ORS) and zinc as appropriate. The CHWs also identify and refer children with severe disease to
health facilities. The success of ICCM also depends on the caregivers of sick children, who should be
able to recognize symptoms, seek care from the CHWs promptly, adhere to treatment or referral,
monitor the condition of the children and seek further care if the children do not get well (WHO and
UNICEF, 2011).
However, there are a number of potential challenges to the implementation of ICCM. It is not clear
whether the CHWs will effectively handle the more extensive roles they need to perform under
ICCM. In addition, it is not clear if the caregivers will be able to perform their roles in this strategy
effectively or whether the intervention will improve drug use as well as treatment outcomes. There
have been limitations in medicines supply management in formal health systems and communitybased programmes. These systems have been marred by frequent drug stock-outs (Blanas et al.,
2013; Achan et al., 2011; Kangwana et al., 2009). The ICCM strategy fitting into this health system
is likely to be similarly affected. Furthermore, community-based strategies are aimed at improving
equity, but it not clear if this will be achieved under ICCM. Another major concern is the balance
between increasing access to antimicrobials and their excess use. The effect of ICCM on rational use
of drugs in the community is not certain.
1.12.3.1 Equity in child health programs
Equity in health care is currently of great concern globally because the populations that suffer the
biggest burden of disease have the lowest access to effective health care. Low-income countries bear
more than 50% of the global disease burden but have only about 2% of the total health expenditure
(WHO, 2011a). These disparities are also visible within countries where the poorest people, those
with lower education, and those living in rural areas have poor access to health care (United Nations,
2011). Many of the interventions aimed at reducing child mortality stress the need to ensure that the
poor people access them. Community-based programmes are likely to address the gaps in equity for
childhood interventions (de Sousa et al., 2011). It is important to ensure that this is achieved under
ICCM.
1.13 UGANDA COUNTRY PROFILE
Uganda is a landlocked country in the eastern part of Africa. It borders Kenya to the east, Tanzania to
the south, Rwanda to the southwest, the Democratic Republic of Congo to the west and South Sudan
to the north. It covers an area of 241,038 square kilometres and has a population of 36 million people.
About 48.5% of the total population is made up of children aged below 15 years. The life expectancy
at birth was 56 years in 2011. The country is mainly rural, with about 16% of the population living in
urban areas. The majority of the population is involved in agriculture. The gross national income per
capita was 1,310 US dollars in 2011 (WHO, 2012a; United Nations Development Program, 2013;
WHO, 2013f).
11
1.13.1 Uganda health indicators
Uganda is classified as a country with high child and very high adult mortality. Its infant mortality
rate was 54 deaths per 1,000 live births in 2011 and the U5MR was 90 deaths per 1,000 live births.
There has been a reduction in the death rate but the birth rate has continued to escalate resulting in a
young age structure. The fertility rate has reduced slightly from 7.1 births in 2000 to 6.1 births in
2010 but this figure is still high. The maternal mortality ratio is also high at 310 deaths per 100,000
live births and a considerable number of children are stunted (38%) (WHO, 2013f; WHO, 2012a). In
addition, the full immunization coverage in 2011 was at only 52% (Kasirye, 2012). Pneumonia and
malaria are common causes of death in under-fives, contributing about 17% and 13% of under-five
mortality respectively (Countdown to 2015, 2012).
Uganda has had slow progress towards achievement of MDG4. Although some reduction in underfive mortality has been noted, the country is not on track to achieving the set targets. In 2000 the
U5MR was 152 deaths per 1,000 live births and this had reduced to 90 deaths per 1,000 live births by
2011. With only four years left to reach 2015 the U5MR in 2011 was still much higher than the
MDG4 target of 56 deaths per 1,000 live births by 2015 (WHO and UNICEF, 2012).
1.13.2 Uganda health system
Health services in Uganda are provided by both public and private sectors with each sector having
about 50% of the health service delivery outlets. The public sector is comprised of a tiered system
with the range of services offered and staffing levels varying according to the level of the health
facility. Highest in the level of health service outlets are the national referral hospitals, followed by
the regional referral hospitals, general hospitals, health centre IVs, health centre IIIs, health centre IIs
and finally the health centre I/village health teams (VHTs). The VHT has no physical structure, but is
a team of people who work at the community level and link the community and health facilities
(Table 1) (Ministry of Health (Uganda), 2010b; Ministry of Health (Uganda) et al., 2012).
The Ministry of Health in Uganda provides leadership for the health sector and has the overall
responsibility for the delivery and oversight of all health services in Uganda. The health services are
decentralized to districts and sub-districts which are responsible for implementation of services. The
health services in all public health facilities are free since the abolition of user fees in 2001.
Table 1. Structure of the Uganda National Health System
Health unit
Physical structure
Location
National Referral
Advanced tertiary care
National
Hospital
Regional Referral
Specialist services
Region (8-10 districts)
Hospital
General Hospital
Hospital, laboratory, surgery, District
and medical imaging
Health Centre IV
Out- and in-patient services,
County
theatre, laboratory and blood
transfusion
Health Centre III
Out-patient services,
Sub-county
maternity, general ward, and
laboratory
Health Centre II
Out-patient services only
Parish
Health Centre I
None
Village
12
Staffing levels
Specialist physician
Specialist physician
Medical officer
Medical officer
Clinical officer
Nurse
CHW
Although Uganda has most of the relevant health policies and regulations in place, there are still
several challenges in the health system. The health sector is under-financed with per capita
expenditure on health at only US$42 in 2011 which was lower than the sub-Saharan regional average
of US$951.6 (World Bank, 2012). Public financing is low at only 22.6% of the total health
expenditure whereas out-of-pocket expenditure is high at about 54% (Ministry of Health (Uganda) et
al., 2012). The low financing implies that the country is unable to deliver the Uganda Minimum
Health Care Package (a minimum package of services comprising the most cost-effective
interventions for major causes of disease burden), and yet this is one of the ways that the country has
articulated as a means to improve the health of the people.
Access to health services has improved over the past two decades but some people especially the
poor and those living in the rural areas still find it difficult to access primary health care and the
quality of services provided is still a major setback. There are frequent stock outs of drugs and
supplies at many public health facilities (Tumwine et al., 2010; Achan et al., 2011; Ministry of
Health (Uganda), 2011). In addition, although there has been some improvement in human resource
generation there are still shortages, and the distribution of the health workers favours the urban areas
(Healthy Child Uganda, 2011).
1.13.3 The role of the private sector
Uganda has a vibrant private sector covering about 50% of the health services in the country
(Ministry of Health (Uganda) et al., 2012). The private health system comprises private not for profit
(PNFP) facility- and non-facility-based health providers, private health practitioners and traditional
and complementary medicine practitioners. The facility-based PNFP sector covers about 41% of the
hospitals and 22% of the lower level health facilities. The PNFP health providers have about 70% of
the health training institutions in the country and these are run with government support. The PNFP
practitioners are more prominent in rural areas. The private for profit (PFP) health facilities focus
mainly on primary level services and are based mainly in urban and semi-urban areas. The
government of Uganda recognizes the role of the private sector and subsidizes the PNFP facilities,
and a few private hospitals. The private sector efforts are coordinated through the public-private
partnerships for health (PPPH) technical working group. In spite of this much of the private sector
remains unregulated, leading to great variations in quality, capacity and availability of services
(Interagency Health Team, 2011). At the top of the ladder are big private hospitals with specialized
services, while at the bottom are the private clinics and drug shops which are often manned by
unqualified personnel and often have poor quality services (Konde-Lule et al., 2010).
1.13.4 Community health workers in Uganda
Uganda has had some experience of using CHWs for provision of health-care related activities under
various programmes including: providing malaria treatment under home-based management of
malaria (HBMF) through CDDs or CMDs, providing nylon filters and educating communities to
control spread of the guinea worm under the Uganda guinea worm eradication programme (UGWEP)
by trained volunteers, distributing and directly observing tuberculosis treatment under the
community-based directly-observed therapy-short-course (CB-DOTS) programme by community
volunteers, distribution of ivermectin for onchocerciasis under the community-directed treatment
with ivermectin (CDTI), and health education and promotion through the VHTs (WHO and Global
Health Workforce Alliance, 2010). Some of the CHW programmes have had nearly country wide
implementation while others have been implemented in a few areas.
13
1.13.4.1 Village health teams
The concept of the VHT became part of Uganda’s national health strategy in 2001 as Uganda’s
commitment to the Alma Ata Declaration of 1978 and the 2008 WHO’s Ouagadougou Declaration
on Primary Health Care and Health Systems in Africa (Ministry of Health (Uganda), 2010c). These
declarations emphasize the involvement of communities in their health. The VHT is meant to serve
as the primary (village level) health contact for all villages in Uganda. VHTs are volunteer CHWs
who are involved in community information management, health promotion and education,
mobilization of communities for utilization of health services and health action, community case
management and follow-up of major killer diseases (malaria, diarrhoea, pneumonia) and
emergencies, care of the newborns, and distribution of health commodities. The strategy aims to have
five VHT members in each village. The VHTs are selected through a popular vote by the
communities in which they live and serve. The district health teams together with Ministry of Health
officials conduct the training for VHTs.
By 2009 about 77% of all districts had trained VHTs but in some districts the VHTs are inactive and
the attrition rate is high (Ministry of Health (Uganda), 2009a). The target of the Health Sector
Strategic and Investment Plan (HSSIP) 2010-2015 is to increase the active VHTs from 31% to 100%
(Ministry of Health (Uganda), 2010a).
Sustainability of the activities of the VHTs is critical. The CHWs need to be motivated to perform
their functions as well as continue in these roles for a sufficient length of time (retention).
Furthermore conflicting commitments may affect performance since many of the CHW programmes
are not salaried and the CHWs need to find other occupations for their sustenance.
1.13.4.2 Home based management of fever in Uganda
HBMF was introduced in Uganda in 2002 as a strategy for community case management of malaria
(Batega et al., 2004). In this strategy CHWs who were commonly known as community drug
distributors (CDDs) treated children with fever using a combination of chloroquine and
sulphadoxine-pyrimethamine (CQ-SP) (also called Homapak®). Each village had two CHWs who
were supervised by health workers from health facilities in the area served. Children with fever were
presumptively treated for malaria.
HBMF was shown to improve access to prompt and appropriate treatment for malaria (NsungwaSabiiti et al., 2007) but its utilization was low, mainly because CHWs were managing malaria only
while some children had other illnesses or they had multiple illnesses. In addition, some caregivers
perceived it as ineffective (Malimbo et al., 2006). HBMF ceased to be functional when the
antimalarial policy in Uganda was changed from CQ-SP to ACTs (specifically artemetherlumefantrine (AL)) due to the widespread resistance to CQ-SP (Nanyunja et al., 2011). However,
before HBMF with AL could be rolled out in Uganda, studies showed that there was a lot of
symptom overlap between malaria and pneumonia (Källander et al., 2004) and mistreatment of
children suffering from other illnesses with antimalarials (Källander et al., 2008) which necessitated
the integrated management of malaria and pneumonia.
1.13.4.3 Integrated community case management of childhood illnesses
Uganda adopted the ICCM strategy in 2010 (Ministry of Health (Uganda) et al., 2010). ICCM in
Uganda is part of the VHT strategy for promoting health and preventing deaths. Its goal is to reduce
child morbidity and mortality by providing case management for malaria, pneumonia, and diarrhea to
sick children in addition to identifying and referring sick newborns. The VHT will be supplied with a
14
kit containing: pre-packaged medicines including amoxicillin for non-severe pneumonia, ACTs for
uncomplicated malaria, low osmolarity ORS and zinc for diarrhoea; rectal artesunate for pre-referral
administration; diagnostic commodities and supplies such as respiratory timers, and mid-upper-arm
circumference tapes. The VHTs will counsel mothers on home care and care-seeking and will
mobilize communities to demand, support, and use the ICCM programme where children with fever,
cough and diarrhea will be treated.
Newborns with danger signs and severely ill children will also be referred by the VHTs and prereferral rectal artesunate will be given as appropriate. Trained staff at health facilities will manage the
referred cases and supervise VHTs in their catchment area as well as monitor the programme’s
progress. The program also includes peer supervision among VHTs (Ministry of Health (Uganda) et
al., 2010). The ICCM strategy has been implemented in more than 25 districts in Uganda mainly by
non-governmental organizations (NGOs).
15
2 RATIONALE FOR THE STUDIES
Uganda adopted the ICCM strategy in 2010 and its scale-up is still ongoing (Ministry of Health
(Uganda) et al., 2010). For effective implementation of this strategy, some critical knowledge gaps
need to be addressed.
Caregivers of children under five have to seek care from the CHWs promptly in order to realize the
advantages of ICCM in providing integrated health care close to where people live. Caregivers have
been found to seek care late (Ahorlu et al., 2006; Källander et al., 2008) or not to seek care at all
(Diaz et al., 2013). Low utilization of the HMM strategy was reported (Kisia et al., 2012; NsungwaSabiiti et al., 2007) and it is unclear if ICCM will be optimally utilized. Having ICCM in the
community will be a waste of resources if caregivers do not seek care for their children from CHWs,
and the desired impact of ICCM on mortality and morbidity will not be realized.
A key assumption of ICCM is that the CHWs will be able to elicit signs and symptoms related to the
three illnesses of interest (malaria, pneumonia, and diarrhea), classify the illnesses, prescribe
appropriate medicines and give appropriate instructions to the caregivers, or refer severely ill children
to health facilities. However, it is unclear how CHWs will perform in the provision of care for
multiple illnesses. The studies that have assessed performance of CHWs in provision of care for
children less than five years old have mostly been in the context of management of single illnesses
(Ashwell and Freeman, 1995; Ajayi et al., 2008b; Chinbuah et al., 2006; Nsungwa-Sabiiti et al.,
2007; Nsungwa-Sabiiti et al., 2005). More complex algorithms will be used for diagnosis, treatment,
and referral under ICCM which require more knowledge and skills from the CHWs. A few studies
that have looked at management of multiple illnesses have shown some gaps in CHW management
of multiple illnesses (Gilroy et al., 2012; Mukanga et al., 2011; Degefie et al., 2009; Kelly et al.,
2001).
In order to complement the work done by the CHWs, caregivers have to ensure good adherence to
the medicines received from CHWs in order to improve treatment outcomes as well as prevent
development of resistance. It is uncertain what effect the complexity of having more drugs will have
on the quality of counselling by CHWs and therefore adherence. In addition, it is uncertain what
effect the increase in pill burden among children receiving antibiotics in addition to antimalarials will
have on adherence to medicines. Another concern is whether the caregivers will adhere to referral
advice which is important for minimizing mortality from severe disease.
The easy availability of antimalarials and antibiotics from the CHWs may lead to their overprescription thus increasing drug pressure and the risk of drug resistance in the community. In
addition, other health providers may increase their prescription of these medicines, further increasing
drug pressure. It is not yet clear how the availability of medicines through the CHWs will affect
community drug use patterns. In addition, although integrated care has been recommended to
improve timeliness of appropriate treatment for common childhood illnesses, it is uncertain whether
this will be achieved and what impact the integrated care will have on treatment failures.
16
3 CONCEPTUAL FRAMEWORK
A conceptual framework has been included in this thesis in order to define the key concepts and
contexts that supported and informed this research (Leshem and Trafford, 2007). The conceptual
framework used in this thesis is modified after the access framework employed in the setting of
livelihood insecurity (Obrist et al., 2007). This conceptual framework, even though centred around
access, has been considered suitable for this thesis, which goes beyond access to look at the outcomes
of the access. This is because the conceptual framework has a broad overview of the contexts
preceding access to care as well as the consequences of access to health care, which are the focus of
this thesis. It looks at the care-seeking, the health services available and the livelihoods of the people
in the setting. In addition, this conceptual framework has been designed to look at the broad aspects
of care-seeking in resource-poor settings such as Uganda where the studies were based.
Although effective interventions for most of the common conditions causing morbidity and mortality
exist, these interventions are not accessed and utilized by the people that need them, resulting in poor
health outcomes (WHO, 2007b). The access to health interventions starts when the illness is
recognized and treatment is sought. The recognition of illness and treatment seeking is influenced by
access to livelihood assets, which include physical capital comprising of knowledge, skills, and local
education; social capital comprising of social networks and affiliations; natural capital like land,
water, and livestock; physical capital such as infrastructure, equipment and means of transport; and
financial capital such as cash and credit. The livelihood assets of people are, however, influenced by
forces over which people have little control, such as the economy, politics, or technology, climatic
variability or shocks e.g. floods, landslides, etc. which are referred to as the vulnerability context.
At the centre of this conceptual framework are the five dimensions of health care access, including
availability, accessibility, affordability, adequacy, and acceptability of health services. The extent of
accessibility reached along the five dimensions depends on the interaction between (a) the health care
services and the broader policies, institutions, organizations and processes and (b) the livelihood
assets people in certain vulnerability contexts have access to. This interaction between factors will
influence the utilization of health services. Coupled with the quality of care provided from different
providers, this will influence the health status of the patients, the patient satisfaction with the care
received as well as equity (Obrist et al., 2007).
In this thesis the access conceptual framework has been extended to include under-five mortality,
which is the key target for many health care interventions in children less than five years of age,
although it has not been studied in this thesis. The quality of care section has also been expanded to
indicate the quality aspects that were focused on in this thesis, specifically the performance of
CHWs, community drug use for malaria and pneumonia, and adherence to medicines. The thesis has
in addition to the above addressed the health status and patient satisfaction as well as the influence of
livelihood assets on utilization of health services and quality of care. Resistance has been included in
the section on outcomes of care since it is a common concern for community use of antimicrobials
(Marsh et al., 2008). The conceptual framework is subsequently used to provide the format against
which the results are presented and discussed.
17
POLICIES, INSTITUTIONS,
ORGANIZATIONS & PROCESSES
HEALTH CARE SERVICES
Health facilities, Private practice, Drug shops,
Traditional Healers, CHWs
ACCESS
Availability
Accessibility
Affordability
Adequacy
Acceptability
UNDER-FIVE
MORTALITY
UTILIZATION &
QUALITY OF CARE
Utilization of services
(Study IV)
Health Status
(Study III)
Quality of Care
- Performance of CHWs
(I & IV)
Patient
satisfaction
(Study I & IV)
- Community drug use
for malaria and
pneumonia (III & IV)
Equity
(Study III & IV)
Drug resistance
- Adherence (II)
Illness recognized and treatment seeking
initiated (Study III & IV)
LIVELIHOOD ASSETS (Study II, III & IV)
Physical capital
Social
capital
Natural
capital
Human
capital
Financial
capital
VULNERABILITY CONTEXT
Figure 3. Conceptual framework showing ICCM in the context of the Access Framework
Adapted from the Access Framework (Obrist et al., 2007)
18
4 AIM AND OBJECTIVES
4.1 GENERAL AIM
To assess care-seeking and quality of care in integrated community case management of malaria and
pneumonia in children aged less than five years in Uganda so as to inform the implementation of the
ICCM strategy.
4.1.1 Specific objectives
1.
To compare the performance of community health workers in the integrated community
case management of malaria and pneumonia versus malaria management alone in under-five
children and to assess the factors influencing their performance (Study I).
2.
To compare adherence to antimalarials and antibiotics versus antimalarials alone in the
integrated community case management of malaria and pneumonia in under-five children
(Study II).
3.
To determine the effect of integrated malaria and pneumonia management on receiving
prompt and appropriate antibiotics for pneumonia symptoms and treatment outcomes as well
as determine associated factors (Study III).
4.
To determine the effect of the integrated community case management of malaria and
pneumonia on utilization of community health workers and community drug use patterns in
under-five children (Study IV).
19
5 METHODS
5.1 STUDY AREA AND POPULATION
The studies were carried out in Iganga-Mayuge Health and Demographic Surveillance Site (HDSS),
which covers an area of 155 square kilometres in the two districts of Iganga and Mayuge in eastern
Uganda. The districts are about 112 kilometres from Kampala, the capital city of Uganda.
Iganga district
Mayuge district
Figure 4. Map of Uganda showing Iganga and Mayuge districts with arrows
The HDSS is made up of 65 villages with a total population of about 70,000 people living in 13,000
households, with the population of children below five years about 11,000. The area is predominantly
rural; about 90% of the population lives in rural areas. The main causes of morbidity and mortality
for children in the area are: malaria, pneumonia, and diarrhea. There are 13 health facilities, and 122
drug shops and private clinics. In addition, 131 CHWs were providing health care in the area to
children aged 4-59 months since 2009 under a cluster randomized controlled trial (cRCT) that was
designed to evaluate the impact of integrated management of illness with antimalarials and antibiotics
on under-five mortality.
In this trial, the HDSS was divided into intervention (30 villages where CHWs treated both malaria
and pneumonia) and control areas (35 villages where CHWs treated malaria only). There were two
CHWs in each of the villages, except one village which had three CHWs because of its large size
(initially it was two villages that had three CHWs but the third CHW in one of the villages withdrew
from the area). The data collection for this thesis commenced about two years after the start of
implementation of the cRCT.
20
5.2 DESCRIPTION OF THE CLUSTER RANDOMIZED TRIAL
Cluster randomized trials are types of randomized controlled trials in which groups of subjects rather
than individuals are randomized to interventions. They are usually employed to study interventions:
that cannot be directed towards individuals, where contamination is likely to occur from subjects
sharing interventions, and where it is socially unacceptable to assign some subjects in a certain
location to one group while others are assigned to another group. They could also be used for
logistical reasons for the ease of ensuring that contamination does not occur (i.e. that only subjects
randomized to the intervention use it). Although they minimize the risk of contamination of
interventions, they involve greater complexity in the design and analysis (Hayes and Moulton, 2009).
This cRCT involved CHWs that could treat children with malaria and/or pneumonia (intervention),
or with malaria only (control) and referred children with pneumonia to health facilities. Cluster
randomization was employed because it would have been socially inappropriate and difficult for the
same CHW to treat some children with pneumonia in their locality and refer others who also had
pneumonia to health facilities. In addition, cluster randomization eased assessment of the impact of
the intervention on mortality and other outcomes.
5.2.1 Selection of CHWs
Although CHWs had been used previously in Uganda to provide home management of malaria under
the HBMF strategy, they were not active at the time of commencement of the cRCT. This was
because CHWs had previously used combinations of chloroquine and sulphadoxine-pyrimethamine
(CQ-SP) for treatment of fever, but the change in antimalarial policy that occurred in 2005
necessitated a change in medicines to the recommended ACTs (Nanyunja et al., 2011). However, use
of ACTs by CHWs had not yet been widely rolled out when recommendations were made to adopt
the ICCM strategy. Therefore, as part of the cRCT, CHWs were selected from the communities
where they were to serve. Persons were eligible for selection as CHWs if they were permanent
residents of the areas where they were to serve, able to read and write, respectable members of the
community, and good mobilizers. The selected persons were trained either in the management of
malaria alone (if they were in the control areas) or in the integrated management of malaria and
pneumonia (if they were in the intervention areas).
5.2.2 Training of CHWs
All the CHWs in the cRCT were trained for three days on various aspects of malaria including signs
and symptoms, danger signs, transmission, prevention, and populations at risk of malaria. The
training on malaria was done before randomization of areas into intervention and control areas. After
randomization the CHWs in the intervention arm received three days’ training on integrated malaria
and pneumonia management, thus receiving additional training on acute respiratory tract infections
(ARI) that included: signs and symptoms, use of respiratory timers in counting breathing rates,
danger signs, and prevention of pneumonia. During the training on malaria, all CHWs (including
those in the control arm) had been informed about the signs and symptoms of ARI, but not how to
assess and classify these signs and symptoms. CHWs in both the intervention and control areas were
also trained on referral, filling in registers, managing drug supplies, counselling caregivers of
children, and adverse reaction monitoring. In addition, the CHWs received refresher training at
monthly meetings with the supervisors of the project and health workers.
Prior to the training of CHWs, members of the district health team; and health workers in public,
NGO, and private health facilities received two days’ training on integrated community management
of malaria and pneumonia, investigation and documentation of adverse events, as well as supervision
of CHWs. The training was conducted by Ministry of Health officials.
21
5.2.3 Randomization
Cluster randomization was used in which randomization was done at the parish level (i.e. each parish
was considered as a cluster). A parish is made up of several villages. The 65 villages of the HDSS
make up 26 parishes and these parishes were randomized to the intervention and control arm using
stratified block randomization. The parishes were stratified into eight urban and 18 rural clusters. The
clusters were further stratified according to the population of children less than five years in each
cluster.
In the rural area three strata based on the population of children less than five years were made, as
follows: (i) 190-320 children, (ii) 321-390 children, and (iii) 391 children, resulting in six clusters
in each of these strata. The clusters from the urban area were grouped into two strata also based on
population sizes of (iv) 280-430, and (v) 431. Random numbers were generated in blocks of six for
the rural clusters and in blocks of four for the urban clusters. In each of the three strata in the rural
area, three clusters were randomized to the intervention arm and three to the control arm. In each of
the two strata in the urban area, two clusters were randomized to the intervention arm and two
clusters to the control arm. There were no buffer zones between the clusters.
5.2.4 Supervision of CHWs
The CHWs received monthly support supervision from health workers at the nearest health facilities.
The health workers involved in supervision of CHWs were nurses, midwives, and clinical officers
(holders of 3-year diplomas in medicine). They visited the CHWs’ homes and checked the CHWs’
records for completeness and correctness of dosing, as well as the medicines boxes for
appropriateness of medicines storage.
5.2.5 Patient management in the cRCT
Diagnosis: The CHWs in the control arm treated children with non-severe malaria and referred
children with severe disease and those with pneumonia symptoms of any severity to nearby health
facilities. The CHWs in the intervention arm treated children with non-severe malaria and/or
pneumonia and referred children with severe disease to nearby health facilities. The algorithm
followed by CHWs is summarized in Figure 5. The CHWs in both the intervention and control arms
classified children’s illnesses based on symptoms similar to IMCI guidelines (WHO and UNICEF,
2008). Specifically, the CHWs classified children with fever or a history of fever in the last 24 hours
as having malaria. In the intervention arm, the CHWs classified children with cough or difficult
breathing and fast breathing as having pneumonia. A child was considered to have fast breathing if
their breathing rate was 50 breaths per minute if they were aged up to 12 months, or 40 breaths per
minute if they were aged 12-59 months. Children were considered to have severe disease if they had
any of the four general danger signs (repeated vomiting, convulsions, failure to feed or
lethargy/unconsciousness) or other danger signs like chest in-drawing, noisy breathing, severe
dehydration or pallor.
22
Control arm
Intervention arm
Pneumonia
symptoms
Fever or History
of fever
No
danger
signs
Danger
signs
No
pneumonia
Refer
Treat
with
AL
Danger
signs
Fever or History
of fever
No
danger
signs
Danger
signs or
pneumonia
symptoms
Pneumonia
symptoms
No
danger
signs
Pneumonia
Treat
with AL
+
AMOX*
Refer
Treat
with
AMOX*
Refer
Treat
with
AL
Refer
* Amoxicillin
Figure 5. Schematic figure of cluster randomized trial
Treatment provided: The CHWs used dispersible AL tablets (Coartem® 20mg artemether, 120mg
lumefantrine) manufactured by Novartis Pharma AG, for treatment of malaria in both the
intervention and control arms. The AL was pre-packaged in two age-specific doses (6 tablets in
yellow pack for children aged 4-35 months and 12 tablets in blue pack for children aged 36
months). In both age-groups, AL was given twice daily for three days.
In the intervention arm, the CHWs used pre-packaged amoxicillin tablets (125mg) manufactured by
Medophar, India or IDA Foundation, The Netherlands for treatment of pneumonia. The amoxicillin
was procured in bulk packages and re-packaged by a local pharmaceutical industry (Kampala
Pharmaceutical Industries) in three age-specific doses (6 tablets for children aged 4-11 months, 12
tablets for children 12-35 months and 18 tablets for children 36-59 months). Amoxicillin was
administered twice daily for three days.
23
5.3 STUDY DESIGN AND DATA COLLECTION METHODS
5.3.1 Study design
The thesis comprises four studies (I-IV) which were nested within the cRCT. Both qualitative and
quantitative methods were used to collect data. A cross-sectional study combined with focus group
discussions (FGDs) was used to assess the performance of CHWs (study I) in June 2011. A cohort
study was employed to assess adherence of children to antimalarial and antibiotic treatments (study
II) and the effectiveness of integrated malaria and pneumonia management on prompt and
appropriate treatment for pneumonia symptoms as well as treatment outcomes (study III) in October
and November 2011. The same cohort study was employed for studies II and III, but some of the
children that were included in study III did not participate in study II.
Cohort studies are types of studies where an investigator selects a group of persons with an exposure
of interest and a comparison group without the exposure. The groups are then followed up for
outcomes of interest (Gordis, 2000). The exposures of interest in study II were antimalarials and
antibiotics (intervention arm), antimalarials only (intervention arm), and antimalarials only (control
arm). In study III, the exposures of interest were seeking care from CHWs in the intervention or
control arms of the cRCT. A cross-sectional study combined with key informant interviews (KIIs)
was conducted in January to February 2011 to assess the utilization of CHWs and the effect of the
intervention on community drug use patterns (study IV). However, data from KIIs was not included
in the publication of study IV but has been presented in this thesis. The data collection methods for
studies I-IV are described in more detail in sections 5.3.2 and 5.3.3.
5.3.2 Overview of the data collection methods
Study I and IV employed both qualitative and quantitative data collection methods. Combining
qualitative and quantitative data has been said to strengthen research because each method
compliments and minimizes the limitations of the other (Creswell et al., 2003; Bryman, 2004). The
data from qualitative and quantitative methods were combined through triangulation. In addition,
data obtained from various sources using quantitative methods have been combined to answer the
study objectives. Triangulation is an approach to data analysis that synthesizes data from multiple
sources for a single problem, often combining qualitative and quantitative data. Different types of
triangulation exist, including combining data collected with different methods (methods
triangulation), data from different sources within the same method (data triangulation), data from
different observers (investigator triangulation) and using different perspectives or theories (theory
triangulation) (Patton, 1999). In this thesis there was mainly methods and data triangulation.
Questionnaires/structured interviews were employed in studies I, II, III and IV. Questionnaires
have been used to collect large amounts of data from a large number of people in a short time. Thus
they were very handy in collecting data for the considerably large sample sizes in the various studies.
Questionnaires/structured interviews can also be used by many people doing the interviewing with
limited influence on the validity, especially when adequate training has been done. This was
important because several research assistants were employed to collect the data. However, they are
not suited to collecting certain types of information, e.g. skills. Some world views state that they are
artificial creations by researchers where limited information without explanations is collected. The
questions may also be interpreted differently or researchers may impose their views (Hardon et al.,
2004). The responses to questions, especially those on knowledge in the various studies, were
unprompted to improve validity of assessment (Franco et al., 2002; Hwang et al., 2007). Prompted
questions were only used to obtain information on the presenting signs and symptoms of the
children’s illness.
24
Focus group discussions (FGDs) were employed in study I. They are a form of data collection
method that capitalizes on interaction between research participants to generate data on a topic which
the moderator helps to keep the group focused on. The method is particularly useful for exploring
people's knowledge, perceptions, feelings, and experiences and can be used to examine not only what
people think but how they think and why they think that way (Kitzinger, 1995; Stewart et al., 1990).
FGDs can provide naturalistic data, but the researcher’s influence on the interaction may affect this
naturalism (Morgan, 1997). In addition, the data cannot be generalized (Liamputtong and Ezzy,
2005) in a similar manner to data from surveys. Furthermore, some people’s opinion may be
suppressed through dominant voices taking over the discussion, and this calls for the moderator’s
expertise at ensuring that all views including minority dissenting ones, are captured (Smithson,
2000).
Key informant interviews were employed in study IV. They are methods of data collection where
individuals with specialized knowledge about the topic of interest are interviewed with semistructured or unstructured interviews. They are used to assess people’s experiences, inner
perceptions, attitudes and feelings of reality, which may not be possible to discern with
observations (Bryman, 2004).
Observations
Structured observations were employed in study I to assess the performance of CHWs in the
intervention arm while they assessed children for respiratory symptoms. Observations provide more
reliable information about events than questionnaires (Bryman, 2004) and are very useful methods
for assessment of skills and actual performance of health workers, but do not reflect the thought
processes that guide decision making in the study subjects (Kak et al., 2001). In addition, the persons
being observed may alter their behavior during the assessment and thus not reflect what they
routinely do. Alteration of behavior is increased in cases where the study subjects are removed from
their natural settings (Bryman, 2004). In this study the observations on CHWs were undertaken at a
nearby health facility.
Use of existing data
Using data that were collected by someone else, usually for another purpose (i.e., secondary data)
saves time and resources that would have otherwise been used to collect the data. Although large
amounts of data become available in a very short time, one of the major drawbacks of this method is
the completeness and comprehensiveness of the data to answer the current research question
(Boslaugh, 2007). Some data used in this thesis already existed from the HDSS database. The data in
the HDSS database are collected through six-monthly censuses of the population in the area. Data are
collected on household characteristics, births, deaths, immigrations and emigrations. Through this
data collection system every person in the population can be traced through their household and
personal identification numbers. Data were extracted from the HDSS database on some household
characteristics of CHWs (study I) and children (study IV). In study I CHWs’ records were also
reviewed.
Methods for assessment of CHW performance
Several methods have been proposed for assessment of performance of health providers, including
direct observations, record reviews, provider interviews or written exams (knowledge tests), exit
interviews with patients, case scenarios, and simulated patients (Cardemil et al., 2012; Franco et al.,
2002; Kak et al., 2001). The methods all have their strengths and drawbacks, which may influence
the decision to use them. The knowledge tests provide information about what providers know and
not necessarily what they do. However, they may be useful in assessing knowledge about rare
25
conditions (Franco et al., 2002). Case scenarios together with the knowledge tests are said to
standardize interviews, have objectivity in scoring, minimize costs of data collection, and present rare
cases of illness; however they have the disadvantage of not assessing skills well (Franco et al., 2002;
Kak et al., 2001). Reviewing records provides reliable information for treatments given and enables
fast data collection, but may provide insufficient information on diagnosis and counseling. The
quality of the records also affects the data collected (Franco et al., 2002). Information obtained from
observations by an independent person provides one of the most reliable and complete pictures of
what health providers do (Franco et al., 2002); however, it could also be influenced by alteration of
behavior due to being observed (Rowe et al., 2006). In this thesis knowledge tests, case scenarios,
record reviews, interviews of caregivers of children treated by CHWs, and to some extent direct
observations were used to assess CHW performance.
5.3.3 Sampling, sample size and data collection methods
5.3.3.1 Study I – Performance of CHWs
Quantitative data were collected from 57 of 61 CHWs in the intervention arm and 68 of 70 CHWs in
the control arm of the cRCT using questionnaires with knowledge tests and case scenarios as well as
questions on CHWs’ demographic characteristics, review of CHWs’ records, and interview of
caregivers of two children treated by CHWs in the week prior to the interview. In addition,
observations were conducted among 57 CHWs in the intervention arm while they assessed children
with respiratory symptoms. The quantitative methods and sample sizes used are summarized in Table
2.
The study aimed to recruit all CHWs that were available during the study period and who consented
to participate in the study. Since there was a fixed number of CHWs, power calculations were used to
assess the adequacy of the number of available CHWs for answering the research objective. A
formula for comparison of two proportions was used along with a two-sided 5% level of significance
and 90.5% of CHWs able to adequately treat malaria (Kelly et al., 2001). The available number of
CHWs was sufficient to detect 25% difference between the intervention and control arms with 85%
power. The interviews were conducted by six trained field assistants supervised by the principal
researcher, who also did the record reviews. The observations were conducted by two medical
officers trained in IMCI. One of the medical officers observed the CHW and filled a checklist while
the other counted the respiratory rates along with the CHW. Data on the wealth index of CHWs were
extracted from the HDSS database.
Table 2. Summary of the quantitative data sources for study I with number of participants
Data collection method
Intervention Control
N
N
Questionnaires
57
68
Review of records
57
68
Observations of respiratory symptoms assessment
57
Interview of children treated by CHWs
114
134
Extraction of data on CHW’s wealth index from HDSS database* 47
47
* Some CHWs missing wealth index data
The qualitative data were collected through four FGDs among CHWs. The FGDs were held
separately for men and women in the intervention and control arms. Each FGD had 8-10 participants
who were selected purposively to include the leaders of CHWs in each arm; CHWs with high or low
patient turn-over as seen in the review of records, with several errors in their records or
26
questionnaires filled during the quantitative data collection; and CHWs that were active in the
monthly meetings. This range of participants was thought to be broad enough to provide a diverse
range of ideas (Powell and Single, 1996). The FGDs were conducted by a skilled moderator assisted
by the principal researcher. The FGDs were conducted subsequent to quantitative data collection
after identifying areas that needed further elaboration; these included training of CHWs for their
roles, community support for the CHWs, referral, effect of CHWs’ roles on their lifestyle and factors
perceived by CHWs to influence their work.
The outcome variable was performance of CHWs, which was based on eliciting signs and symptoms,
classification of illness, identifying and responding to danger signs, prescribing medicines (dosing,
giving instructions on how medicines should be administered) and storing medicines appropriately.
5.3.3.2 Study II – Adherence to medicines
Children that were treated by CHWs were consecutively sampled from the CHWs’ registers on the
day they received treatment, and those whose homes were traced and their caregivers gave informed
consent to participate in the study were enrolled. In the control areas 667 children were enrolled
while in the intervention areas 323 children prescribed antimalarials only and 266 children prescribed
antimalarials plus antibiotics were enrolled. The minimum required sample size was estimated using
the formula for comparison of two proportions with adjustment for clustering (Adamchak et al.,
2000). The assumptions used were 5% level of significance, 80% power, design effect of two, 81%
adherence among children taking AL alone as seen in a previous study (Ajayi et al., 2008a), 20%
reduction in adherence among children taking both AL and amoxicillin giving adherence of 64.8% in
AL plus amoxicillin group, and 10% loss to follow-up. Based on these assumptions, a minimum
sample size of 258 children in each group was sufficient.
The data were collected using a questionnaire that was administered to caregivers of the enrolled
children. The questionnaire was divided into two parts. Part one of the questionnaire collected
information on demographic and illness characteristics and was administered on day one of receiving
treatment from CHWs. Part two of the questionnaire was administered on day four after receiving
treatment from the CHWs and was used to collect information on understanding of medicines
administration instructions, vomiting of administered medicines, and adherence to treatment. The day
four visits were unannounced to minimize their influence on adherence as well as to minimize
discarding of pills that may occur when patients expect their medicines to be checked (Minzi and
Naazneen, 2008).
Adherence was assessed using pill counts and caregiver reports of how the treatment was
administered over a three-day period (Osterberg and Blaschke, 2005). The caregiver report was used
to compute the percentage level of adherence (i.e. the percentage of prescribed medicines that were
taken appropriately) and was validated by checking the medicines’ packet to determine if there were
any pills and the number of pills that were left on the day of evaluation (pill counts). The computed
level of adherence was categorized into adherent (those that had taken all their prescribed drugs
appropriately) and non-adherent (those that had not taken some of their prescribed drugs or had taken
them inappropriately) (Kachur et al., 2004). The caregivers were also asked if the children had
vomited any of the doses within 30 minutes of taking the medicines and whether they had
administered replacement medicines for the vomited doses. Children that had been classified as
adherent were considered as non-adherent if they were not re-administered doses after vomiting
(Beer et al., 2009).
27
Further evaluations of adherence were made based on what the medicines were administered with or
what foods were taken before the medicines were administered. Based on this, the participants were
considered to have optimal adherence if they had taken all their medicines as prescribed and had
taken them with a fatty meal; good adherence if they had taken all their medicines as prescribed but
had either not taken the medicines consistently with a fatty meal or had not taken the medicines with
a fatty meal at all; and non-adherence if some of the medicines were not taken or were not taken
according to the correct schedule (Achan et al., 2009).
5.3.3.3 Study III – Effectiveness of integrated malaria and pneumonia management
The data for this study were the same as for study II, but a few more children were included in this
study who did not participate in study II because they received amoxicillin only from the CHWs,
which was not included in the groups to be compared on adherence. Study III included children
treated by CHWs: 609 from the intervention arm and 667 from the control arm. The required sample
size was estimated using the formula for comparison of two proportions in cRCTs (Hayes and
Bennett, 1999). The assumptions used were proportions of children receiving prompt and appropriate
pneumonia treatment of 13% in the control arm and 68% in the intervention arm, as seen in a
previous study (Yeboah-Antwi et al., 2010), 90% power, 5% level of significance, coefficient of
variation of 0.4 (estimated to fall in range of recommended values (Hayes and Bennett, 1999),
minimum of 12 children per cluster, 20% of children having pneumonia symptoms, and a drop-out
rate of 10%. Using these calculations 111 children with pneumonia symptoms were needed in each
arm, translating to about 610 children treated by CHWs in each arm.
For this study the day one data included child and caregiver demographics, illness characteristics, and
treatment seeking. In addition, the children’s temperatures were measured and their breathing rates
were counted. The children were visited again on day four of treatment seeking and the caregivers
were interviewed about additional treatment received, adherence, hospitalization, and illness
resolution. Temperature and breathing rates were measured again.
The children were classified as having pneumonia symptoms if the caregiver reported cough
accompanied by fast breathing and/or difficult breathing (Diaz et al., 2013; UNICEF, 2013b).
The primary outcome was receiving prompt and appropriate antibiotics for pneumonia symptoms and
the secondary outcome was treatment outcomes. Prompt and appropriate antibiotics was defined as
receiving the recommended antibiotic for pneumonia on the day of symptom presentation or the next
day. Treatment outcomes were defined as: having high temperature ( 37.5oC) or raised respiratory
rate ( 50 breaths per minute in children aged 4-12 months and 40 breaths per minute in children 1259 months) on day four; receiving additional antimalarials after treatment by CHW; receiving
additional antibiotics after receiving treatment from the CHWs in the intervention arm or receiving
additional antibiotics beyond those received from the health facility where referred children sought
care in the control arm; hospitalization, or death (Yeboah-Antwi et al., 2010). A child with any of
these treatment outcomes was considered to have treatment failure. Self-reported treatment failure
was also assessed by the caregiver’s report on the child’s illness resolution.
5.3.3.4 Study IV – Effect of integrated management on utilization and community drug use
patterns
Children aged 6-59 months were sampled from the HDSS database using random sampling to obtain
700 children from each arm of the cRCT. The households of the sampled children were visited and
caregivers that were available at home and gave informed consent were interviewed (548 from the
28
intervention arm, 547 from the control arm). Questionnaires were used to collect data on knowledge
of fever, malaria, and pneumonia, illness experienced in two weeks prior to the interview, health
care-seeking for illness and treatment received. In addition, data on child, caregiver and household
characteristics were extracted from the HDSS database. This data included a wealth index (in the
form of quintiles) computed using principal components analysis based on household items similar to
those used in the Uganda Demographic and Health Survey (Uganda Bureau of Statistics and Macro
International Inc., 2007), which has been described elsewhere (Waiswa et al., 2010).
Illness classifications for malaria and pneumonia were made based on the symptoms reported by the
caregiver. A child was considered as having had malaria if the caregiver reported fever in the
previous two weeks (Gove, 1997). Self-reported pneumonia symptoms were defined as caregiver
report of cough and fast breathing with or without fever; cough and difficult breathing with or
without fever; difficult and fast breathing with or without fever; and fever with difficult breathing
(UNICEF, 2013b; Department of Health Policy Planning and Management, 2010).
The treatment received was classified as appropriate if the child used the recommended drug, dose,
frequency, and duration. In addition, if the child used appropriate treatment promptly, i.e. if it was
administered on the day of onset of symptoms or the next day, (Ajayi et al., 2008b; Rutebemberwa et
al., 2009a) it was categorized as prompt and appropriate (Nsungwa-Sabiiti et al., 2007). The
assessment for appropriateness of treatment was done for the first treatment given to the child (i.e.
first treatment action) and for the second treatment given to the child if they needed further treatment
(i.e. second treatment action).
Medicines used for treatment were assessed for appropriateness based on national (Ministry of
Health (Uganda), 2010b) or CHWs’ treatment guidelines (Department of Health Policy Planning and
Management, 2010) or treatment recommendations of the British National Formulary (Joint
Formulary Committee, 2007), a widely used reference in Uganda.
The KIIs involved health service providers found in the facilities within the area including health
centres, drug shops and private clinics. The purpose of these KIIs was mainly to explore malaria and
pneumonia treatment practices by health workers within the area. The findings from the KIIs were
not included in the publication of study IV but have been included in this thesis.
5.3.4 Data management
The FGDs and KIIs involved note-taking in the field as well as tape recording. The notes were used
mainly to record non-verbal communication while the tape recording allowed an accurate account of
the discussions to be taken (Liamputtong and Ezzy, 2005). The notes were taken by the assistant
moderator and this eased the moderator’s demanding task of conducting FGDs, allowing her to
concentrate on moderation of the discussion. The FGDs with CHWs were recorded in Lusoga and
then translated into English during transcription by two fluent Lusoga speakers. All the KIIs were
recorded and transcribed in English.
The questionnaires used in all four studies were field edited for errors by members of the study team
and corrections were made while still in the field. A further check for errors was done by the
principal investigator and the data were corrected the next day. The data from questionnaires were
entered into FoxPro computer software and then exported to STATA version 10 for statistical
analysis.
29
The observations made during study I were recorded on an observation checklist by medical officers
with IMCI training. The data from the record reviews were abstracted using a data extraction form.
They were then entered into EPIDATA and subsequently exported to STATA 10 for analysis.
5.3.5 Data analysis
Content analysis: The data from the FGDs in study I and KIIs in study IV were analyzed using
manifest content analysis. This analysis focused on what the text said, thereby dealing with the
content aspect, and described the visible, obvious components (Graneheim and Lundman, 2004;
Downe-Wamboldt, 1992). The units of analysis were the transcripts from the FGDs and KIIs. The
transcripts were read through several times, and meaning units from which codes were generated
were identified. The codes were subsequently grouped into categories.
Descriptive statistics were computed in all four studies for general description of the participants
and performance of CHWs in study I.
Performance of CHWs: Scores on performance of CHWs were generated for various dimensions
including knowledge of malaria or pneumonia, eliciting signs and symptoms for malaria or
pneumonia, prescribing for malaria or pneumonia and giving instructions on how the medicines
should be administered. Since these dimensions were assessed with different numbers of questions or
had different numbers of possible responses, percent scores were obtained for each of the questions.
The percent scores were summarized as means to give mean percentage scores on each of the
dimensions assessed for each CHW (Kumar et al., 2009; Venkatachalam et al., 2011). Performance
scores were also generated using principal components analysis (PCA) and they were compared with
the mean percentages. Because the percentage scores are easier to interpret and compare between
studies (Streiner and Norman, 2003), a decision was made to present them after finding fair to high
correlations (0.64-0.88) between them and the scores generated using PCA. The mean percentages
for each dimension were then summarized separately for CHWs in the intervention or control arm as
means or medians.
Comparative analysis: t-tests or Mann-Whitney U tests were used to compare numerical
characteristics between the intervention and control arms in study I, II and IV, as well as scores for
performance on malaria (study I). These tests are used to compare numerical characteristics between
two groups. Performance on malaria and pneumonia in the intervention arm was compared using the
Wilcoxon signed rank tests (study I). These tests compare paired numerical characteristics that are
not normally distributed. In study II the three treatment groups were compared using analysis of
variance (ANOVA) or Kruskal-Wallis tests. These tests compare three or more groups on numerical
characteristics. Pairwise comparisons between the groups following ANOVA or Kruskal-Wallis tests
were conducted using t-tests or Mann-Whitney U tests with adjustments of levels of significance
using the Bonferroni correction. Chi-squared or Fisher’s exact tests were used to compare categorical
variables between groups in study I, II and IV.
In study III study outcomes were compared in the intervention and control arms and unadjusted
relative risks were estimated using Mantel-Haenszel methods, which considered the strata used in the
randomization.
Multivariable analysis: Logistic regression was used in study II to determine the factors associated
with non-adherence; and in study IV to determine the factors associated with the utilization of CHWs
and appropriateness of treatment. Adjusted and unadjusted analyses were done. Factors that had p-
30
values less than 0.2 at unadjusted analysis were considered for multivariable analysis. All analyses in
study II and III were adjusted for clustering.
In order to simultaneously control for the effect of covariates in study III, both cluster level and
individual level analyses were done. The cluster level analysis was the recommended level of
analysis since there were fewer than 15 clusters in each arm (Hayes and Moulton, 2009). However in
order to simultaneously account and present results for other covariates, the individual level analysis
was presented after checking that the conclusions drawn regarding the intervention effect were not
different from those drawn after cluster level analysis. The individual level analysis was done using
generalized estimating equations (GEE), while the cluster level analysis was done by estimating rates
of the outcome in each cluster and then comparing the mean rates in the intervention and control arm
using t-tests.
5.3.6 Summary of methods
The methods of the four studies in this thesis are summarized in Table 3.
Table 3. Summary of the methods used in the four studies
Study
Study design
Study
Data analysis
population &
sample size
Performance Mixed methods:
CHWs (4
Descriptives, Chiof CHWs (I) FGDs
FGDs, 125
squared or Fisher’s
Questionnaires
interviews, 57
exact tests, Mann(knowledge tests, observations)
Whitney U tests,
case scenarios)
Wilcoxon signed
Record reviews
Caregivers of
rank tests, manifest
Observations in
children treated content analysis
intervention arm by CHWs (248)
Adherence to Cohort within
Children
Descriptives,
medicines
cRCT
treated by
ANOVA or Kruskal(II)
CHWs and
Wallis, Chi-squared
their caregivers or Fisher’s tests, t(1256)
tests or MannWhitney U test,
logistic regression
Effectiveness Cohort within
Children
Descriptives, Mantelof integrated cRCT
treated by
Haenszel methods,
malaria &
CHWs and
GEE (individual
pneumonia
their caregivers level analysis), t-tests
management
(1276)
(cluster level
(III)
analysis)
Utilization of Cross-sectional
1095 children
Descriptives, ChiCHWs and
with KIIs
under five and
squared or Fisher’s
effect on
their caregivers tests, logistic
drug use
regression
patterns (IV)
13 KIs from
health facilities Content analysis
31
Period of data
collection
June 2011
October to
November
2011
October to
November
2011
January to
February 2011
5.3.7 Ethical Issues
Initial ethical approval and continuing review for all sub-studies were obtained from the School of
Public Health Higher Degrees Research and Ethics Committee (IRB00005876) and the Uganda
National Council of Science and Technology (HS898). In addition, sub-study IV received ethical
approval from the Regional Ethics Committee of Karolinska Institutet, Sweden (Dnr 2011/167931/4).
Written informed consent was obtained from each participant involved in the quantitative studies
while verbal consent was obtained from the participants in the FGDs. The participants also consented
to the use of a tape-recorder.
Permission to conduct the studies was also obtained from the local administration of the villages.
Confidentiality was maintained throughout data collection, management and analysis.
Potential ethical risks from this study included the possibility of creating misunderstandings between
CHWs and the community they serve, because the CHWs would know when people from their
villages were enrolled into the study. However, attempts were made to mitigate this by maintaining
confidentiality of the information given by the caregivers.
32
6 RESULTS
The results are presented in line with the modified access framework (Figure 3). They include
utilization of services (study IV), quality of care (studies I, II, III and IV), health status (study III),
patient satisfaction (studies I, III and IV), equity (study IV), and influence of livelihood assets and
other factors on utilization of services and quality of care (studies II, III and IV).
6.1 UTILIZATION OF SERVICES (STUDY IV)
Children that were ill in the two weeks prior to the interview were assessed for promptness of careseeking, and the source of first and second treatment outside the home.
Promptness of care-seeking: About 94% (689/732) of caregivers that sought care outside the home
(from any type of health provider) did so promptly: 93% (354/382) in the intervention arm, and 96%
(335/350) in the control; p=0.11.
Sources of care: For the first source of treatment outside the home, about 26% of caregivers sought
care for their children from CHWs and this was higher in the intervention (29%) compared with the
control arm (21%; p=0.01). Overall, caregivers mostly sought treatment from the private sector (32%
private clinics, 25% drug shops) (Figure 6).
35
Overall
30
Intervention
25
Control
20
%
15
10
5
Traditional
healer
(p=0.95)
General
shop
(p=0.61)
NGO unit
(p=0.21)
G ov ernment
Unit
(p=0.07)
Drug shops
(p=0.26)
Private
clinics
(p=0.97)
CHWs
(p=0.01)
0
Source of first tre atment
Figure 6. First source of care among children in intervention and control areas
When the second source of treatment outside the home was considered, the utilization of CHWs
increased to 31% in the intervention arm and 22% in the control arm (adjusted OR=1.60, 95%
CI=1.09-2.35; adjusted for wealth index, knowledge of malaria prevention, knowledge of danger
signs, and child’s symptoms, i.e. not having fever).
33
6.2 QUALITY OF CARE
This section includes the performance of CHWs in management of malaria and pneumonia and
factors perceived by the CHWs to influence performance; community drug use for malaria and
pneumonia; and adherence to medicines.
Performance of CHWs (studies I and IV)
It was hypothesized that addition of pneumonia management to malaria management would lower
performance on malaria and that CHWs managing pneumonia would find this a challenge because of
the more complex diagnostic algorithm and additional medicines to prescribe. Performance of CHWs
on knowledge of malaria, eliciting signs and symptoms, and prescribing was compared between the
intervention and control arms (study I). In addition, performance of CHWs on pneumonia
management was determined in the intervention arm and compared with malaria management. Drug
use for malaria symptoms was compared between children treated by CHWs and those treated by
other health providers (study IV). Drug use for self-reported pneumonia symptoms was compared
between children treated by CHWs and those treated by other health providers in the intervention
arm.
CHWs’ performance on malaria management
CHWs in the intervention and control arms had similar scores on knowledge of malaria (intervention
72%, control 70%), eliciting malaria signs and symptoms (50% both groups), prescribing
antimalarials based on case scenarios (80% both groups), and prescribing antimalarials based on
records (99% intervention, 100% control) (study I). Table 4 summarizes findings on performance on
malaria in study I.
Table 4. CHWs’ performance on malaria in study I
Percentage score
Intervention
Knowledge of malaria, median
72%
Eliciting signs and symptoms, median
50%
Prescribing antimalarials, case scenarios, median 80%
Prescribing antimalarials, record reviews, mean 99%
Control
70%
50%
80%
100%
p-value
0.37
0.74
0.70
0.10
Comparison of children treated by CHWs with those treated by all other health providers (study IV)
showed that a higher proportion of children treated by CHWs (37%) received prompt and appropriate
treatment of malaria symptoms than those treated by other health providers (9%; p<0.001). The
children treated by CHWs were more likely to receive the recommended drug (100% vs 38%;
p<0.001), recommended drug and dose (71% vs 22%; p<0.001), recommended drug and frequency
of administration (78% vs 22%; p<0.001) and recommended drug for recommended duration (79%
vs 21%; p<0.001).
CHWs’ performance on pneumonia management in the intervention arm
Assessment of CHWs in the intervention arm in study I showed median overall pneumonia
knowledge of 40% from questionnaires, and median scores on prescribing of 82% from case
scenarios and 96% from record reviews. The mean percentage of children with fast breathing that
received antibiotics was 82% and that of children without fast breathing that received antibiotics was
12%. About 49% and 35% of CHWs were observed to count respiratory rates within five breaths and
two breaths respectively of those counted by the physician, and 75% of them correctly classified
children as having pneumonia or not (Table 5). This level of correct classification of pneumonia
status was observed even though there was a lot of discordance between the breath counts made by
the CHWs and physicians, because the breath counts of many children were far from the cut-off
34
points such that gross over- or under-estimation by CHWs would not change the “pneumonia”
classification of the child.
Table 5. Performance of CHWs in intervention arm on pneumonia in study I
Source
Score
40%
Questionnaires Overall knowledge, median score
Eliciting signs & symptoms (integrated)*, median score
25%
Prescribing from case scenarios (integrated)*, median score 82%
With fast breathing received antibiotics, mean percent
82%
Records
Without fast breathing received antibiotics, mean percent
12%
Correct doses of antibiotics, mean score
96%
Counted respiratory rates ±5 breaths of physician
49%
Observations
Counted respiratory rates ±2 breaths of physician
35%
Correctly classified children as ± pneumonia
75%
Misclassified children as having pneumonia
14%
Misclassified children as not having pneumonia
4%
Could not decide if child had pneumonia or not
7%
Showed correct position for evaluation of chest in-drawing 88%
* Integrated refers to combining malaria and pneumonia.
Comparison of malaria (Table 4) and pneumonia management (Table 5) in the intervention arm,
showed higher scores for malaria than for pneumonia on overall knowledge (72% vs 40%; p<0.001)
and correct doses of medicines from record reviews (99% vs 96%; p<0.001). The performance on
eliciting signs and symptoms from case scenarios fell from 50% for malaria to 25% when pneumonia
signs and symptoms were integrated (study I).
When CHWs were compared with other health providers on pneumonia (study IV), the proportion of
children that received an antibiotic for self-reported pneumonia symptoms did not differ between the
children treated by CHWs and those treated by other health providers (42% vs 45%; p=0.39). There
was also no statistically significant difference in the proportion that received prompt and appropriate
treatment of pneumonia symptoms (CHW -users 7%, non-CHW -users 13%; p=0.18).
Factors perceived to influence CHW performance (study I)
The factors perceived by CHWs as influencing performance were grouped into community-,
programme-, and health facility-related factors. The community factors included mobilization of
communities to seek care from CHWs by local leaders, community members seeking care from
CHWs because of their confidence in medicines used (which mainly arose from finding CHWs with
medicines that were similar to medicines children would receive from public health facilities), and
caregivers’ non-compliance with referral:
“The LCs [local councils] [local leaders] also play very important roles here. They tell people to
come for treatment from me and as a result they come whenever the children are sick.” (FGD with
female CHWs, control arm)
35
There was also lack of community appreciation for age restrictions of children treated. The caregivers
usually wanted children older than five years also to be treated. This finding was also reflected in the
interviews with caregivers of children treated by CHWs:
“Sometimes the people say that it seems these health workers want to kill our children with those
drugs reason being that why are we very strict on who we treat? They said that it would be good if
everyone is treated. If possible we could increase the age group to 7 years.” (FGD with female
CHWs, control arm)
Programme factors included reinforcement of knowledge through continued training during monthly
meetings, interrupted supply of medicines, using safe medicines, transport refund for collecting
medicines, difficulty of working at night and during rainy weather, lack of transport for follow-up of
treated children, and large coverage areas making follow-up difficult:
“The three days [of initial training] weren’t enough for us because we trained to treat malaria and
pneumonia at the same time. When we were trained to treat pneumonia, it was a bit difficult for us so
what they decided to do was to put monthly meetings where we meet as CHWs and we share
experiences on how we are handling situations in our villages.” (FGD with female CHWs,
intervention arm)
“When we don’t have drugs, everything goes down. Also transport to do follow up of the children
that you treated is really hard. Sometimes you even use your own personal money to do follow up.”
(FGD with female CHWs, control arm)
The health facility-related factor influencing performance was the lack of cooperation from health
workers at facilities:
“There are times when you refer a person to the health centre but when this patient reaches there,
he/she doesn’t get the needed attention. They say that they look at the referral note over and over
again instead of treating the patient.” (FGD with female CHWs, control arm)
Community drug use for malaria and pneumonia in intervention and control arms (studies III
& IV)
Overall medicines use (IV)
There was no difference in the mean number of medicines received by children in the intervention
and control arms (mean=2, SD=1.1 in both arms). However, the mean number of medicines used was
lower among children treated by CHWs in both the intervention (1.6 vs 2.4; p<0.001) and control
arms (1.4 vs 2.3; p<0.001) compared to children treated by other health providers.
Medicines use for malaria symptoms (study IV)
About 62% of the children in the intervention arm and 59% in the control arm with fever received
antimalarials. However, 54% in the intervention arm and 49% in the control arm received
recommended antimalarials. Overall, 18% of the children with fever in the intervention arm received
prompt and appropriate treatment for malaria symptoms compared with 12% in the control arm
(p=0.03).
Malaria management practices by non-CHW health providers
Through KIIs in study IV, exploration of medicines used for treatment of malaria revealed that some
of the KIs use artemether-lumefantrine to treat malaria while several others still use the old
antimalarials that are no longer recommended. In particular, some health providers said they use
36
chloroquine as a first-line treatment for malaria alone while others follow it with sulphadoxinepyrimethamine, and they use quinine in case the treatment with chloroquine fails:
“I treat according to the age of the child let’s say if a mother brings a child at the age between two
and four months I can use chloroquine syrup, I don’t have tablets to give children at that age. And if
they are five months and above I can give chloroquine or quinine tablets. If the child completes the
dose of quinine or chloroquine I accompany with Fansidar® [sulphadoxine-pyrimethamine]
according to the age and body weight.” (KI, nursing assistant, private health facility)
A few KIs mentioned that they use quinine in cases of severe malaria or as a second line in case the
first line treatment failed to work:
“Okay, in the facility normally they are treated with the first line of Coartem® [artemetherlumefantrine], if they don’t respond they are put on the second-line quinine.” (KI, health worker,
public health facility)
One KI said that she can use sulphadoxine-pyrimethamine alone for treatment of fever depending on
how the child presented:
“If a child has fever, normally I treat using Lornat® [artemether-lumefantrine] tablets or Fansidar®
[sulphadoxine-pyrimethamine] or quinine basing [it] on how the child has presented.” (KI, nurse,
private facility)
In addition to antimalarials, many KIs said they use other drugs including antipyretics, hematinics
and anti-emetics, depending on the symptoms the child presented with or if the caregiver could afford
the drugs. Some KIs said they use dexamethasone as an antipyretic while others use it to boost the
child’s strength, as shown by these quotes below:
“I do give it in conditions like high temperature. I mix chloroquine tablets and Fansidar®. I also give
dexa [dexamethasone].” (KI, nursing assistant, drug shop)
“You will look at this child and she or he will be very weak and without energy so in order for this
child to get energy, I add dexa [dexamethasone] in the treatment I give her/him.” (KI, nursing
assistant, drug shop)
Medicines use for pneumonia symptoms
Assessment of use of antibiotics for pneumonia symptoms in the community regardless of the source
showed that the proportion of children with self-reported pneumonia symptoms that received
antibiotics was not significantly different in the intervention and control arms (45% vs 37%; p=0.44).
About 11% of the children in the intervention arm and 5% in the control arm (p=0.10) received
prompt and appropriate treatment of pneumonia symptoms (study IV).
In study III, assessment of use of antibiotics for pneumonia symptoms focused on children that were
treated by CHWs in the intervention and control arms. About 63% of the children in the intervention
arm with self-reported pneumonia symptoms received antibiotics from the CHWs, while in the
control arm 51% of such children were referred to health facilities for further care. The proportion of
children that received prompt and appropriate antibiotics for self-reported pneumonia symptoms was
higher in the intervention arm (45%) than in the control arm (10%), with a relative risk of 3.51 (95%
CI = 1.75-7.03) after adjusting for education of caregiver, urban or rural residence, and child’s
presenting symptoms.
37
Pneumonia management practices by non-CHW health providers
Most of KIs cited cotrimoxazole as the drug used in pneumonia treatment. Some of them used it
alone as the first drug while others used it as a continuation drug after giving injections. Benzyl
penicillin injections were cited for the treatment of pneumonia and were usually followed with oral
drugs such as cotrimoxazole or amoxicillin owing to the difficulty of continuing injections especially
for health facilities that do not admit patients. Other KIs said they use oral drugs only, e.g.
amoxicillin or ampicillin plus cloxacillin.
Another KI mentioned that when patients get to his health facility, they have usually been treated
with many drugs such as cotrimoxazole and amoxicillin so he treats pneumonia with amoxicillin plus
clavulanate. Chloramphenicol was also cited as medicine commonly used for pneumonia.
From the quantitative data collection in study IV, some children had received medicines that were
crushed and mixed by the health provider, with no labels showing the constituents. KIIs revealed that
these mixtures were prepared by the health providers and the constituents varied depending on the
illness. The quote below shows the typical mixture that was prepared for a 3-year-old child with
pneumonia:
“We make the whole mixture once and it’s the mother to get a small portion of the mixture on the
spoon and give the child for about 3 days and gives twice a day ....we mix 2 capsules of amoxicillin
[total of 500mg amoxicillin] with 1 Panadol® [paracetamol 500mg] and 1 prednisone [5mg].” (KI,
drug shop)
Adherence to medicines (study II)
Adherence to medicines in the intervention arm did not differ among children taking antimalarials
only (mean adherence 99.2%) and those taking a combination of antimalarials and antibiotics (mean
adherence 98.5%). However, adherence to medicines in the intervention arm was higher than in the
control arm (antimalarials only) (mean 96%; p<0.001). Although the proportion of children that were
adherent to their medicines (i.e. took all medicines as recommended) was high, most of them did not
take the artemether-lumefantrine with a fatty meal as recommended by the manufacturers (i.e.
optimal adherence was much lower) (Figure 7). In addition, some of the children that vomited within
30 minutes of taking the medicines did not receive replacement doses, and this also lowered
adherence.
The reasons given for missing medicines or not giving them according to the prescribed schedule
included forgetfulness (38.1%), the child improved or recovered (14.3%), vomiting (12.7%), the
tablets were too many (11.1%), not having food or drink to give the child (4.8%), not understanding
the medicines administration instructions well (4.8%), and other reasons such as caregiver being
away from home, the child experiencing adverse reactions, and having other medicines to take.
38
Adherent
120%
Percent adherent
100%
86% 83%
96% 95%
94%
92%
80%
Optimal adherence
(taken with fatty meal)
60%
40%
20%
Adherent (vomiting
considered)
18%
26%
23%
0%
AM (control)
AM (intervention)
AM + AB
(intervention)
Treatment group
Figure 7. Non-adherence to medicines among children treated by CHWs
6.3 HEALTH STATUS (study III)
There was no difference in overall treatment failure between the intervention and control arms.
However, the proportion of children with temperature 37.5oC on day four since seeking care from
CHWs was lower in the intervention arm (1%) compared with the control arm (4%). The reduction in
fast breathing from day one to day four was higher in the intervention (9%) compared with the
control arm (4%) (Table 6).
Table 6. Treatment outcomes of children treated by CHWs
Characteristic
Intervention Control
RR (95% CI)
All children treated by CHWs
N=609
N=667
Overall treatment failure
90 (14.8%) 101 (15.1%)
0.88 (0.67-1.18)
Received additional* antibiotics
7 (1.2%)
13 (2.0%)
0.77 (0.32-1.86)
Received additional**
4 (0.7%)
10 (1.5%)
0.44 (0.15-1.28)
antimalarials
Hospitalization
10 (1.6%)
9 (1.4%)
1.47 (0.64-3.38)
Temperature 37.5oC on day 4
7 (1.2%)
23 (3.5%)
0.29 (0.11-0.78)
Had fast breathing on day 4
66 (10.8%)
48 (7.2%)
1.29 (0.87-1.92)
Difference in fast breathing,
9.2%
4.2%
p-value=0.01
day 1 & 4†
Deaths
Self-reported treatment failure
83 (13.7%) 170 (25.5%)
0.58 (0.46-0.74)
* Additional antibiotics received after CHW treatment in intervention arm or after treatment from other
source where children were referred in control arm.
** Additional antimalarials received after treatment by CHWs.
† Difference between percentage of children with fast breathing on day one and four,
p-value comparing difference in intervention and control.
39
6.4 PATIENT SATISFACTION (STUDIES I, III AND IV)
Caregivers rated the care provided by CHWs highly in all studies. In study I, 90% (103/113) in the
intervention arm and 84% (113/135) of caregivers in the control arm rated the quality as good, with
only one (0.7%) caregiver in the control arm rating care as poor. In study III, 99.5% (605/608) of
caregivers in the intervention arm rated the care from CHWs as good and 0.5% (3/608) rated it as fair
while in the control arm 97% (657/667) rated the care as good, 2.8% (19/667) rated the care as fair,
and only one caregiver (0.2%) rated the care received from CHWs as poor. In study IV, some
caregivers of children seen in government facilities (6/125, 4.8%), private clinics (3/231, 1.3%) and
drug shops (1/180, 0.6%) rated the quality of care as poor while no caregivers rated the CHWs’ care
as poor.
6.5 EQUITY (STUDY IV)
Utilization of CHW services was compared across the five wealth quintiles in study IV. There was a
tendency for more children in the lower four wealth quintiles to seek care from CHWs compared
with the highest quintile (p=0.05) (Figure 8).
35%
30%
30%
30%
27%
26%
25%
%
20%
18%
15%
10%
5%
0%
Poorest
Poorer
Poor
Less Poor
Least Poor
Figure 8. Utilization of CHWs according to wealth index among children in
Iganga-Mayuge HDSS
40
6.6 INFLUENCE OF LIVELIHOOD ASSETS ON UTILIZATION OF SERVICES
AND QUALITY OF CARE
Some livelihood assets including human capital such as post-primary education, increasing
knowledge of malaria, pneumonia, and danger signs and caregivers’ perception that illness was
severe and financial capital, such as the wealth index, were associated with different outcomes in
various studies (studies II, III and IV) (Table 7). Children whose caregivers had attained post-primary
education were more likely to receive prompt and appropriate antibiotics (study III), while increasing
knowledge of malaria prevention was associated with utilization of CHWs (study IV). Knowledge of
how malaria is transmitted, knowledge of danger signs, and knowledge of pneumonia signs were
associated with receiving prompt and appropriate malaria treatment (study IV). Non-adherence was
higher among children whose caregivers perceived the illness as not severe and caregivers who did
not understand the medicines administration instructions (study II). Poorer children were more likely
to utilize the services of CHWs (study IV).
Table 7. Influence of livelihood assets on utilization of CHWs' services and quality of care
Variable Predictor
Outcome
Measure of
grouping
association
Human
Post primary education
Prompt and appropriate RR=1.56,
capital
antibiotics (III)
CI=1.04-2.36
Knowledge of malaria
Utilization of CHWs
OR=1.39,
prevention
(IV)
CI=1.12-1.73
Knowledge of malaria
Prompt and appropriate OR=2.05,
transmission
malaria treatment (IV) CI=1.23-3.41
Knowledge of danger
Prompt and appropriate OR=1.22,
signs
malaria treatment (IV) CI=1.03-1.43
Knowledge of
Prompt and appropriate OR=1.40,
pneumonia signs
malaria treatment (IV) CI=1.14-1.72
Perception that illness
Non-adherence (II)
OR=2.0,
was not severe
CI=1.1-3.8
Not understanding
Non-adherence (II)
OR=24.5,
medicines use
CI=2.7-224.5
instructions
Financial Wealth index (poorest to Utilization of CHWs
OR=1.92,
capital
less poor vs least poor)
(IV)
CI=0.99-3.71
RR = relative risk, OR = odds ratio, CI = confidence interval
41
6.7 OTHER FACTORS ASSOCIATED WITH UTILIZATION OF CHWS
AND QUALITY OF CARE
Other factors that were associated with utilization of CHWs’ services (study IV), adherence (study
II), and receiving prompt and appropriate malaria and pneumonia treatment (studies III and IV)
included presenting symptoms of the child’s illness, and age of the household head and child as well
as residence (Table 8). Children that had no fever were more likely to utilize the services of CHWs
(study IV) and not adhere to medicines (study II) compared to children that had fever. Children that
were vomiting were more likely to be non-adherent (study II), while those that had no diarrhoea were
more likely to receive prompt and appropriate antibiotics (study III). Increasing age of the household
head was associated with the child receiving prompt and appropriate treatment for pneumonia
symptoms (study IV) and for malaria symptoms (study IV). In addition, younger children were less
likely to receive prompt and appropriate malaria treatment (study IV). Children that were staying in
urban areas were more likely to receive prompt and appropriate antibiotics (study III).
Table 8. Influence of presenting symptoms and age on CHW utilization and quality of care
Variable Predictor
Outcome
Measure of
grouping
association
Presenting No fever
Utilization of CHWs
OR=2.55, CI=1.26symptoms
(IV)
5.16
No fever
Non-adherence (II)
OR=3.3, CI=1.6-6.9
Vomiting
Non-adherence (II)
OR=2.6, CI=1.2-5.5
No diarrhoea
Prompt and appropriate RR=1.33, CI=1.01antibiotics for
1.76
pneumonia symptoms
(III)
Other
Age of
Prompt and appropriate OR=1.05, CI=1.01factors
household head pneumonia treatment
1.08
(IV)
Urban residence Prompt and appropriate RR=1.56 (1.20-2.00)
antibiotics for
pneumonia symptoms
(III)
Age of
Prompt and appropriate OR=1.02, CI=1.00household head malaria treatment (IV)
1.04
Age of the child Prompt and appropriate OR=0.98, CI=0.96malaria treatment (IV)
0.99
Seeking care
Non-adherence (II)
OR=2.2, CI=1.3-3.7
after 2 days
42
7 DISCUSSION
7.1 MAIN FINDINGS
The additional responsibility of pneumonia management by CHWs did not affect their performance
on malaria management. The good performance of CHWs on malaria was reflected in the community
drug use patterns where children treated by CHWs had more rational drug use for malaria compared
with children treated by other health providers. However, the CHWs had some difficulties in
assessment of pneumonia symptoms. Care-seeking from CHWs that provide integrated malaria and
pneumonia treatment increased prompt and appropriate antibiotic use for pneumonia symptoms
compared with care-seeking from CHWs that treated malaria alone and referred children with
pneumonia symptoms to health facilities. Care-seeking for childhood illnesses was, however, mostly
from private facilities (clinics and drugs shops), although utilization of CHWs was higher in areas
where CHWs could treat both malaria and pneumonia and among children of lower socio-economic
status. High adherence was observed in both children who received antimalarials only from CHWs
and those who received combinations of antimalarials and antibiotics.
The key findings are presented and discussed in relation to the Obrist Access Framework, which has
been adapted in this thesis (Figure 3), and are sequenced as follows utilization of CHWs; quality of
care including performance of CHWs, community drug use for malaria and pneumonia management,
and adherence; health status; patient satisfaction, equity; resistance; and influence of livelihood assets
on utilization of CHWs and quality of care. This section is followed by a discussion of health systems
and methodological considerations in this thesis.
Utilization of services
The proportion of children that sought treatment from CHWs (26%) was lower than what was seen in
another study in eastern Uganda (40%) (Mukanga et al., 2012a) and one in Ghana (59%) (Ajayi et
al., 2008b), but was higher than that in a study in the same area where treatment seeking from CHWs
was only 2% (Rutebemberwa et al., 2009b). The higher treatment seeking from CHWs in Ghana
could result from familiarity of the community with the CHWs, since they had previously performed
this role. In contrast, the previous study in the same area by Rutebemberwa et al. was done at the
time when the malaria policy in Uganda had changed from use of CQ-SP to ACTs and the CHWs at
the time did not have ACTs.
The relatively low care-seeking from CHWs reported in this thesis is possibly due to the high
competition for market share between the CHWs and other health providers especially the private
clinics and drug shops which are numerous (n=122) within the area and provide a wider range of
drugs than the CHWs. Most children received treatment from these private clinics and drug shops.
The caregivers may prefer to seek care from these health providers, who can treat a wide range of
illnesses as well as treat all age groups including adults and children. In addition, the interrupted
supply of medicines that CHWs experienced from time to time may have affected treatment seeking
from them. Low care-seeking from CHWs implies that positive effects of the integrated community
illness management may not have sufficient impact in the community. Quality of care received by the
children may be lowered by seeking care from sources with inappropriate practices.
Some inappropriate malaria and pneumonia treatment practices were noted among some health
providers in the area, including the use of drugs that are no longer recommended, inappropriate use
of steroids, and crushing medicines together before dispensing to the caregivers, which would make
the measurement of appropriate dosages difficult.
43
Care-seeking from CHWs was higher in areas where CHWs provided integrated malaria and
pneumonia treatment, suggesting that increasing the roles of CHWs may increase the likelihood of
caregivers seeking care from them. Similar findings were noted in Cambodia (Hasegawa et al.,
2013). Having CHWs that treat only one illness has been identified as a limitation to their utilization,
because children often suffer from multiple illnesses for which they need a holistic intervention
(Nsabagasani et al., 2007; Nsungwa-Sabiiti et al., 2007; Källander et al., 2004).
Quality of care
Performance of CHWs on malaria: The CHWs managing malaria alone and those managing
malaria and pneumonia performed equally well on malaria management, suggesting that CHWs’
additional roles may not affect established knowledge and skills negatively. The findings could be
due to the long familiarity with malaria in the general population in Uganda (Batega, 2004). In
addition, there has been some experience in using CHWs to manage malaria under the HBMF
strategy in Uganda (Uganda Ministry of Health, n.d.). CHWs have been found to be effective in
managing malaria in previous studies (Kamal-Yanni et al., 2012). The performance of the CHWs in
prescription of medicines was higher than that in eliciting signs and symptoms and overall
knowledge of malaria suggesting that the major weakness for CHWs is malaria diagnosis as opposed
to prescription. The use of pre-packaged medicines by CHWs may have improved their prescription
practices, since these have been reported to improve drug use (Yeboah-Antwi et al., 2001). The
performance scores obtained by the CHWs on malaria were comparable to those from a study done
in Kenya (Kelly et al., 2001) and Rwanda (MoH Republic of Rwanda, 2009). Contrary to the current
recommendation by WHO for universal parasite-based diagnosis of malaria (WHO, 2010a), RDTs
were not used in the cRCT on which this thesis is based because the trial commenced before the
recommendation was made. It is not clear what impact the RDTs would have had on the performance
of CHWs. Some studies have suggested that RDTs minimize over-use of antimalarials (YeboahAntwi et al., 2010; Mukanga et al., 2012b).
The CHWs’ performance on malaria was reflected at community level, where a higher proportion of
children treated by CHWs (37%) compared with those treated by other health providers (9%)
received prompt and appropriate malaria treatment, similar to a study in western Uganda (NsungwaSabiiti et al., 2007). The other source of treatment was mainly the private sector, including drug
shops and clinics which are commonly manned by unqualified personnel (Ministry of Health
(Uganda), 2009b). The better malaria treatment practices among CHWs compared with other health
providers, coupled with the increased use of CHWs in the intervention compared with the control
arm, could explain why the intervention arm (18%) had a higher proportion of children that received
prompt and appropriate treatment for malaria than the control arm (12%).
However, the proportion of children with prompt and appropriate malaria treatment was much lower
than the WHO targets of 80% (Roll Back Malaria, 2005) but this malaria classification was based on
symptoms and not parasite-based diagnosis as recommended (WHO, 2010a). Thus, many of the
children may not have had malaria and therefore it was not desirable to have a 100% treatment rate.
However, some studies done in high malaria transmission areas of Uganda similar to the HDSS have
found high RDT-positivity (79-88%) in febrile children (Mukanga et al., 2012b; Batwala et al.,
2011), implying that many of these febrile children seen in Iganga-Mayuge HDSS would probably
have been positive on RDTs. In addition, the proportion of children with prompt and appropriate
treatment of malaria symptoms could have been under-estimated due to misreporting of the dosing
schedule by the caregivers. The data was collected on treatment of illnesses experienced over a
period of two weeks prior to the interview and the caregivers could have forgotten how the medicine
44
was given. CHWs dispense complete doses of pre-packaged medicines and therefore higher
proportions of children treated correctly are expected.
Performance of CHWs on pneumonia: The CHWs’ performance on management of pneumonia was
lower than that for malaria, with lower overall knowledge (70% vs 40%), prescribing from record
reviews (100% vs 96%) and performance on eliciting signs and symptoms, which fell from 50% to
25% when pneumonia assessment was added. Pneumonia assessment could be particularly difficult
since it involves a complex algorithm with counting and categorization of breathing rates, a task of
some difficulty even among formal health providers (Gadomski et al., 1993). There may also be less
familiarity with pneumonia compared with malaria.
In Uganda, previous efforts at illness control in children at community level have focused largely on
malaria. The use of CHWs to manage pneumonia has been mainly through small pilot studies. The
main areas of weakness for the CHWs were in assessment and classification of pneumonia symptoms
as opposed to prescribing medicines, a trend similar to that for malaria. This difficulty in assessment
of pneumonia was further shown by challenges exhibited in counting breathing rates. Only 35% of
CHWs counted breathing rates within two breaths of those counted by the physician, similar to the
proportion observed in Malawi (39%) (Gilroy et al., 2012). However, the proportion of CHWs that
counted breathing rates within five breaths of those counted by the gold standard in the current study
(49%) was lower than the 71% found in another study in Uganda (Källander et al., 2006a). The
difference in findings could be explained by differences in lapse of time between training and CHW
assessment. The latter study was done within two weeks of CHW training while the current study
was done about two years after initial training. Other studies have also reported difficulties in
respiratory assessments among CHWs (Mukanga et al., 2011; Kelly et al., 2001). It is important that
efforts are made towards improving diagnosis of pneumonia.
There was no significant difference in the proportion of children that received prompt and
appropriate treatment for pneumonia symptoms when children treated by CHWs were compared with
children treated by other health providers, and neither was there a difference between the intervention
and control arms. This may have resulted from insufficient power to detect differences or from
insufficient impact of the intervention on pneumonia treatment practices in the community. Contrary
to the expectation that CHWs would over-use antibiotics, the proportion of children that received
antibiotics did not differ significantly among children treated by CHWs and those treated by other
health providers. However, it is important to ensure that health providers are not missing pneumonia.
When comparisons were focused on children treated by CHWs in the intervention and control arms, a
three-fold difference was found in proportions receiving prompt and appropriate antibiotics in the
two arms, suggesting that CHWs who provide integrated malaria and pneumonia care increase the
promptness of appropriate antibiotics compared with CHWs who treat malaria only and refer
children with pneumonia symptoms to health facilities. These findings are similar to those in a study
in Zambia (Yeboah-Antwi et al., 2010).
Several methods were used to assess the performance of CHWs, but although some results are not
conclusive on pneumonia, some useful information was obtained including that CHWs perform
highly on prescription of antibiotics, as seen from case scenarios and record reviews. However, they
have lower scores on assessment of signs and symptoms of pneumonia and classifying them, as
shown by the relatively lower scores on overall knowledge of pneumonia (40%), on eliciting signs
and symptoms (25%) and counting breathing rates (35% able to count within two breaths of the gold
standard). When breathing rates are determined, a higher proportion is able to categorize them
correctly into fast breathing or otherwise. In addition, having a CHW who provides integrated care
increases promptness of appropriate antibiotics compared with a CHW who treats malaria only and
45
refers children with pneumonia symptoms to nearby health facilities. This effect was, however, not
apparent at the community level, and this may be due to the small number of children with
pneumonia symptoms in the community that were treated by CHWs, the problems of recall that were
inherent in the survey that was used in study IV to evaluate community drug use, and the use of selfreported symptoms. Over-prescription of antibiotics by CHWs was not apparent.
The CHWs are contributing to the use of effective medicines in a setting where children continue to
receive medicines such as chloroquine and sulphadoxine-pyrimethamine that are no longer
recommended for malaria treatment. Some children also receive drugs such as cotrimoxazole which
although not yet replaced in Ugandan treatment guidelines for pneumonia, has widespread resistance
(Hoa et al., 2010; Joloba et al., 2001). CHWs were the most common source of amoxicillin, a drug
that is currently recommended for treatment of pneumonia in the community (Grant et al., 2009). To
corroborate this finding, several health providers in the qualitative study mentioned as first-line drugs
for treatment of malaria or pneumonia drugs that are no longer recommended. They also had other
inappropriate practices such as routine use of dexamethasone, a steroid for treatment of respiratory
symptoms.
Misuse of corticosteroids in respiratory infections has been reported in India (Raveenthiran, 2008;
Shrivastava, 2009) and it may predispose children to many complications including Cushing’s
syndrome (Rajendra et al., 2013). Other inappropriate practices included crushing and mixing of all
medicines into powder, of which the caregivers were directed to measure off small quantities to
administer to the children. This may cause unwanted chemical reactions between different drugs,
resulting in inactivation of medicines or toxic reactions in children. Inadequate quantities of
medicines were also likely to be administered. ICCM is likely to reduce such practices through
provision of pre-packaged and dispersible medicines which may be easier for caregivers to
administer (Yeboah-Antwi et al., 2001; Abdulla and Sagara, 2009), and by availing a limited range of
drugs which minimizes misuse of other drugs. Children treated by CHWs in both the intervention and
control areas received fewer drugs on average than children treated by other health providers,
suggesting more rational use of medicines by CHWs.
Adherence: Caregivers performed well in adherence to medicines. Contrary to the hypothesis that
increased pill burden would lower adherence (Osterberg and Blaschke, 2005), there was high
adherence in children that received both antimalarials and antibiotics and in children that received
antimalarials alone. Caregivers of children that received both antimalarials and antibiotics may have
perceived the children’s illness as severe and thus they may have made more effort to adhere to
medicines. Perceived severity of illness has been found to influence medicines use (Bush and
Iannotti, 1990). However, adherence could have been over-estimated, since the exact timing of doses
was not assessed. The timing rather than the complete omission of doses has been reported as a very
common aspect of non-adherence (Lemma et al., 2011). The adherence found in this study is similar
to that in previous studies of CHWs (Ajayi et al., 2008a; Ajayi et al., 2008b; Chinbuah et al., 2006).
The main reasons cited for non-adherence including forgetfulness, improvement of the child and
vomiting have also been reported in previous studies (Beer et al., 2009; Kolaczinski et al., 2006).
Appropriate caregiver counselling should be undertaken to ensure that all the medicines
administration instructions have been understood, that caregivers continue to give medicines even
when children seem well, and that additional doses are given to the children when they vomit
medicines. High adherence should be maintained so as to protect the medicines against resistance.
Health status
There was no difference in overall treatment failure among children treated by CHWs in the
intervention and control arms (study III), similar to a study in Zambia (Yeboah-Antwi et al., 2010).
46
In contrast to that study, children treated by CHWs in the intervention arm were less likely to have
high temperature (i.e. 37.5oC) on day four compared with children in the control arm. Furthermore,
the reduction in the proportion of children with fast breathing between day one and four in the
intervention arm (9%) was more than twice that in the control arm (4%), suggesting that integrated
management of illness improves health status compared with single illness management. In the
Zambian study, high temperature, fast breathing, and difficult breathing on day four were grouped
together, and differences on the specific outcomes could have been missed. The treatment outcomes
for pneumonia treatment reported in this thesis may have been lower than what the intervention could
achieve because of the three-day doses of amoxicillin used in the cRCT, rather than the five-day
doses recommended for high HIV prevalence areas like Uganda (Grant et al., 2009). In addition, the
age-specific doses were lower than the recommended 50mg per kilogram of body weight daily
(Grant et al., 2009), mostly for the older children in each age category who were likely to be heavier.
Children may not have had adequate drug exposure for good treatment outcomes. However, selfreported treatment failure was lower in the intervention arm compared to the control arm suggesting
improved treatment outcomes from the caregivers’ perspective.
Patient satisfaction
Caregivers of children that sought care from CHWs rated the care they received highly (study IV).
There were slightly higher proportions of caregivers in the intervention arm that rated the quality of
care received from CHWs as good compared with the control arm, although it was not significantly
different. In studies I and III, only one caregiver from each study rated care from CHWs as poor
(both times from the control arm), while in study IV some caregivers who sought care from
government facilities (4.8%), private clinics (1.3%) and drug shops (0.6%) rated the care received as
poor. Another study in Uganda also found high levels of caregivers’ satisfaction with CHW-care
(Mukanga et al., 2012a). Caregivers’ high rating of health care may reflect its acceptability or patientcentredness, which is one of the attributes of quality of health care (Institute of Medicine, 1990).
Equity
Children in the lower wealth quintiles were more likely to utilize the services of the CHWs (study
IV). This is a positive finding since it suggests that the CHW programme in this setting is reaching
the poorer children as planned for community-based programs (de Sousa et al., 2011). The findings
are in agreement with studies that found the poorest people more likely to utilize CHWs’ services
compared to the least poor (Hasegawa et al., 2013; Kisia et al., 2012) but contrast with findings in
western Uganda where children from richer households utilized CHW services more than the poor
(Nsungwa-Sabiiti et al., 2007).
Resistance
In this thesis resistance was not directly studied, but some drivers of resistance were studied and are
outlined here. There was higher appropriate use of antimalarials among children treated by CHWs
compared with other health providers (study IV) and this may reduce the risk of resistance to
antimalarials (Bloland, 2001). However, no RDTs were used in malaria diagnosis, which may
contribute to an unnecessary increase in drug pressure and result in increased risk of resistance to
antimalarials (Bloland, 2001). Resistance to antibiotics would also be increased by over-use of
antibiotics and inadequate drug exposure (Cars and Nordberg, 2005) which could occur through use
of poor quality antibiotics, use of inadequate doses of antibiotics or non-adherence. The CHWs did
not use antibiotics more than other health providers and thus may not be increasing drug pressure
unnecessarily. However, to further reduce this risk, they should have adequate knowledge and skills
in respiratory assessment, together with reliable instruments to aid diagnosis, so that only children
that need antibiotics receive them. This area, as shown in study I, needs further improvement with
increased training and provision of necessary diagnostic aids.
47
Influence of livelihood assets
Social capital: In study I, the performance of CHWs was influenced by community support. The
local leaders in the areas where the CHWs did their work helped to mobilize caregivers to seek care
from the CHWs, and this enhanced the CHWs’ performance. This is in contrast to a study in western
Uganda where CHWs felt that the community was not supportive at all (Uganda Program for Human
and Holistic Development (UPHOLD), 2007). This difference in findings could stem from
differences in contexts. In the current study, the local leaders were the first persons sensitized about
the programme and they may have continued with the responsibility of sensitizing the community.
The caregivers sometimes did not comply with the referral advice given by the CHWs and this
negatively impacted on their work. Caregiver’s non-compliance with referral has been reported in
another study in western Uganda (Källander et al., 2006b). In addition, the caregivers’ lack of
appreciation of age cut-offs for children treated sometimes created conflicts. These findings are
similar to what was reported in Uganda under HBMF where caregivers would lie about the ages of
their children so that they could be treated by the CHWs (Uganda Program for Human and Holistic
Development (UPHOLD), 2007). Furthermore, there was lack of cooperation from health workers at
the health facilities where some of the children were referred; they did not treat the referred children
promptly. There is need for strong collaboration and cooperation between the CHWs, the
communities that they serve, and the formal health providers if the ICCM strategy is to succeed. The
lack of support from the formal health system was also reported in Mali (Callaghan-Koru et al.,
2013).
Human capital: In study IV knowledge of malaria and pneumonia was associated with improved
malaria treatment practices and care-seeking from CHWs, while in study III caregiver education was
shown to influence prompt and appropriate pneumonia treatment. Children whose caregivers had
received post-primary education were more likely to receive prompt and appropriate pneumonia
treatment. In study II caregivers’ perception that illness was not very severe, as well as their not
understanding the medicines administration instructions, were associated with increased nonadherence. These factors have been associated with treatment practices in other studies (Hwang et al.,
2007; Assefa et al., 2008).
Financial capital: The wealth index of households influenced treatment seeking from CHWs (study
IV) and has been discussed under the section on equity.
The financial capital also affected the CHWs because they felt that the money they were given to
enable them collect medicines, although helpful, was not sufficient for them to do their work
effectively. This factor linked to physical capital, because the CHWs reported that they lacked
transport to follow up treated children, as well as materials such as torches or weather-proof wear that
would enable them to do their work at night or during rainy weather. The CHWs also reported that
sometimes they had to use their money to follow up treated children because of lack of transport.
This may demotivate the CHWs and lower their performance. Dissatisfaction with remuneration was
also reported among CHWs in India (Bajpai and Dholakia, 2011).
Influence of other factors on utilization of CHWs and quality of care
Presenting symptoms: Children who did not have fever were more likely to utilize the services of
CHWs and they were also more likely not to adhere to treatment. Caregivers may perceive illness
without fever as ‘not severe’ and therefore may seek care from the CHWs who are nearby. This could
48
also explain the non-adherence to treatment. Fever may be perceived by the caregivers as a serious
illness and therefore seek care from other health providers whom they perceive to be more qualified
than the CHWs. A study in western Uganda reported that caregivers considered care from CHWs as
‘first aid’ treatment which they gave to their children as they monitored them to determine if they
needed further care (Nsabagasani et al., 2007). Another study in Kenya found that caregivers whose
children had fever were more likely to seek care for those illnesses (Taffa and Chepngeno, 2005).
7.2 CONSIDERATIONS FOR THE HEALTH SYSTEMS
This thesis has assessed a model of health service delivery for under-five children that integrates
malaria and pneumonia management using volunteer CHWs. The findings show that integrated care
increases utilization of CHW services and increases prompt and appropriate antibiotics for
pneumonia symptoms. The CHWs continue to perform well on malaria despite their added roles of
pneumonia management, and their performance translates to better community drug use patterns for
malaria among the children they treat compared with other health providers. The findings have been
presented according to the modified access framework by Obrist (Figure 3). I would now like to turn
to the constituent building blocks of the health system as presented by WHO (Figure 1). In order to
successfully implement and scale up ICCM, some aspects of the building blocks of health system
need to be addressed as outlined below.
The CHWs are contributing to the health workforce and are important sources of health services
especially in areas that are under-served by formal health facilities. However, there are still questions
regarding their motivation and retention. CHWs in Uganda and some other countries are volunteer
workers and a question arises as to whether it is possible to sustain the CHWs providing care under
ICCM as volunteers. Some countries have used paid CHWs (Bhutta et al., 2010) but can countries
like Uganda, which already have difficulties in remunerating their formal health workers (Hagopian
et al., 2009; Kanyesigye, 2003; Nalugo, 2013), be able to pay CHWs? And if the CHWs cannot be
paid, what will be the motivation for them to provide high quality health care. Studies to find how
CHWs could be motivated were conducted under HBMF and they reported that CHWs could be
motivated with monetary and/or non-monetary incentives (Ludwick et al., 2013; Uganda Program for
Human and Holistic Development (UPHOLD), 2007) and further studies are still ongoing that may
provide more information regarding motivation of CHWs (Innovations at Scale for Community
Access and Lasting Effects (inSCALE), 2013).
For the information management, strategies for linking the CHWs’ data to the health management
information system that already exists in the country should be devised. Otherwise challenges of lack
of accurate data on diseases will be increased.
There is need for consistent availability of medical products and supplies required by the CHWs.
This has been a major problem of the formal health system with frequent shortages of drugs and
supplies (Achan et al., 2011). There will be more CHWs than health facilities, requiring innovative
ways of maintaining drugs and other supplies. Drugs and supply shortages have also been
encountered in CHW-programmes (Stekelenburg et al., 2003), and they not only interrupt health
service provision but also affect the community’s trust in the CHWs (Puett et al., 2013). In addition
the appropriate supplies and technologies that have high accuracy in diagnosis should be made
available to the CHWs.
The financial implications of maintaining the CHWs whether paid or volunteers and maintaining the
medical products and supplies will be great, requiring innovative financing mechanisms. Questions
arise as to whether the services should be free or they should be paid for. One study has estimated
49
that $3,750 per year is required to train, equip and support each CHW in sub-Saharan Africa
(McCord et al., 2012). Can the government of Uganda afford to consistently provide the ICCM
services free-of-charge given the current funding rates? There is need for more data on the costing of
ICCM and studies are ongoing in response to this need (Kasteng, 2013).
Another area that will need to be strengthened is leadership and governance. Strong leadership will be
required to ensure high quality of care and efficient resource use. Supervision of the CHWs has to be
strengthened and questions on the best modality of supervision of the CHWs arise. A study in Malawi
reported low supervision of CHWs mainly arising from having clinician supervisors that found it
difficult to leave their clinical duties to supervise CHWs (Callaghan-Koru et al., 2013). In addition,
health facilities have critical roles to play in terms of treating referred children and supervising CHWs,
and will they be able and willing to support the system? The ICCM strategy might require a stronger
health system than the facility-based system of care, and this creates a dilemma because the
community-based systems were started to support the weak health systems that existed especially in
hard to reach areas (de Sousa et al., 2011). The scale-up of ICCM should proceed cautiously until all
of the important pillars in the health system have been addressed. Or perhaps conversely, ICCM
implementation needs to come with health systems strengthening components.
Should antibiotics be available through CHWs?
One of the common concerns of community case management involving the use of antibiotics has
been putting antibiotics out into the community through persons with no formal medical training
(Marsh et al., 2008). This could increase drug pressure and thereby the risk of resistance. In addition,
if the drugs were misused through inadequate dosages and non-adherence, the risk of resistance
would increase further. This thesis did not substantiate such fears. There was high adherence reported
to medicines from CHWs. The CHWs also provide complete dosages of pre-packaged medicines
which combined with appropriate counselling of caregivers should improve drug use compared with
the common practice of caregivers buying incomplete doses of medicines (Minzi and Manyilizu,
2013; Wafula, 2013). The CHWs could also contribute to the rational use of antibiotics through
respiratory rate assessment prior to their prescription. The key issue is whether they are able to do
respiratory rate assessments and classify children into those that require antibiotics and those that do
not. The CHWs need a lot of support in this area and there is need for technical innovation of better
respiratory rate counting devices. The alternatives to CHWs in most of these hard to reach areas are
the drug shops and small private clinics; these are often manned by unqualified personnel whose
quality of care is questionable (Awor et al., 2012; Konde-Lule et al., 2010). These private clinics will
also need to be addressed, and ICCM tools could have a positive effect on private sector quality.
There is ongoing research to evaluate implementation of ICCM in drug shops, and findings from
baseline evaluations have been published (Awor et al., 2012).
Another fear is whether the CHWs will not give in to pressure from caregivers to prescribe
antibiotics. The CHWs are accountable to the communities that they serve and may therefore feel
compelled to give in to such demands. In addition to strengthening the CHWs with credible
diagnostics and good support supervision, the communities should be adequately sensitized about the
principles of rational use of drugs, so that they do not expect to receive medicines for every illness.
7.3 METHODOLOGICAL CONSIDERATIONS
Measurement errors
In study I CHWs’ performance was assessed using knowledge tests, case scenarios and review of
records. This may not have reflected the actual practice and therefore performance. However, the use
of several methods may have strengthened the assessment. Although direct observation with re50
examination by an expert is the most valid method of assessing performance, it is often not feasible
because it requires a lot of time and money to have the experts stationed at the CHWs’ posts, waiting
for patients with varied illness presentations as well as assessing a large number of CHWs (Cardemil
et al., 2012). The knowledge tests and case scenarios are likely to show the clinical knowledge of the
CHWs and their ability to apply knowledge, rather than actual practice (Kak et al., 2001). They are
however, able to present a wide range of illness possibilities, including rare conditions that might not
have been encountered in the course of the study, if real cases of sick children had been used
(Cardemil et al., 2012). However, these scenarios also had to be limited so as to minimize the length
of the questionnaire, and therefore they did not assess all situations comprehensively. The record
reviews present a larger number of cases seen over a longer period of time than can be seen with
direct observation but they still lack sufficient information to make adequate assessment of
performance (Cardemil et al., 2012). The record reviews were, however, able to show that some
CHWs would record high respiratory rates for children and not give antibiotics, while other children
with low respiratory rates were given antibiotics. They also showed some CHWs giving wrong doses
of medicines according to the age in the records, as well as some gaps in respiratory assessment;
some CHWs had one value of respiratory rate recorded for all children, which is unlikely to occur in
real life.
The methods used do not assess skills well but an attempt was made to assess the skills of CHWs in
assessing respiratory symptoms through observations of CHWs in the intervention arm. However,
CHWs may have altered their practice as a result of being observed (Hawthorne effect) (Rowe et al.,
2006) or as a result of being removed from the setting where they usually practise, since the
observations were done at health facilities. In addition, the IMCI-trained medical officers used to
observe counting of respiratory rates may not have been true gold standard.
The performance scores were generated using mean percentage scores from each of the items
assessed. This method gave equal weight to each of the responses, and yet some responses may have
been more important in prediction of performance than others. To overcome this problem, methods
such as PCA, which can take into account this relative importance, are recommended (Streiner and
Norman, 2003). Performance scores generated using PCA had fairly high correlations with scores
generated as mean percentages (0.64-0.88). The mean percentages were thus chosen for their relative
ease of interpretability (Streiner and Norman, 2003). The mean percentage scores may mask
deficiencies in performance on specific tasks (Agency for Healthcare Research and Quality (AHRQ),
n.d.), but given the high number of items assessed it was not possible to present each item separately.
Problems of recall could have occurred in studies II, III and IV which could have led to
misclassification of children. In study II, the caregivers were asked to recall the way medicines had
been taken since the day they got the treatment from the CHWs. However, this problem was most
likely minimized due to the short duration of time over which they were required to recall, since the
interview was conducted on day four. In studies III and IV the caregivers were required to recall the
symptoms that the child had, and how long they had taken before seeking care. This problem was
more likely to occur in study IV, where the caregivers were supposed to recall illness over the two
weeks prior to the interview. Recall of illness over a two-week period has been shown to
underestimate disease rates (Feikin et al., 2010). Surprisingly there was a very high fever illness rate
reported (69%, study IV), higher than that previously reported in the same area (Rutebemberwa et al.,
2009a). Fever illness may have been over-estimated and this may explain the low overall prompt and
appropriate malaria treatment reported. In addition, caregivers in study IV had to recall the treatment
that the child had received. Posters bearing the most commonly used medicines for childhood
illnesses in the study area were used to aid the caregivers to recall the medicines the child had taken
51
for the illness. The misclassifications in all three studies were more likely to be non-differential rather
than differential.
Reporting bias in study II, arising from use of caregiver reports of how medicines were administered,
could have led to over-estimation of adherence. Caregivers may have reported what they thought the
interviewer wanted to hear and not what they actually did. Self-reports of adherence lead to overestimation of adherence (Osterberg and Blaschke, 2005). The validity of adherence assessment,
however, was improved by combining caregiver reports with unannounced pill counts, which have
been shown to be better than announced pill counts (Minzi and Naazneen, 2008). However, the
HDSS is a unique setting where lots of studies have been done and caregivers may have experienced
similar studies, making them more likely to give answers that they expect interviewers to like (i.e.
social desirability bias). Nevertheless, some caregivers reported non-adherence and factors associated
with this non-adherence were determined, which offers opportunities for intervening to improve
adherence.
Caregiver reports of illness symptoms were used in the classification of pneumonia in studies III and
IV, which could also have led to misclassification. The algorithm used in IMCI and also by CHWs to
classify pneumonia involves counting breathing rates and categorizing them at different thresholds
into fast breathing or otherwise. Caregivers’ reports of fast breathing may not have met the
thresholds. Caregiver reports of pneumonia symptoms have been used in demographic health surveys
and in multiple indicator cluster surveys (UNICEF, 2013b) and in other studies (Diaz et al., 2013) to
identify children with suspected pneumonia, but they are likely to over-estimate the number of
children with pneumonia (Campbell et al., 2013). However, this over-estimation is unlikely to have
differed between the intervention and the control arm, thus providing valid comparisons of antibiotic
use for the two arms.
Caregivers may have changed their behaviour in study II due to the day one visit, which does not
occur routinely. They may have increased their adherence. In addition, the correct timing of doses
was not assessed in adherence assessment and could therefore have led to over-estimation of
adherence. Incorrect timing of medicines has been shown to contribute significantly to non-adherence
(Mace et al., 2011).
Selection bias
In study I a small number of CHWs could not be contacted for inclusion into the study. This may
have led to selection bias, since CHWs who were difficult to find may have been inactive in their
CHW roles and therefore had poor performance due to lack of experience. However, the impact of
this non-participation on the findings is likely to have been small, since they were few (five).
Study IV had a number of children with missing information, mostly of data extracted from the
HDSS data base. This missing information could have led to selection bias during the analysis of data
that involved these variables. The wealth index of CHWs extracted from the HDSS database in study
I was similarly affected. The missing data did not differ between the intervention and control arms,
thus affected both groups similarly. In addition, it was unlikely to be related to the study outcomes,
i.e. performance of CHWs in study I and utilization of CHWs and drug use in study IV.
Clustering
Clustering could have occurred in studies II, III and IV owing to the designs of the studies. In studies
II and III clustering was anticipated at the CHW level, with children treated by the same CHW likely
to have similar care, which could affect adherence and treatment outcomes similarly. In addition,
study III had clustering at the level of the sub-parishes that were used in randomization to the
52
intervention and the control arm, since it was in the setting of a cRCT. In study IV clustering was
adjusted for at the village level. It is necessary to adjust for clustering when it exists in order to
estimate valid standard errors and therefore valid confidence intervals. If not adjusted for, clustering
can cause narrow confidence intervals and therefore erroneous conclusion of statistical significance
when it does not exist (Hayes and Moulton, 2009; Wears, 2002). The sample sizes estimated for
these studies were also adjusted for the anticipated clustering.
Triangulation
This thesis has involved triangulation of data and methods to answer the various objectives (Patton,
1999). Sometimes qualitative and quantitative methods have been combined and other times several
data sources within the quantitative methods have been combined, either from the same study or even
across different studies. Study I included information obtained through qualitative (FGDs) and
quantitative methods. The quantitative methods also used a multi-method approach (knowledge tests,
case scenarios, record reviews and observations) to assess performance, as well as interviews with
caregivers of children that had been treated in the week prior to the interview. Study IV also used
information from qualitative (KIIs) and quantitative methods. Data from the four studies have been
combined as much as possible to answer objectives. The triangulation has helped to highlight the
good performance of CHWs on malaria (studies I and IV). It has also highlighted one reason why the
performance of other health providers was lower than of CHWs, i.e. use of antimalarials that are not
recommended (KIIs in study IV). Through the several methods used to assess performance,
triangulation has helped to highlight where the CHWs’ challenges in performance on pneumonia are.
Use of diagnostics
The cRCT in which studies for this thesis were nested did not employ RDTs which are currently
recommended for inclusion into the CHWs’ illness management. In addition, the CHWs in the
intervention arm used watches to count respiratory rates instead of respiratory timers as has been
recommended. The impact of lack of these diagnostics on the performance of CHWs is unclear, but it
is likely that non-use of RDTs could have led to over-treatment with antimalarials. However, Iganga
and Mayuge are in high malaria transmission areas of Uganda and the RDT positivity in this area is
likely to be high, implying that most children would still be eligible to receive antimalarials. There is
literature suggesting that parasite-based diagnosis may not be cost-effective in areas of high malaria
transmission (Graz et al., 2011; English and Scott, 2008). In addition, it is important to note that
health facilities and programmess in Uganda and other resource-limited areas experience frequent
drug and supplies shortages, and there might be times when CHWs are faced with such situations
(Githinji et al., 2013; Achan et al., 2011; Mubi et al., 2013; UNICEF, 2013a). The lack of respiratory
timers is one of the possible reasons why performance on counting respiratory rates was low.
53
8 CONCLUSIONS AND POLICY IMPLICATIONS
CHWs continue to perform well on malaria management despite the additional duties of
pneumonia management (studies I and IV). Higher proportions of children treated by
CHWs received prompt and appropriate treatment for malaria symptoms compared to
children treated by other health providers (study IV).
Management of pneumonia presents some difficulties for CHWs, mainly in assessment of
symptoms (study I). However, contrary to thinking that CHWs will over-use antibiotics,
the use of antibiotics in children treated by CHWs was not higher than that among other
health providers (study IV).
CHWs that provided integrated malaria and pneumonia management increased the
likelihood of children receiving prompt and appropriate antibiotics for pneumonia
symptoms compared with CHWs that provided malaria treatment only and referred
children with pneumonia symptoms to health facilities (study III).
The utilization of CHWs is higher among children from poorer households and where
CHWs can provide integrated malaria and pneumonia management compared to areas
where CHWs treat malaria only (study IV). However, most caregivers seek care from the
private sector.
High adherence to medicines was observed and was not lowered by the additional pill
burden of antibiotics. However, adherence was lower among children that were vomiting
and those whose caregivers did not understand medicines administration instructions
(study II).
Policy implications for Ministry of Health and ICCM implementing partners
CHWs may take on additional roles without lowering their performance in already established roles.
However, they should be supported with more training, with special emphasis on pneumonia
assessment and treatment, provision of necessary drugs, diagnostic materials and other supplies, and
by ensuring a supportive health system.
The scale-up of CHW programmes should include strategies aimed at improving their utilization.
Given the high rate of care-seeking from the private sector, strategies that improve management of
childhood illnesses should also address the private sector.
Health service providers especially in the private sector should implement policy recommendations
so that medicines that are no longer recommended are not used by patients. In addition, efforts should
be made to provide the recommended, but more expensive medicines at subsidized prices.
54
9 ACKNOWLEDGEMENTS
Many people have contributed to the development and completion of thesis and my sincere gratitude
goes to them all. I would like to thank the persons mentioned below in a special way:
The study participants for the various studies: caregivers, children, community health workers and
other health providers who generously shared their experiences.
My current main supervisor: Dr Tobias Alfven, thank you for your insightful suggestions and critical
reviews on all the papers and the thesis. Thank you for all the support that you have given me
throughout my doctoral studies. Your great support throughout the period of writing the thesis and in
making the final preparations is highly appreciated.
My former main supervisor: Prof. Stefan Peterson, for all the support (academic and non-academic)
that you have given me throughout my doctoral studies. Thank you for your great advice during
development of the project as you guided me to “sharpen the research question”, and subsequently as
we collected the data and wrote the papers. There was never a problem too big that you did not have
a solution for.
Assoc Prof. Charles Karamagi, my main supervisor at Makerere University, thank you for the
support that you gave right from the conception of this project to its completion. You attracted me to
the field of child health and I am grateful for your showing me the light. Your great words of wisdom
and encouragement always changed what seemed like a difficult situation into one where there was
victory. Thank you.
Dr Elizeus Rutebemberwa, it is through your generosity that I was able to work in Iganga-Mayuge
HDSS. Thank you for stepping aside, forfeiting what could have been yours and letting me have it.
Thank you for all the advice and support you gave during the data collection, for availing to me the
resources that you did, and for the extensive reviews that you did on the papers and the thesis. I am
very grateful.
Dr Sarah Ssali, thank you for the time that you have given me, and especially for your very valuable
comments on my qualitative work.
My external mentor: Prof. Florence Mirembe, thank you for being a pillar of support. Amidst the
many things and people seeking your attention, you always found time for me. You have been a great
inspiration.
The Health Systems Research group at KI: Prof. Goran Tomson, head of research group, thank you
for the advice you have given me right from the beginning when I was developing my project. For
your great guidance on the thesis, I am very grateful. You are a pillar of excellence. I also thank the
rest of the health systems group, for their input at various points of my doctoral education. Thank you
for the critical comments at my pre-dissertation seminars. Special thanks go to Assoc. Prof. Rolf
Wahlstrom, Dr Andreas Martensson, and Meena Daivadanum.
Faculty at Department of Public Health Sciences (PHS) and IHCAR: I thank the Director of the PHS
department, Prof. Lucie Laflamme, and the Director of Doctoral education, Prof. Marie Hasselberg;
thank you for all the support you gave me throughout my doctoral education. All the teachers at
IHCAR: Prof. Vinod Diwan, Prof. Cecilia Stalsby Lundborg, Prof. Elizabeth Faxelid, Assoc. Prof.
55
Asli Kulane, Assoc. Prof. Birgitta Rubenson, Assoc. Prof. Johan von Schreeb, Assoc. Prof Ayesha
De Costa, Assoc. Birger Forsberg, Prof. Anna Mia Ekstrom, Prof. Anna Thorson, Assoc. Prof Sarah
Thomsen, and Prof. Cecilia Magnusson, Kersti Radmark, Gun-Britt Eriksson, Marie Dokken,
Elisabeth Kaven, Marita Larsson, Bo Planstedt, Csaba Goszleth, Andreas Sandin, and Jacob
Carlsson.
My pre-defence committee: Assoc. Prof. Karin Kallander, Dr Jaran Eriksen and Dr David Mukanga
for critically reviewing my work and giving me very useful comments.
My half-time committee: Prof Otto Cars (Uppsala University), Assoc. Prof. George Nasinyama
(Makerere University), Prof. Inge Axelsson (Mitt Universitetet) for the critical review and the
comments that helped to improve my work.
I am very grateful to the Swedish International Development Cooperation Agency (SIDA) for
funding my doctoral education. I also thank WHO/TDR, which funded the parent trial where I nested
my studies and also funded some aspects of my studies. I thank the coordinators of SIDA at
Makerere: Prof. James Tumwine, Prof. Josephat Byamugisha, and Prof. Elly Katunguka. Ms Maria
Nakyewa and your team, thank you for your constant assistance. Prof Nelson Sewankambo, the
Principal of the College of Health Sciences whose intervention enabled me to get this scholarship. I
am very grateful.
Friends and colleagues at IHCAR: Lisa Blom (thank you for being a great person to share office
space with), Anna Bergstrom, Senia Rosales-Klintz, Patricia Awiti, Netta Beer, Christine Nalwadda,
David Mukanga, Roy Mayega, Agnes Nanyonjo, Barageine Justus, Hanani Tabana, Helga Namburi,
Meena Daivadanum, Gorrette Nalwadda, Simon Walusimbi, Muhammad Lubega, Elin Larsson,
Kristi Sidney, Susanne Stromdahl, Sandeep Nerkar, Dorcus Kiwanuka, Ulrika Baker, Annelie
Eriksson, Jessica Pafs, Sara Eriksson, Sigga Baldursdottir, Myroslava Protsiv (thanks for all the
assistance with the dissertation application), Helle Molsted (it was a great joy to share the final lap
with you, thank you for the laughter in difficult times), my twin: Dr Ann Lindstrand (it was great
braving the hot January sun of Iganga with you, and thank you for the hospitality that you showed me
in Stockholm).
Fellow PhD students at Makerere: Esther Buregyeya, Achan Jane, Nantanda Rebecca, Norah
Mwebaza, Joaniter Nankabirwa, Jane Nakibuuka, and Prossy Naluyima. Thank you for sharing the
joys and the trials. PhD forum at School of Public Health, Makerere: Babirye Juliet, Nabiwemba
Elizabeth, Sarah Bunoti, Richard Mangwi, Allen Kabagenyi, Catherine Kansiime, Eleanor Turyakira,
Bonnie Wandera, Barbara Kirunda, Gertrude Nakigozi, Joseph Matovu, Fred Nalugoda, Aloysius
Mutebi, Aggrey Mukose, Raymond Tweheyo, Vincent Batwala, Florence Tushemereirwe, Simon
Kamukama and Godfrey Kigozi. PhD forum of College of Health Sciences: Doris Mwesigire,
Kenneth Opio, Ndibazza Juliet, Josephine Bwogi, Emmy Okello, Nakimuli Annette, Mwaka Amos,
Ronald Kiguba, Sam Ononge, Lydia Nakiyingi, Allan Lugaaju, Ian Munabi, Miriam Nakalembe,
Fatuma Namusoke, Sarah Nakubulwa, Denise Meya, Ocan Moses, Ochieng Francis, and Musa
Sekikubo.
Other friends: Victoria Walusansa (thank you for looking out for my children when I was away) and
Ivy Kasirye. My landlords in Sweden: Monique and Omar Morales, thank you.
Research assistants: Jennifer Kataike, Clare Nsubuga, Dr Tusubira Sheila, Dr Amanda Wanyana,
Zach Ojiek and the 30-man team that did data collection for study II and III, the HDSS field teams.
56
Other staff at DSS: Michael Kibuuka, Edward Galiwango, Judith Kaija, Dan Kadobera, James
Kalungi, Jane Mirembe, Kirunda, and Najib.
Leverne Gething, thank you for the language edits.
My colleagues at work: Department of Pharmacy: Mr Oria Hussein, Mr Kalidi Rajab, Ms Pamela
Blessed, Mr Mike Mugisha, Ms Eva Mugisa, Mr Bruhan Kaggwa, Mr Edson Ireeta, Ms Winnie
Nambatya, Dr Onegi Bernadina, Prof Richard Odoi-Adome, Mr Paul Kutyabami, Mr Freddy Kitutu,
Mr Nobert Anyama, Mr Otto Robert, Mr Malcolm Kavuma Gwayambadde, Ms Damalie Namusabi,
Mr Okello, Mr Sabila Bosco, Mr Aguma Bush, and Mr Kimani Karenye; CEU: Prof. Karamagi
Charles, Dr Achilles Katamba, Dr Joaniter Nankabirwa, and Ms. Christine Maholo
My family: My mother, my late father. Mum, you are a true hero and a great inspiration. Thank you
for all that you are. My Aunt Florence Nanteza, thank you for being there. My sisters: Aisha,
Bernadette, Catherine, Florence, Teo. Thank you for supporting me all the times that I needed you.
My parents-in-law, Rtd Bishop George and Edith Sinabulya (for your encouragement and blessings
all the time I travelled). My brothers- and sisters-in-law: Ruth, Rachel and Grace, thank you for
always checking on me when I was in Sweden. I thank the rest of the Sinabulya family for the
support given to my husband and I. Last but not least: My husband Peter, you are truly “the rock”.
Thank you for all your support, taking care of me and our children, especially during the times I was
away in Stockholm or in Iganga during data collection. My children: Norah, Julius, George and
Samuel. Thank you dears for bearing with the “absentee mum”. Your love and understanding made a
whole lot of difference to my struggles.
To Him who is able to do exceedingly more than we can imagine … Be glorified
57
10 REFERENCES
Abdulla, S. & Sagara, I. 2009. Dispersible formulation of artemether/lumefantrine: specifically
developed for infants and young children. Malaria Journal, 8, S7.
Abuya, T., Mutemi, W., Karisa, B., Ochola, S., Fegan, G. & Marsh, V. 2007. Use of over-the-counter
malaria medicines in children and adults in three districts in Kenya: implications for private
medicine retailer interventions. Malaria Journal, 6, 57.
Achan, J., Tibenderana, J., Kyabayinze, D., Mawejje, H., Mugizi, R., Mpeka, B., Talisuna, A. &
D'Alessandro, U. 2011. Case management of severe malaria - a forgotten practice:
Experiences from health facilities in Uganda. PLoS ONE, 6, e17053.
Achan, J., Tibenderana, J., Kyabayinze, D., Wabwire Mangen, F., Kamya, M. R., D'Alessandro, U.,
Rosenthal, P. J. & Talisuna, A. 2009. Effectiveness of quinine versus artemether-lumefantrine
for treating uncomplicated falciparum malaria in Ugandan children: randomised trial. BMJ,
339, b2763.
Action for Global Health. 2010. Health workers urgently needed to reach the health MDGs. The
human resources for health crisis [Online], 8. Available:
http://www.actionforglobalhealth.eu/fileadmin/AfGH_Intranet/AFGH/Publications/HRH_brie
fing_paper.pdf [Accessed 18 Jul 2013].
Adamchak, S., Bond, K., MacLaren, L., Magnani, R., Nelson, K. & Seltzer, J. 2000. A guide to
monitoring and evaluating adolescent reproductive health programs. In: FOCUS on Young
Adults (ed.). Washington DC: FOCUS on Young Adults.
Agency for Healthcare Research and Quality (AHRQ). n.d. Methodological considerations in
generating provider performance scores for use in public reporting: A guide for community
quality collaboratives [Online]. Rockville: AHRQ. Available:
http://www.ahrq.gov/legacy/qual/value/perfscoresmethods/perfsctask6.htm [Accessed 4
October 2013].
Ahorlu, C. K., Koram, K. A., Ahorlu, C., De Savigny, D. & Weiss, M. G. 2006. Socio-cultural
determinants of treatment delay for childhood malaria in southern Ghana. Tropical Medicine
& International Health, 11, 1022-1031.
Ajayi, I., Browne, E., Bateganya, F., Yar, D., Happi, C., Falade, C., Gbotosho, G., Yusuf, B.,
Boateng, S., Mugittu, K., Cousens, S., Nanyunja, M. & Pagnoni, F. 2008a. Effectiveness of
artemisinin-based combination therapy used in the context of home management of malaria:
A report from three study sites in sub-Saharan Africa. Malaria Journal, 7, 190.
Ajayi, I. O., Browne, E. N., Garshong, B., Bateganya, F., Yusuf, B., Agyei-Baffour, P., Doamekpor,
L., Balyeku, A., Munguti, K., Cousens, S. & Pagnoni, F. 2008b. Feasibility and acceptability
of artemisinin-based combination therapy for the home management of malaria in four
African sites. Malaria Journal, 7, 8.
Ashwell, H. E. & Freeman, P. 1995. The clinical competency of community health workers in the
eastern highlands province of Papua New Guinea. Papua New Guinea Medical Journal, 38,
198-207.
Assefa, T., Belachew, T., Tegegn, A. & Deribew, A. 2008. Mothers' health care seeking behavior for
childhood illnesses in Derra District, Northshoa Zone, Oromia Regional State, Ethiopia
Ethiopian Journal of Health Sciences, 18, 87-94.
Awad, A. I. & Eltayeb, I. B. 2007. Self-medication practices with antibiotics and antimalarials among
Sudanese undergraduate university students. The Annals of Pharmacotherapy, 41, 1249-1255.
Awor, P., Wamani, H., Bwire, G., Jagoe, G. & Peterson, S. 2012. Private sector drug shops in
integrated community case management of malaria, pneumonia, and diarrhea in children in
Uganda. The American Journal of Tropical Medicine and Hygiene, 87, 92-96.
Baird, K. J. 2005. Effectiveness of antimalarial drugs. New England Journal of Medicine, 352, 15651577.
Bajpai, N. & Dholakia, R. H. 2011. Improving the performance of accredited social health activists in
India. Working paper series. Mumbai: Columbia Global Centres.
Baqui, A. H., Arifeen, S. E., Williams, E. K., Ahmed, S., Mannan, I., Rahman, S. M., Begum, N.,
Seraji, H. R., Winch, P. J., Santosham, M., Black, R. E. & Darmstadt, G. L. 2009.
Effectiveness of home-based management of newborn infections by community health
workers in rural Bangladesh. Pediatric Infectious Disease Journal, 28, 304-310.
Batega, W. D. 2004. Knowledge, attitudes, and practices about malaria treatment and prevention in
Uganda: A literature review. Kampala: Ministry of Health (Uganda).
58
Batega, W. D., Greer, G. J. & Plowman, B. A. 2004. Home based management of fever strategy in
Uganda. Assessment of the implementation and operation at district and community levels.
Arlington, VA: MoH (Uganda), WHO, BASICS II for the United States Agency for
International Development.
Batwala, V., Magnussen, P. & Nuwaha, F. 2011. Comparative feasibility of implementing rapid
diagnostic test and microscopy for parasitological diagnosis of malaria in Uganda. Malaria
Journal, 10, 373.
Baume, C., Helitzer, D. & Kachur, P. S. 2000. Patterns of care for childhood malaria in Zambia.
Social Science and Medicine, 51, 1491-1503.
Beer, N., Ali, A. S., Rotllant, G., Abass, A. K., Omari, R. S., Al-mafazy, A.-w. H., Bjorkman, A. &
Kallander, K. 2009. Adherence to artesunate-amodiaquine combination therapy for
uncomplicated malaria in children in Zanzibar, Tanzania. . Tropical Medicine and
International Health, 14, 766-774.
Bhutta, Z. A., Lassi, Z. S., Pariyo, G. & Huicho, L. 2010. Global experience of community health
workers for delivery of health related Millennium Development Goals: A systematic review,
country case studies, and recommendations for integration into national health systems.
Geneva: WHO. Global Health Workforce Alliance.
Blanas, D., Ndiaye, Y., Nichols, K., Jensen, A., Siddiqui, A. & Hennig, N. 2013. Barriers to
community case management of malaria in Saraya, Senegal: training, and supply-chains.
Malaria Journal, 12, 95.
Bloland, P. B. 2001. Drug resistance in malaria [Online]. Geneva: WHO. Available:
http://www.who.int/csr/resources/publications/drugresist/malaria.pdf [Accessed 2 October
2013].
Bojalil, R., Guiscafre, H., Espinosa, P., Martinez, H., Palafox, M., Guillermina, R. & Gutierrez, G.
1998. The Quality of Private and Public Primary Health Care Management of Children With
Diarrhoea and Acute Respiratory Infections in Tlaxcala, Mexico. Health Policy and Planning,
13, 323-331.
Boschi-Pinto, C., Bahl, R. & Martines, J. 2009. Limited Progress in Increasing Coverage of Neonatal
and Child-health Interventions in Africa and Asia. Journal of Health, Population and
Nutirition, 27.
Boslaugh, S. 2007. Secondary data sources for public health: A practical guide, New York,
Cambridge University Press.
Breman, J. G. 2012. Resistance to artemisinin-based combination therapy. The Lancet Infectious
Diseases, 12, 820-822.
British Medical Association & Royal Pharmaceutical Society of Great Britain 2012. British National
Formulary (BNF), London, Pharmaceutical Press.
Bryce, J., el Arifeen, S., Pariyo, G., Lanata, C. F., Gwatkin, D. & Habicht, J.-P. 2003. Reducing child
mortality: can public health deliver? The Lancet, 362, 159-164.
Bryman, A. 2004. Social research methods, Oxford, Oxford University Press.
Bush, P. J. & Iannotti, R. J. 1990. A Children's Health Belief Model. Medical Care, 28, 69-86.
Byarugaba, D. K. 2004. Antimicrobial resistance in developing countries and responsible risk factors.
International Journal of Antimicrobial Agents, 24, 105-110.
Callaghan-Koru, J., Gilroy, K., Hyder, A., George, A., Nsona, H., Mtimuni, A., Zakeyo, B., Mayani,
J., Cardemil, C. & Bryce, J. 2013. Health systems supports for community case management
of childhood illness: lessons from an assessment of early implementation in Malawi. BMC
Health Services Research, 13, 55.
Campbell, H., el Arifeen, S., Hazir, T., O'Kelly, J., Bryce, J., Rudan, I. & Qazi, S. A. 2013.
Measuring Coverage in MNCH: Challenges in Monitoring the Proportion of Young Children
with Pneumonia Who Receive Antibiotic Treatment. PLoS Medicine, 10, e1001421.
Cardemil, C. V., Gilroy, K. E., Callaghan-Koru, J. A., Nsona, H. & Bryce, J. 2012. Comparison of
Methods for Assessing Quality of Care for Community Case Management of Sick Children:
An Application with Community Health Workers in Malawi. American Journal of Tropical
Medicine and Hygiene, 87, 127-136.
Carr, D. 2004. Improving the Health of the World's Poorest People. In: Bureau, P. R. (ed.) Health
Bulletin. Population Reference Bureau.
Cars, O. & Nordberg, P. 2005. Antibiotic resistance – the faceless threat. International Journal of
Risk & Safety in Medicine, 17 103-110.
Chandy, S. J., Thomas, K., Mathai, E., Antonisamy, B., Holloway, K. A. & Stalsby Lundborg, C.
2012. Patterns of antibiotic use in the community and challenges of antibiotic surveillance in a
59
lower-middle-income country setting: a repeated cross-sectional study in Vellore, south India.
Journal of Antimicrobial Chemotherapy, 68, 229-236.
Chibwana, A., Mathanga, D., Chinkhumba, J. & Campbell, C. 2009. Socio-cultural predictors of
health-seeking behaviour for febrile under-five children in Mwanza-Neno district, Malawi.
Malaria Journal, 8, 219.
Chinbuah, A. M., Gyapong, J. O., Pagnoni, F., Wellington, E. K. & Gyapong, M. 2006. Feasibility
and acceptability of the use of artemether-lumefantrine in the home management of
uncomplicated malaria in children 6-59 months old in Ghana. Tropical Medicine and
International Health, 11, 1003 - 1016.
Chopra, M., Binkin, N. J., Mason, E. & Wolfheim, C. 2012. Integrated management of childhood
illness: what have we learned and how can it be improved? Archives of Disease in Childhood,
97, 350-354.
Chuc, N. T., Larsson, M., Falkenberg, T., Do, N. T., Binh, N. T. & Tomson, G. B. 2001.
Management of childhood acute respiratory infections at private pharmacies in Vietnam.
Annals of Pharmacotherapy, 35, 1283-1288.
Colvin, C. J., Smith, H. J., Swartz, A., Ahs, J. W., de Heer, J., Opiyo, N., Kim, J. C., Marraccini, T.
& George, A. 2013. Understanding careseeking for child illness in sub-Saharan Africa: A
systematic review and conceptual framework based on qualitative research of household
recognition and response to child diarrhoea, pneumonia and malaria. Social Science &
Medicine, 86, 66-78.
Commission on social determinants of health 2008. Closing the gap in a generation: health equity
through action on the social determinants of health. Final Report of the Commission on Social
Determinants of Health. In: WHO (ed.). Geneva: WHO.
Comoro, C., Nsimba, S. E. D., Warsame, M. & Tomson, G. 2003. Local understanding, perceptions
and reported practices of mothers/guardians and health workers on childhood malaria in a
Tanzanian district: implications for malaria control. Acta Tropica, 87, 305-313.
CORE Group, Save the Children, BASICS & MCHIP 2012. Community case management
essentials: Treating common childhood illnesses in the community. A guide for program
managers. Washington DC: CORE Group.
Countdown to 2015. 2012. Maternal, child and newborn survival [Online]. Geneva: WHO.
Available: http://www.countdown2015mnch.org/documents/2012Report/2012-profiles-KZ.pdf [Accessed 23 August 2013].
Creswell, J. W., Plano Clark, V. L., Gutmann, M. L. & Hanson, W. E. 2003. Advanced mixed
methods research designs. In: Tashakkori, A. & Eddlie, C. (eds.) Handbook of Mixed
Methods in Social & Behavioral Research. Thousand Oaks, CA: Sage.
Das, A. & Ravindran, T. S. 2010. Factors affecting treatment-seeking for febrile illness in a malaria
endemic block in Boudh district, Orissa, India: policy implications for malaria control.
Malaria Journal, 9, 377.
Das, L. K., Jambulingam, P. & Sadanandane, C. 2008. Impact of community-based presumptive
chloroquine treatment of fever cases on malaria morbidity and mortality in a tribal area in
Orissa State, India. Malaria journal, 7, 75.
de Savigny, D. & Adam, T. (eds.) 2009. Systems thinking for health systems strengthening. Alliance
for health policy and systems research, Geneva: Alliance for health policy and systems
research, WHO.
de Sousa, A., Tiedje, K. E., Recht, J., Bjelic, I. & Hamer, D. H. 2011. Community case management
of childhood illnesses: policy and implementation in Countdown to 2015 countries. Bulletin
of the World Health Organization, 90, 183-190.
Degefie, T., Marsh, D., Gebremariam, A., Tefera, W., Osborn, G. & Waltensperger, K. 2009.
Community case management improves use of treatment for childhood diarrhea, malaria and
pneumonia in a remote district of Ethiopia. Ethiop. J. Health Dev, 23, 120-126.
Del Fiol, F. d. S., Lopes, L. C., Barberato-Filho, S. & Motta, C. d. C. B. 2012. Evaluation of the
prescription and use of antibiotics in Brazilian children. Brazilian Journal of Infectious
Diseases, 17, 332-337.
Department of Health Policy Planning and Management 2010. Community Medicine Distributor
training guide on Integrated Management of Malaria and Pneumonia. In: School of Public
Health Makerere University (ed.). Kampala.
Diaz, T., George, A., Rao, S., Bangura, P., Baimba, J., McMahon, S. & Kabano, A. 2013. Healthcare
seeking for diarrhoea, malaria and pneumonia among children in four poor rural districts in
Sierra Leone in the context of free health care: results of a cross-sectional survey. BMC Public
Health, 13, 157.
60
Downe-Wamboldt, B. 1992. Content analysis: Method, applications, and issues. Health Care for
Women International, 13, 313-321.
Enato, E. F. & Uwaga, C. F. 2011. Profile of antimicrobial drug use patterns in a Nigerian
metropolitan city. International Journal of Health Research, 4, 37-44.
English, M., Punt, J., Mwangi, I., McHugh, K. & Marsh, K. 1996. Clinical overlap between malaria
and severe pneumonia in African children in hospital. Transactions of the Royal Society of
Tropical Medicine and Hygiene, 90, 658-662.
English, M. & Scott, J. A. G. 2008. What is the future for global case management guidelines for
common childhood diseases? PLoS Medicine, 5, e241.
Feikin, D. R., Audi, A., Olack, B., Bigogo, G. M., Polyak, C., Burke, H., Williamson, J. & Breiman,
R. F. 2010. Evaluation of the optimal recall period for disease symptoms in home-based
morbidity surveillance in rural and urban Kenya. International Journal of Epidemiology, 39,
450-458.
Franco, L. M., Franco, C., Kumwenda, N. & Nkhoma, W. 2002. Methods for assessing quality of
provider performance in developing countries. International Journal for Quality in Health
Care, 14 suppl 1, 17-24.
Gadomski, A. M., Khallaf, N., el Ansary, S. & Black, R. E. 1993. Assessment of respiratory rate and
chest indrawing in children with ARI by primary care physicians in Egypt. Bulletin of the
World Health Organization, 71, 523-527.
GAVI Alliance. n.d. GAVI's strategy [Online]. GAVI Alliance. Available:
http://www.gavialliance.org/about/strategy/ [Accessed 20 June 2013].
Geissler, P. W., Nokes, K., Prince, R. J., Achieng' Odhiambo, R., Aagaard-Hansen, J. & Ouma, J. H.
2000. Children and medicines: self-treatment of common illnesses among Luo schoolchildren
in western Kenya. Social Science & Medicine, 50, 1771-1783.
Getahun, A., Deribe, K. & Deribew, A. 2010. Determinants of delay in malaria treatment-seeking
behaviour for under-five children in south-west Ethiopia: a case control study. Malaria
Journal, 9, 320.
Gilroy, K. E., Callaghan-Koru, J. A., Cardemil, C. V., Nsona, H., Amouzou, A., Mtimuni, A.,
Daelmans, B., Mgalula, L. & Bryce, J. 2012. Quality of sick child care delivered by Health
Surveillance Assistants in Malawi. Health Policy and Planning, 28, 573-585.
Githinji, S., Kigen, S., Memusi, D., Nyandigisi, A., Mbithi, A. M., Wamari, A., Muturi, A. N., Jagoe,
G., Barrington, J., Snow, R. W. & Zurovac, D. 2013. Reducing stock-outs of life saving
malaria commodities using mobile phone text-Messaging: SMS for life study in Kenya. PLoS
ONE, 8, e54066.
Global Health Initiative. 2009. The Future of Global Health: Ingredients for a bold and effective US
initiative [Online]. Available: http://www.theglobalhealthinitiative.org/documents/report.pdf
[Accessed 23 Jul 2013].
Gordis, L. 2000. Epidemiology, Philadelphia, W.B Saunders.
Gouws, E., Bryce, J., Habicht, J.-P., Amaral, J., Pariyo, G., Schellenberg, J. A. & Fontaine, O. 2004.
Improving antimicrobial use among health workers in first-level facilities: results from the
multi-country evaluation of the integrated management of childhood illness strategy. Bulletin
of the World Health Organization, 82, 509-515.
Gove, S. 1997. Integrated management of childhood illness by outpatient health workers: technical
basis and overview. The WHO working group on guidelines for integrated management of the
sick child. Bulletin of the World Health Organization, 75 Suppl 1, 7-24.
Graham, S. M., English, M., Hazir, T., Enarson, P. & Duke, T. 2008. Challenges to improving case
management of childhood pneumonia at health facilities in resource-limited settings. Bulletin
of the World Health Organization, 86, 349-355.
Graneheim, U. H. & Lundman, B. 2004. Qualitative content analysis in nursing research: concepts,
procedures and measures to achieve trustworthiness. Nurse Education Today, 24, 105-112.
Grant, G. B., Campbell, H., Dowell, S. F., Graham, S. M., Klugman, K. P., Mulholland, E. K.,
Steinhoff, M., Weber, M. W. & Qazi, S. 2009. Recommendations for treatment of childhood
non-severe pneumonia. The Lancet Infectious Diseases, 9, 185-196.
Graz, B., Willcox, M., Szeless, T. & Rougemont, A. 2011. "Test and treat" or presumptive treatment
for malaria in high transmission situations? A reflection on the latest WHO guidelines.
Malaria Journal, 10, 136.
Gwimile, J. J., Shekalaghe, S. A., Kapanda, G. N. & Kisanga, E. R. 2012. Antibiotic prescribing
practice in management of cough and/or diarrhoea in Moshi Municipality, Northern Tanzania:
cross-sectional descriptive study The Pan African Medical Journal, 12, 103.
61
Hagopian, A., Zuyderduin, A., Kyobutungi, N. & Yumkella, F. 2009. Job satisfaction and morale in
the Ugandan health workforce. Health Affairs, 28, w863-w875.
Hardon, A., Hodgkin, C. & Fresle, D. 2004. How to investigate the use of medicines by consumers.
Geneva: WHO, University of Amsterdam.
Hasegawa, A., Yasuoka, J., Ly, P., Nguon, C. & Jimba, M. 2013. Integrating child health services
into malaria control services of village malaria workers in remote Cambodia: service
utilization and knowledge of malaria management of caregivers. Malaria Journal, 12, 292.
Hayes, R. J. & Bennett, S. 1999. Simple sample size calculation for cluster-randomized trials.
International Journal of Epidemiology, 28, 319-326.
Hayes, R. J. & Moulton, L. H. 2009. Cluster randomized trials, Boca Raton (Florida), CRC Press.
Healthy Child Uganda. 2011. Community Health Workers and Community Case Management:
Policies and Progress in Uganda within a Global Context: Review of the Literature and
Situational Analysis [Online]. Available:
http://www.healthychilduganda.org/assets/Community-Health-Workers-and-CommunityCase-Management-April-2011.pdf [Accessed 20 February 2013].
Hildenwall, H. 2009. Beyond fever: Managing children with pneumonia symptoms in malaria
endemic Uganda. PhD thesis, Karolinska Institutet.
Hildenwall, H., Tomson, G., Kaija, J., Pariyo, G. & Peterson, S. 2008. "I never had the money for
blood testing" - Caretakers' experiences of care-seeking for fatal childhood fevers in rural
Uganda - a mixed methods study. BMC International Health and Human Rights, 8, 12.
Hoa, N., Trung, N., Larsson, M., Eriksson, B., Phuc, H., Chuc, N. & Lundborg, C. 2010. Decreased
Streptococcus pneumoniae susceptibility to oral antibiotics among children in rural Vietnam:
a community study. BMC Infectious Diseases, 10, 85.
Holmes, D. 2012. Margaret Chan: committed to universal health coverage. The Lancet, 380, 879.
Hopkins, H., Talisuna, A., Whitty, C. & Staedke, S. 2007. Impact of home-based management of
malaria on health outcomes in Africa: a systematic review of the evidence. Malaria Journal,
6, 134.
Hui, L., Li, X.-s., Zeng, X.-J., Dai, Y.-h. & Foy, H. M. 1997. Patterns and determinants of use of
antibiotics for acute respiratory tract infection in children in China. The Pediatric Infectious
Disease Journal, 16, 560-564.
Hwang, J., Graves, P. M., Jima, D., Reithinger, R., Kachur, P. S. & the Ethiopia MIS Working Group
2007. Knowledge of malaria and its association with malaria-related behaviors--results from
the malaria indicator survey, Ethiopia. PLoS One, 5, e11692.
Ilic, K., Jakovljevic, E. & Skodric-Trifunovic, V. 2012. Social-economic factors and irrational
antibiotic use as reasons for antibiotic resistance of bacteria causing common childhood
infections in primary healthcare. European Journal of Pediatrics, 171, 767-777.
Innovations at Scale for Community Access and Lasting Effects (inSCALE). 2013. Achieving
impact: supporting and motivating frontline health workers [Online]. Kampala: Malaria
Consortium. Available: http://www.malariaconsortium.org/inscale/pages/latestnews/achieving-impact-supporting-and-motivating-frontline-health-workers [Accessed 2
October 2013].
Institute of Medicine 1990. Medicare: A strategy for quality assurance, Washington, National
Academies Press.
Interagency Health Team. 2011. Global health initiative: Uganda: A Strategy for Accelerating
Reductions in Maternal and Neonatal Mortality [Online]. US Mission Uganda. Available:
http://www.ghi.gov/documents/organization/184707.pdf [Accessed 20 June 2013].
Isturiz, R. E. & Claude, C. 2000. Antibiotic use in developing countries. Infection Control and
Hospital Epidemiology, 21, 394-397.
Joint Formulary Committee 2007. British National Formulary, London, British Medical Association
and Royal Pharmaceutical Society of Great Britain.
Joloba, M. L., Bajaksouzian, S., Palavecino, E., Whalen, C. & Jacobs, M. R. 2001. High prevalence
of carriage of antibiotic-resistant Streptococcus pneumoniae in children in Kampala Uganda.
International journal of antimicrobial agents, 17, 395-400.
Kachur, S. P., Khatib, R. A., Kaizer, E., Fox, S. S., Abdulla, S. M. & Bloland, P. B. 2004. Adherence
to antimalarial combination therapy with sulfadoxine-pyrimethamine and artesunate in rural
Tanzania. American Journal of Tropical Medicine and Hygiene, 71, 715-722.
Kak, N., Burkhalter, B. & Cooper, M.-A. 2001. Measuring the competence of healthcare providers.
Operations Research Issue Paper, 2, 1-28.
Källander, K. 2006. Case management of childhood fevers in the community : Exploring malaria and
pneumonia care in Uganda. PhD thesis, Karolinska Institutet.
62
Källander, K., Hildenwall, H., Waiswa, P., Galiwango, E., Peterson, S. & Pariyo, G. 2008. Delayed
care seeking for fatal pneumonia in children aged under five years in Uganda: A case-series
study. Bulletin of the World Health Organization, 86, 332-338.
Källander, K., Nsungwa-Sabiiti, J. & Peterson, S. 2004. Symptom overlap for malaria and
pneumonia—policy implications for home management strategies. Acta Tropica, 90, 211-214.
Källander, K., Tomson, G., Nsabagasani, X., Nsungwa-Sabiiti, J., Pariyo, G. & Peterson, S. 2006a.
Can community health workers and caretakers recognize pneumonia in children? Experiences
from western Uganda. Transactions of the Royal Society of Tropical Medicine and Hygiene,
100, 956-963.
Källander, K., Tomson, G., Nsungwa-Sabiiti, J., Senyonjo, Y., Pariyo, G. & Peterson, S. 2006b.
Community referral in home management of malaria in Western Uganda: A case series study.
BMC International Health and Human rights, 6, 2.
Kamal-Yanni, M., Potet, J. & Saunders, P. 2012. Scaling-up malaria treatment: a review of the
performance of different providers. Malaria Journal, 11, 414.
Kangwana, B. B., Njogu, J., Wasunna, B., Kedenge, S. V., Memusi, D. N., Goodman, C. A.,
Zurovac, D. & Snow, R. W. 2009. Malaria drug shortages in Kenya: A major failure to
provide access to effective treatment. American Journal of Tropical Medicine and Hygiene,
80, 737-738.
Kanyesigye, E. K. 2003. Selective salary reward for heath workers: Realistic of a distortion? . Health
Policy and development, 1, 3-5.
Kasirye, I. 2012. Ensuring adequate early childhood development for Uganda’s children [Online].
Kampala: EPRC. Available:
http://www.eprc.or.ug/pdf_files/Policy%20Brief%20NO%2018.pdf [Accessed 24 July 2013].
Kasteng, F. 2013. Sustaining ICCM: focus on costing [Online]. London: Malaria Consortium,
Innovations at Scale for Community Access and Lasting Effects (inSCALE). Available:
http://www.malariaconsortium.org/userfiles/9-sustaining-ICCM-focus-on-costing.pdf
[Accessed 2 October 2013].
Kelly, J. M., Osamba, B., Garg, R. M., Hamel, M. J., Lewis, J. J., Rowe, S. Y., Rowe, A. K. &
Deming, M. S. 2001. Community health worker performance in the management of multiple
childhood illnesses: Siaya District, Kenya, 1997-2001. American Journal of Public Health,
91, 1617-1624.
Kemble, S. K., Davis, J. C., Nalugwa, T., Njama-Meya, D., Hopkins, H., Dorsey, G. & Staedke, S.
G. 2006. Prevention and treatment strategies used for the community management of
childhood fever in Kampala, Uganda American Journal of Tropical Medicine and Hygiene,
74, 999-1007.
Kidane, G. & Morrow, R. H. 2000. Teaching mothers to provide home treatment of malaria in
Tigray, Ethiopia: a randomised trial. The Lancet, 356, 550-555.
Kisia, J., Nelima, F., Otieno, D., Kiilu, K., Emmanuel, W., Sohani, S., Siekmans, K., Nyandigisi, A.
& Akhwale, W. 2012. Factors associated with utilization of community health workers in
improving access to malaria treatment among children in Kenya. Malaria Journal, 11, 248.
Kitzinger, J. 1995. Qualitative research: Introducing focus groups. BMJ, 311, 299-302.
Kolaczinski, J., Ojok, N., Opwonya, J., Meek, S. & Collins, A. 2006. Adherence of community
caretakers of children to pre-packaged antimalarial medicines (HOMAPAK(R)) among
internally displaced people in Gulu district, Uganda. Malaria Journal, 5, 40.
Konde-Lule, J., Gitta, S., Lindfors, A., Okuonzi, S., Onama, V. & Forsberg, B. 2010. Private and
public health care in rural areas of Uganda. BMC International Health and Human Rights, 10,
29.
Kristiansson, C., Gotuzzo, E., Rodriguez, H., Bartoloni, A., Strohmeyer, M., Tomson, G. & Hartvig,
P. 2009. Access to health care in relation to socioeconomic status in the Amazonian area of
Peru. International Journal for Equity in Health, 8, 11.
Kumar, D., Aggarwal, A. K. & Kumar, R. 2009. The effect of interrupted 5-day training on
integrated management of neonatal and childhood illness on the knowledge and skills of
primary health care workers. Health Policy and Planning, 24, 94-100.
Lawford, H., Zurovac, D., O'Reilly, L., Hoibak, S., Cowley, A., Munga, S., Vulule, J., Juma, E.,
Snow, R. & Allan, R. 2011. Adherence to prescribed artemisinin-based combination therapy
in Garissa and Bunyala districts, Kenya. Malaria Journal, 10, 281.
Laxminarayan, R., Bhutta, Z., Duse, A., Jenkins, P., O'Brien, T., Okeke, I. N., Pablo-Mendez, A. &
Klugman, K. P. 2006. Chapter 55 Drug resistance. In: Jamison DT, Breman JG, Measham
AR, Alleyne G, Claeson M, Evans DB, Jha P, Mills A & Musgrove P (eds.) Disease Control
Priorities in Developing Countries. 2nd ed. Washington DC: World Bank.
63
Lemma, H., Lofgren, C. & San Sebastian, M. 2011. Adherence to a six-dose regimen of artemetherlumefantrine among uncomplicated plasmodium falciparum patients in the Tigray Region,
Ethiopia. Malaria Journal, 10, 349.
Leshem, S. & Trafford, V. 2007. Overlooking the conceptual framework. Innovations in Education
and Teaching International, 44, 93-105.
Liamputtong, P. & Ezzy, D. 2005. Qualitative research methods, Melbourne, Oxford University
Press.
Ludwick, T., Brenner, J. L., Kyomuhangi, T., Wotton, K. A. & Kabakyenga, J. K. 2013. Poor
retention does not have to be the rule: retention of volunteer community health workers in
Uganda. Health Policy and Planning.
Luque, J. S., Whiteford, L. M. & Tobin, G. A. 2008. Maternal recognition and health care-seeking
behavior for acute respiratory infection in children in a rural Ecuadorian county. Maternal
and Child Health Journal, 12, 287-297.
Mace, K. E., Mwandama, D., Jafali, J., Luka, M., Filler, S. J., Sande, J., Ali, D., Kachur, P. S.,
Mathanga, D. P. & Skarbinski, J. 2011. Adherence to treatment with artemether-lumefantrine
for uncomplicated malaria in rural Malawi. Clinical Infectious Diseases, 53, 772-779.
Malimbo, M., Mugisha, E., Kato, F., Karamagi, C. & Talisuna, A. O. 2006. Caregivers' perceived
treatment failure in home-based management of fever among Ugandan children aged less than
five years. Malaria Journal, 5, 124.
Marsh, D. R., Gilroy, K. E., Van de Weerdt, R., Wansi, E. & Qazi, S. 2008. Community case
management of pneumonia: at a tipping point? Bulletin of the World Health Organization, 86,
321-416.
Maslove, D., Mnyusiwalla, A., Mills, E., McGowan, J., Attaran, A. & Wilson, K. 2009. Barriers to
the effective treatment and prevention of malaria in Africa: A systematic review of qualitative
studies. BMC International Health and Human Rights, 9, 26.
Mason, E., Scherpbier, R. & Lawe-Davies, O. 2009. WHO position statement on IMCI. The Lancet,
374, 782-783.
Mbonye, A. K., Bygbjerg, I. C. & Magnussen, P. 2008. Prevention and treatment practices and
implications for malaria control in Mukono district Uganda. Journal of Biosocial Science, 40,
283-296.
McCord, G. C., Liu, A. & Singh, P. 2012. Deployment of community health workers across rural
sub-Saharan Africa: financial considerations and operational assumptions. Bulletin of the
World Health Organization, 91, 244-253.
McGarock, H. 2002. Unjustified antibiotic prescribing in the community: a major determinant of
bacterial antimicrobial resistance. Pharmacoepidemiology and Drug Safety, 11, 407-408.
Ministry of Health (Uganda). 2005. National Policy on Malaria Treatment 2005 [Online]. Kampala.
Available: http://www.health.go.ug/mcp/NationalPolicyonMalariaTreatment(07_03_06).pdf
[Accessed 24 August 2010].
Ministry of Health (Uganda). 2009a. Situation analysis: Village Health Teams Uganda 2009
[Online]. Kampala. Available:
http://www.malariaconsortium.org/inscale/downloads/uganda/VHT-situation-analysis.pdf
[Accessed 2 August 2013].
Ministry of Health (Uganda). 2009b. Uganda national communication strategy for promoting
rational use of medicines [Online]. Available:
http://apps.who.int/medicinedocs/documents/s16511e/s16511e.pdf [Accessed 15 April 2013].
Ministry of Health (Uganda) 2010a. Health Sector Strategic Plan III: 2010/11-2014-15. In: Health
(ed.). Kampala: Government of Uganda.
Ministry of Health (Uganda) 2010b. Uganda Clinical Guidelines, Kampala, Ministry of Health
(Uganda).
Ministry of Health (Uganda). 2010c. Village Health Teams: Strategy and Operational Guidelines
[Online]. Kampala. Available:
http://www.malariaconsortium.org/inscale/downloads/uganda/VHT-strategy-and-operationalguidelines.pdf [Accessed July 2013].
Ministry of Health (Uganda). 2011. Assessment of the Uganda Essential Medicines Kit - Based
Supply System in Uganda [Online]. Kampala. Available:
http://health.go.ug/docs/THE%20KIT%20Assessment%20Report%20%20August%202011.pdf [Accessed 22 April 2013].
Ministry of Health (Uganda), Health Systems 20/20 & Makerere University School of Public Health
2012. Uganda Health System Assessment 2011. Kampala and Bethesda: Health systems
20/20.
64
Ministry of Health (Uganda), WHO & UNICEF. 2010. Integrated community case management of
childhood malaria, pneumonia, and diarrhea: Implementation guidelines [Online]. Kampala.
Available: http://www.malariaconsortium.org/inscale/downloads/uganda/ICCMimplementation-guidelines.pdf [Accessed 28 July 2013 2013].
Minzi, M. O. S. & Naazneen, S. A. 2008. Validation of self reports and hospital pill count using
announced pill count as methods for determination of adherence to antiretroviral therapy.
Tanzania Journal of Health Research, 10, 84-88.
Minzi, O. M. S. & Manyilizu, V. S. 2013. Application of basic pharmacology and dispensing practice
of antibiotics in accredited drug-dispensing outlets in Tanzania. Drug, Healthcare and Patient
Safety, 5, 5-11.
MoH Republic of Rwanda 2009. Community IMCI / Community Case Management: Evaluation
Report of Community Health Workers Performance.
Morgan, L. D. 1997. Focus groups as qualitative research, Thousand Oaks, CA, Sage Publications
Mubi, M., Janson, A., Warsame, M., Martensson, A., Kallander, K., Petzold, M. G., Ngasala, B.,
Maganga, G., Gustafsson, L. L., Massele, A., Tomson, G., Premji, Z. & Bjarkman, A. 2011.
Malaria rapid testing by community health workers is effective and safe for targeting malaria
treatment: randomised cross-over trial in Tanzania. PLoS ONE, 6, e19753.
Mubi, M., Kakoko, D., Ngasala, B., Premji, Z., Peterson, S., Björkman, A. & Mårtensson, A. 2013.
Malaria diagnosis and treatment practices following introduction of rapid diagnostic tests in
Kibaha District, Coast Region, Tanzania. Malaria Journal, 12, 293.
Mukanga, D., Babirye, R., Peterson, S., Pariyo, G. W., Ojiambo, G., Tibenderana, J. K., Nsubuga, P.
& Kallander, K. 2011. Can lay community health workers be trained to use diagnostics to
distinguish and treat malaria and pneumonia in children? Lessons from rural Uganda.
Tropical Medicine and International Health, 16, 1234-1242.
Mukanga, D., Tibenderana, J., Peterson, S., Pariyo, G., Kiguli, J., Waiswa, P., Babirye, R., Ojiambo,
G., Kasasa, S., Pagnoni, F. & Kallander, K. 2012a. Access, acceptability and utilization of
community health workers using diagnostics for case management of fever in Ugandan
children: a cross-sectional study. Malaria Journal, 11, 121.
Mukanga, D., Tiono, A. B., Anyorigiya, T., Källander, K., Konaté, A. T., Oduro, A. R., Tibenderana,
J. K., Amenga-Etego, L., Sirima, S. B., Cousens, S., Barnish, G. & Pagnoni, F. 2012b.
Integrated community case management of fever in children under five using rapid diagnostic
tests and respiratory rate counting: A multi-country cluster randomized trial. American
Journal of Tropical Medicine and Hygiene, 87, 21-29.
Mukanga, D. O. 2012. Community case management of malaria and pneumonia in children:
Exploring use of diagnostics by community health workers in Uganda. PhD, Karolinska
Institutet.
Najnin, N., Bennett, C. M. & Luby, S. P. 2011. Inequalities in care-seeking for febrile illness of
under-five children in urban Dhaka, Bangladesh. Journal of Health, Population, and
Nutrition, 29, 523-531.
Nalugo, M. 2013. MPs want health workers’ pay doubled in next financial year. Monitor Newspaper.
Kampala: Nation Media Group.
Nanyunja, M., Nabyonga, O. J., Kato, F., Kaggwa, M., Katureebe, C. & Saweka, J. 2011. Malaria
treatment policy change and implementation: The case of Uganda. Malaria Research and
Treatment, 2011.
Ndugwa, R. P. & Zulu, E. M. 2008. Child morbidity and care-seeking in Nairobi slum settlements:
the role of environmental and socio-economic factors. Journal of Child Health Care, 12, 314328.
Nguyen, D. T. K., Leung, K. K., McIntyre, L., Ghali, W. A. & Sauve, R. 2013. Does integrated
management of childhood illness (IMCI) training improve the skills of health workers? A
systematic review and meta-analysis. PLoS ONE, 8, e66030.
Nsabagasani, X., Nsungwa-Sabiiti, J., Kallander, K., Peterson, S., Pariyo, G. W. & Tomson, G. 2007.
Home based management of fever in rural Uganda: community perceptions and provider
opinions. Malaria Journal, 6, 11.
Nshakira, N., Kristensen, M., Ssali, F. & Whyte, R. S. 2002. Appropriate treatment of malaria? Use
of antimalarial drugs for children's fevers in district medical units, drug shops and homes in
eastern Uganda. Tropical Medicine and International Health, 7, 309-316.
Nsungwa-Sabiiti, J. 2009. Home based management of fever in Uganda - towards improved malaria
treatment practices PhD, Karolinska Institutet.
65
Nsungwa-Sabiiti, J., Peterson, S., Pariyo, G. W., Ogwal-Okeng, J., Petzold, M. & Tomson, G. 2007.
Home-based management of fever and malaria treatment practices in Uganda. Transactions of
the Royal Society of Tropical Medicine and Hygiene, 101, 1199-1207.
Nsungwa-Sabiiti, J., Tomson, G., Pariyo, G. W., Ogwal-Okeng, J. & Peterson, S. 2005. Community
effectiveness of malaria treatment in Uganda - a long way to Abuja targets. Annals of
Tropical Pediatrics, 25, 91-100.
O'Brien, M. J., Squires, A. P., Bixby, R. A. & Larson, S. C. 2009. Role development of community
health workers: An examination of selection and training processes in the intervention
literature. American Journal of Preventive Medicine, 37, S262-S269.
Obrist, B., Iteba, N., Lengeler, C., Makemba, A., Mshana, C., Nathan, R., Alba, S., Dillip, A., Hetzel,
M. W., Mayumana, I., Schulze, A. & Mshinda, H. 2007. Access to health care in contexts of
livelihood insecurity: A framework for analysis and action. PLoS Medicine, 4, e308.
Okeke, T. A. & Okeibunor, J. C. 2010. Rural-urban differences in health-seeking for the treatment of
childhood malaria in south-east Nigeria. Health Policy, 95, 62-68.
Olofsson, S. K. & Cars, O. 2007. Optimizing drug exposure to minimize selection of antibiotic
resistance. Clinical Infectious Diseases, 45, S129-S136.
Onyango, E., Ayodo, G., Watsierah, C., Were, T., Okumu, W., Anyona, S., Raballah, E., Okoth, J.,
Gumo, S., Orinda, G. & Ouma, C. 2012. Factors associated with non-adherence to
artemisinin-based combination therapy (ACT) to malaria in a rural population from
holoendemic region of western Kenya. BMC Infectious Diseases, 12, 143.
Oreagba, I. A., Ashorobi, R. B., Olayemi, S. O., Oparah, A. C. & Awodele, O. E. 2007. Case
management of children with malaria in a Nigerian community. International Journal of
Pharmacy Practice, 15, 327-329.
Oshikoya, K. A., Njokanma, O. F., Bello, J. A. & Ayorinde, E. O. 2007. Family self-medication for
children in an urban area of Nigeria. Paediatric and Perinatal Drug Therapy, 8, 124-130.
Osterberg, L. & Blaschke, T. 2005. Adherence to medication. New England Journal of Medicine,
353, 487-497.
Osterholt, D. M., Rowe, A. K., Hamel, M. J., Flanders, W. D., Mkandala, C., Marum, L. H. &
Kaimila, N. 2006. Predictors of treatment error for children with uncomplicated malaria seen
as outpatients in Blantyre district, Malawi. Tropical Medicine and International Health, 11,
1147-1156.
Pan American Health Organization & WHO. n.d. About integrated management of childhood illness
(IMCI) [Online]. WHO. Available: http://www.paho.org/english/hcp/hct/imci/imci-aiepi.htm
[Accessed 20 May 2011].
Patton, M. Q. 1999. Enhancing the quality and credibility of qualitative analysis. Health Services
Research, 34, 1189-1208.
Pecoul, B., Chirac, P., Trouiller, P. & Pinel, J. 1999. Access to essential drugs in poor countries: A
lost battle? JAMA, 281, 361-367.
Perry, H. & Zulliger, R. 2012. How effective are community health workers?: Overview of current
evidence with recommendations for strengthening community health worker programs to
accelerate progress in achieving the health - related Millennium Development Goals
[Online]. Available:
http://www.coregroup.org/storage/Program_Learning/Community_Health_Workers/review%
20of%20chw%20effectiveness%20for%20mdgs-sept2012-condensedversion.pdf [Accessed
24 Jul 2013].
Powell, R. A. & Single, H. M. 1996. Focus groups. International Journal for Quality in Health Care,
8, 499-504.
Prassad, B. M. & Muraleedharan, V. R. 2007. Community health workers: a review of concepts,
practice, and policy concerns [Online]. CREHS. Available:
http://r4d.dfid.gov.uk/PDF/Outputs/Equitable_RPC/CHW_Prasad_Muraleedharan.pdf
[Accessed 13 June 2013].
Puett, C., Alderman, H., Sadler, K. & Coates, J. 2013. 'Sometimes they fail to keep their faith in us':
community health worker perceptions of structural barriers to quality of care and community
utilisation of services in Bangladesh. Maternal and Child Nutrition.
Rajendra, N. P., Karthik, T. S. & Amaresh, R. P. 2013. Iatrogenic cushing’s syndrome in an infant.
Journal of Clinical and Diagnostic Research, 7, 353-354.
Raveenthiran, V. 2008. Misuse of corticosteroids in infants of rural Tamilnadu. Indian Pediatrics, 45,
1008.
Rodin, J. & de Ferranti, D. 2012. Universal health coverage: the third global health transition? The
Lancet, 380, 861-862.
66
Roll Back Malaria. 2005. Global strategic plan 2005-2015 roll back malaria [Online]. Roll Back
Malaria Partnership. Available: http://www.rollbackmalaria.org/forumV/docs/gsp_en.pdf
[Accessed 20 May 2012 ].
Roll Back Malaria / World Health Organization. 2003. The Abuja declaration and the plan of action
[Online]. Geneva: World Health Organization. Available:
http://www.rollbackmalaria.org/docs/abuja_declaration.pdf [Accessed Jun 2012].
Rowe, S. Y., Olewe, M. A., Kleinbaum, D. G., McGowan, J. E., McFarland, D. A., Rochat, R. &
Deming, M. S. 2006. The influence of observation and setting on community health workers'
practices. International Journal for Quality in Health Care, 18, 299-305.
Rutebemberwa, E. 2009. Access to health care for febrile children in Uganda: Symptom recognition,
care seeking practices and provider choice. PhD thesis, Karolinska Institutet.
Rutebemberwa, E., Pariyo, G., Peterson, S., Tomson, G. & Kallander, K. 2009a. Determinants of
delay in care-seeking for febrile children in eastern Uganda. Tropical Medicine and
International Health, 14, 472-479.
Rutebemberwa, E., Pariyo, G., Peterson, S., Tomson, G. & Kallander, K. 2009b. Utilization of public
or private health care providers by febrile children after user fee removal in Uganda. Malaria
Journal, 8, 45.
Santos, D. B., Barreto, M. L. & Coelho, H. L. L. 2009. Drug use and associated factors in children
living in poor areas. Revista de Saúde Pública, 43, 768-778.
Save the Children. 2012. Health in the post-2015 development agenda: Save the children’s
submission to the UN consultation call for papers [Online]. London: Save the Children.
Available: www.worldwewant2015.org/file/295602/download/320481 [Accessed 25 August
2013].
Sazawal, S. & Black, R. E. 2003. Effect of pneumonia case management on mortality in neonates,
infants, and preschool children: a meta-analysis of community-based trials. The Lancet
Infectious Diseases, 3, 547-556.
Shankar, P. R., Partha, P. & Shenoy, N. 2002. Self-medication and non-doctor prescription practices
in Pokhara valley, Western Nepal: a questionnaire-based study. BMC Family Practice, 3, 17.
Shrivastava, S. K. 2009. Misuse of corticosteroids in Madhya Pradesh. Indian Pediatrics, 46, 361.
Sirima, S. B., Konate, A., Tiono, A., Convelbo, N., Cousens, S. & Pagnoni, F. 2003. Early treatment
of childhood fevers with pre-packaged antimalarial drugs in the home reduces severe malaria
morbidity in Burkina Faso Tropical Medicine and International Health., 8, 133-139.
Smith, L., Bruce, J., Gueye, L., Helou, A., Diallo, R., Gueye, B., Jones, C. & Webster, J. 2010. From
fever to anti-malarial: the treatment-seeking process in rural Senegal. Malaria Journal, 9, 333.
Smithson, J. 2000. Using and analysing focus groups: limitations and possibilities. International
Journal of Social Research Methodology, 3, 103-119.
Staedke, S. G., Mwebaza, N., Kamya, M. R., Clark, T. D., Dorsey, G., Rosenthal, P. J. & Whitty, C.
J. 2009. Home management of malaria with artemether-lumefantrine compared with standard
care in urban Ugandan children: a randomised controlled trial. The Lancet, 373, 1623-1631.
Stekelenburg, J., Kyanamina, S. S. & Wolffers, I. 2003. Poor performance of community health
workers in Kalabo District, Zambia. Health Policy, 65, 109-118.
Stewart, D. W., Shamdasani, P. N. & Rook, D. W. 1990. Focus groups: theory and practice,
Thousand Oaks, Sage Publications Inc.
Streiner, D. L. & Norman, G. R. 2003. From items to scales. Health measurement scales: A practical
guide to their development and use. 3rd ed. New York: Oxford University Press.
Sturm, A. W., van der Pol, R., Smits, A. J., van Hellemondt, F. M., Mouton, S. W., Jamil, B., Minai,
A. M. & Sampers, G. H. 1997. Over-the-counter availability of antimicrobial agents, selfmedication and patterns of resistance in Karachi, Pakistan. Journal of Antimicrobial
Chemotherapy, 39, 543-547.
Taffa, N. & Chepngeno, G. 2005. Determinants of health care seeking for childhood illnesses in
Nairobi slums. Tropical Medicine and International Health, 10, 240-245.
Task Team for the Global Thematic Consultation on Health. 2013. Health in the post-2015 agenda:
Report of the global thematic consultation on health [Online]. Available:
http://www.post2015hlp.org/wp-content/uploads/2013/04/health-in-the-post-2015agenda_LR.pdf [Accessed 25 August 2013].
The Global Fund to fight AIDS tuberculosis and malaria 2005. Annual report 2005. In: The Global
Fund (ed.). Geneva.
The Partnership for Maternal Newborn and Child Health. n.d. Global strategy for women's and
children's health [Online]. Geneva: WHO. Available:
67
http://www.who.int/pmnch/activities/jointactionplan/en/index.html [Accessed 20 August
2013].
Theodoratou, E., Al-Jilaihawi, S., Woodward, F., Ferguson, J., Jhass, A., Balliet, M., Kolcic, I.,
Sadruddin, S., Duke, T., Rudan, I. & Campbell, H. 2010. The effect of case management on
childhood pneumonia mortality in developing countries. International Journal of
Epidemiology, 39, i155-i171.
Tinuade, O., Iyabo, R.-A. & Durotoye, O. 2010. Health-care-seeking behaviour for childhood
illnesses in a resource-poor setting. Journal of Paediatrics and Child Health, 46, 238-242.
Trap, B. & Hansen, E. H. 2002. Cotrimoxazole prescribing by dispensing and non-dispensing
doctors: do they differ in rationality? Tropical Medicine and International Health, 7, 878-885.
Tumwine, Y., Kutyabami, P., Odoi, R. A. & Kalyango, J. N. 2010. Availability and expiry of
essential medicines and supplies during the ‘pull’and ‘push’ drug acquisition systems in a
rural Ugandan hospital. Tropical Journal of Pharmaceutical Research, 9, 557-564.
Uganda Bureau of Statistics & Macro International Inc. 2007. Uganda demographic and health
survey 2006 [Online]. Available: http://www.measuredhs.com/pubs/pdf/FR194/FR194.pdf
[Accessed 03 June 2010].
Uganda Ministry of Health. n.d. Malaria case management: home based management of fever
[Online]. Available: http://www.health.go.ug/mcp/mt.html [Accessed 20 May 2010].
Uganda Program for Human and Holistic Development (UPHOLD) 2007. Action research on
improving motivation of community medicine distributors (CMDs) and other community
based health workers. Kampala: USAID.
Ukwaja, K. N. & Olufemi, O. A. 2010. Home management of acute respiratory infections in a
Nigerian district. African Journal of Respiratory Medicine 18-22.
UN Inter-agency group for child mortality estimation 2013. Levels & trends in child mortality:
Report 2013. New York: UN Inter-agency Group for Child Mortality Estimation.
UNICEF. 1996. The 1980s: Campaign for child survival. Available:
http://www.unicef.org/sowc96/1980s.htm.
UNICEF. 2012. Pneumonia and diarrhoea: Tackling the deadliest diseases for the world’s poorest
children [Online]. New York: UNICEF. Available:
http://www.unicef.org/media/files/UNICEF_P_D_complete_0604.pdf [Accessed 26 August
2013].
UNICEF 2013a. Commiting to child survival: A promise renewed. In: Anthony, D. & Mullerbeck, E.
(eds.). New York: UNICEF.
UNICEF. 2013b. Pneumonia [Online]. Available: www.childinfo.org/pneumonia.html [Accessed
April 2013].
UNICEF. 2013c. UNICEF in action: Malaria [Online]. UNICEF. Available:
http://www.unicef.org/health/index_malaria.html [Accessed 22 September 2013].
UNICEF & WHO. 2006. Pneumonia: The forgotten killer of children [Online]. UNICEF/WHO.
Available:
http://www.unicef.org/mdg/mortalitymultimedia/Pneumonia_The_Forgotten_Killer_of_Child
ren.pdf [Accessed].
United Nations. 2000. 55/2. United Nations millennium declaration [Online]. Available:
http://www.un.org/millennium/declaration/ares552e.htm [Accessed 18 June 2013 ].
United Nations 2011. The Millennium Development Goals Report. New York: United Nations.
United Nations. 2013. The millennium development goals report 2013 [Online]. New York: United
Nations. Available: http://www.un.org/millenniumgoals/pdf/report-2013/mdg-report-2013english.pdf [Accessed 24 September 2013].
United Nations Development Program. 2013. Uganda country profile: Human development
indicators [Online]. Available: http://hdrstats.undp.org/en/countries/profiles/UGA.html
[Accessed 25 July 2013].
Venkatachalam, J., Kumar, D., Gupta, M. & Aggarwal, A. K. 2011. Knowledge and skills of primary
health care workers trained on integrated management of neonatal and childhood illness:
Follow-up assessment 3 years after the training. Indian Journal of Public Health, 55, 298-302.
Wafula, F. 2013. Availability and dispensing practices for antimalarials and antimicrobials in western
Kenyan pharmacies. Pharmaceutical Regulatory Affairs, 2, 1.
Waiswa, P., Peterson, S., Tomson, G. & Pariyo, G. 2010. Poor newborn care practices - a population
based survey in eastern Uganda. BMC Pregnancy and Childbirth, 10, 9.
Warsame, M., Kimbute, O., Machinda, Z., Ruddy, P., Melkisedick, M., Peto, T., Ribeiro, I., Kitua,
A., Tomson, G. & Gomes, M. 2007. Recognition, perceptions and treatment practices for
68
severe malaria in rural Tanzania: Implications for accessing rectal artesunate as a pre-referral.
PLoS ONE, 2, e149.
Wears, R. L. 2002. Advanced statistics: statistical methods for analyzing cluster and clusterrandomized data. Academic Emergency Medicine, 9, 330-341.
White, N. 1999. Antimalarial drug resistance and combination chemotherapy. Philosophical
Transactions of the Royal Society of London. Series B: Biological Sciences, 354, 739-749.
WHO 1985. Rational use of drugs: Report of the conference of experts. Geneva: WHO.
WHO 2001. Antimalarial drug combination therapy: Report of a WHO Technical Consultation.
Geneva: WHO.
WHO. 2002. Promoting rational use of medicines: core components [Online]. Geneva: WHO.
Available: http://apps.who.int/medicinedocs/pdf/h3011e/h3011e.pdf [Accessed 18 April
2013].
WHO 2003. Adherence to long-term therapies - Evidence for action. Geneva: WHO.
WHO. 2005a. Malaria control today: current WHO recommendations [Online]. Geneva: WHO.
Available: http://www.emro.who.int/sudan/pdf/malaria%20control%20today.pdf [Accessed
18 May 2010].
WHO 2005b. The roll back malaria strategy for improving access to treatment through home
management of malaria. In: WHO (ed.). Geneva: WHO.
WHO 2005c. Social health insurance: Sustainable health financing, universal coverage and social
health insurance Geneva: WHO
WHO. 2006a. The Stop TB strategy: Building on and enhancing DOTS to meet the TB-related
Millennium Development Goals [Online]. Geneva. Available:
http://whqlibdoc.who.int/hq/2006/WHO_HTM_STB_2006.368_eng.pdf [Accessed 20 July
2013].
WHO. 2006b. Vaccine-preventable diseases [Online]. Geneva: WHO. Available:
http://www.who.int/mediacentre/events/2006/g8summit/vaccines/en/ [Accessed 20 Jul 2013].
WHO. 2007a. Community health workers: What do we know about them? The state of the evidence
on programmes, activities, costs and impact on health outcomes of using community health
workers [Online]. Geneva: WHO. Available:
http://www.who.int/hrh/documents/community_health_workers_brief.pdf [Accessed 22
September 2010].
WHO. 2007b. Everybody's business: Strengthening health systems to improve health outcomes:
WHO's framework for action [Online]. Geneva: WHO Available:
http://www.who.int/healthsystems/strategy/everybodys_business.pdf [Accessed 20 April
2013].
WHO. 2008. The Bulletin series: primary health care 30 years on [Online]. Geneva: WHO.
Available: http://www.who.int/bulletin/primary_health_care_series/en/index.html [Accessed].
WHO. 2009. Medicines use in primary care in developing and transitional countries: Fact book
summarizing results from studies reported between 1990 and 2006 [Online]. Geneva: WHO
Press. Available: http://www.who.int/medicines/publications/primary_care_8April09.pdf
[Accessed 18 April 2013].
WHO. 2010a. Guidelines for the treatment of malaria [Online]. Geneva: WHO Available:
http://whqlibdoc.who.int/publications/2010/9789241547925_eng.pdf [Accessed 2 August
2012].
WHO 2010b. World Malaria report. Geneva.
WHO. 2011a. WHO global health expenditure atlas [Online]. Geneva: WHO Available:
http://www.who.int/nha/atlas.pdf [Accessed 25 July 2013].
WHO. 2011b. World Health Statistics [Online]. Geneva: WHO Press. Available:
http://www.who.int/whosis/whostat/EN_WHS2011_Full.pdf [Accessed].
WHO. 2011c. The world medicines situation [Online]. Geneva: WHO Available:
http://apps.who.int/medicinedocs/documents/s18064en/s18064en.pdf [Accessed 18 April
2013].
WHO. 2012a. African Region: Uganda statistics summary (2002 - present) [Online]. Geneva: WHO.
Available: http://apps.who.int/gho/data/node.country.country-UGA?lang=en [Accessed 25
July 2013].
WHO. 2012b. Cause specific mortality: Regional estimates for 2000-2011 [Online]. Geneva: WHO.
Available:
http://www.who.int/healthinfo/global_burden_disease/estimates_regional/en/index.html
[Accessed 18 July 2013].
69
WHO. 2012c. Global AMDP database - WHO African region [Online]. Geneva: WHO. Available:
http://www.who.int/malaria/am_drug_policies_by_region_afro/en/index.html [Accessed 13
July 2012].
WHO 2012d. Healthy and happy children: Investing in our common future. Brazaville: WHO
Regional Office for Africa.
WHO. 2012e. Life expectancy [Online]. Geneva: WHO. Available:
http://www.who.int/gho/mortality_burden_disease/life_tables/situation_trends_text/en/
[Accessed 20 June 2013].
WHO. 2012f. Millennium Development Goals (MDGs): Fact sheet No. 290 [Online]. Geneva: WHO
Media Centre. Available: http://www.who.int/mediacentre/factsheets/fs290/en/ [Accessed 23
August 2013].
WHO. 2012g. Positioning health in the Post - 2015 Development Agenda: WHO discussion paper
[Online]. Geneva: WHO. Available:
http://www.who.int/topics/millennium_development_goals/post2015/WHOdiscussionpaper_
October2012.pdf [Accessed 4 August 2013].
WHO 2012h. The world health report - Health systems financing: the path to universal coverage.
Geneva: WHO.
WHO. 2012i. World Health Statistics 2012 [Online]. Geneva: WHO Press. Available:
http://www.who.int/gho/publications/world_health_statistics/2012/en/ [Accessed 1 May
2013].
WHO. 2013a. Fact sheet No331: Pneumonia [Online]. Geneva: WHO Media Centre. Available:
http://www.who.int/mediacentre/factsheets/fs331/en/ [Accessed 22 July 2013].
WHO. 2013b. Immunization coverage: Fact sheet No 378 [Online]. Geneva: WHO. Available:
http://www.who.int/mediacentre/factsheets/fs378/en/ [Accessed 2 October 2013].
WHO. 2013c. Malaria: Fact sheet No.94 [Online]. Geneva: WHO Media Centre,. Available:
http://www.who.int/mediacentre/factsheets/fs094/en/ [Accessed 20 Sep 2013 2013].
WHO. 2013d. MDG 4: reduce child mortality [Online]. Geneva: WHO. Available:
http://www.who.int/topics/millennium_development_goals/child_mortality/en/ [Accessed 5
May 2013].
WHO. 2013e. Pneumonia: key facts [Online]. Geneva: WHO Media Center. Available:
http://www.who.int/mediacentre/factsheets/fs331/en/ [Accessed 10 July 2013].
WHO. 2013f. Uganda: health profile [Online]. Geneva: WHO. Available:
http://www.who.int/gho/countries/uga.pdf [Accessed 25 July 2013].
WHO & Global Health Workforce Alliance 2010. Global experience of community health workers
for delivery of health related Millennium Development Goals: A systematic review, country
case studies, and recommendations for integration into national health systems. Geneva.
WHO, UNAIDS & UNICEF. 2005. Joint WHO/UNAIDS/UNICEF statement on use of
cotrimoxazole as prophylaxis in HIV exposed and HIV infected children [Online]. Available:
http://www.who.int/3by5/mediacentre/en/Cotrimstatement.pdf [Accessed 22 July 2013].
WHO & UNICEF. 2004. WHO/UNICEF joint statement: Managment of pneumonia in community
settings [Online]. Available: www.unicef.org/publications/files/EN_Pneumonia_reprint.pdf
[Accessed 18 May 2010].
WHO & UNICEF. 2008. Integrated management of childhood illness chart booklet [Online].
Geneva: WHO Press. Available:
http://whqlibdoc.who.int/publications/2008/9789241597289_eng.pdf [Accessed].
WHO & UNICEF. 2011. Manual for the community health worker: Caring for the sick child in the
community: Treat diarrhoea, confirmed malaria, and fast breathing [Online]. Geneva: WHO
Press. Available:
http://whqlibdoc.who.int/publications/2011/9789241548045_Manual_eng.pdf [Accessed 17
July 2013].
WHO & UNICEF. 2012. Global immunization data [Online]. Geneva: WHO. Available:
http://www.who.int/immunization_monitoring/Global_Immunization_Data.pdf [Accessed 17
July 2013].
World Bank. 2012. Health expenditure per capita [Online]. Available:
http://data.worldbank.org/indicator/SH.XPD.PCAP/countries [Accessed 20 August 2013].
Yeboah-Antwi, K., Gyapong, J. O., Asare, I. K., Barnish, G., Evans, D. B. & Adjei, S. 2001. Impact
of prepackaging antimalarial drugs on cost to patients and compliance with treatment Bulletin
of the World Health Organization, 79, 394-399.
Yeboah-Antwi, K., Pilingana, P., Macleod, W. B., Semrau, K., Siazeele, K., Kalesha, P., Hamainza,
B., Seidenberg, P., Mazimba, A., Sabin, L., Kamholz, K., Thea, D. M. & Hamer, D. H. 2010.
70
Community case management of fever due to malaria and pneumonia in children under five in
Zambia: A cluster randomized controlled trial. PLoS Medicine, 7, e1000340.
Yeung, S. & White, N. J. 2005. How do patients use their drugs? A review of the evidence. . Tropical
Medicine and International Health, 10, 121-128.
Zurovac, D., Rowe, A. K., Ochola, S. A., Noor, A. M., Midia, B., English, M. & Snow, R. W. 2004.
Predictors of the quality of health worker treatment practices for uncomplicated malaria at
government health facilities in Kenya. International Journal of Epidemiology 33, 1080-1091.
71