* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download The Etiology of Genital Ulcer Disease in Zimbabwe: Implications for
Survey
Document related concepts
Schistosomiasis wikipedia , lookup
Marburg virus disease wikipedia , lookup
Carbapenem-resistant enterobacteriaceae wikipedia , lookup
Human cytomegalovirus wikipedia , lookup
Epidemiology of syphilis wikipedia , lookup
Middle East respiratory syndrome wikipedia , lookup
Epidemiology of HIV/AIDS wikipedia , lookup
Diagnosis of HIV/AIDS wikipedia , lookup
Neonatal infection wikipedia , lookup
Oesophagostomum wikipedia , lookup
Herpes simplex wikipedia , lookup
Herpes simplex virus wikipedia , lookup
Microbicides for sexually transmitted diseases wikipedia , lookup
Transcript
1 The Etiology of Genital Ulcer Disease in Zimbabwe: Implications for Syndromic 2 Management 3 More Mungati, MBChB, MPH1 4 Anna Machiha, DNS1 5 Owen Mugurungi, MD1 6 Mufuta Tshimanga, MD, MPH2 7 Peter H. Kilmarx, MD3,4 8 Justice Nyakura, MPH1 9 Gerald Shambira, MD, MPH2 10 Vitalis Kupara, DNS5 11 David A. Lewis FRCP (UK), PhD 6.7 12 Elizabeth Gonese, MPH3 13 Beth Barr, DrPH3 14 Cornelis Rietmeijer, MD, PhD, MSPH8,9 15 for the 16 The Zimbabwe STI Etiology Study Group 17 Affiliations: 18 1 Zimbabwe Ministry of Health and Child Care 19 2 Department of Community Medicine, University of Zimbabwe, College of Health Sciences 20 3 21 CDC, Atlanta. USA 22 4 23 5 Zimbabwe Community Health Intervention Research (ZiCHIRe) Project 24 6 Western Sydney Sexual Health Centre, Parramatta, New South Wales, Australia 25 7 Marie Bashir Institute for Infectious Diseases and Biosecurity & Sydney Medical School- 26 Westmead, University of Sydney, Sydney, New South Wales, Australia 27 8 Colorado School of Public Health, University of Colorado Denver, USA 28 9 Rietmeijer Consulting LLC, Denver, Colorado, Denver, USA U.S. Centers for Disease Control and Prevention, Zimbabwe and Division of Global HIV/AIDS, Fogarty International Center, National Institutes of Health, Bethesda, MD, USA. 29 30 Words in Summary: 41 31 Words in Abstract: 325 32 Words in manuscript: 3475 33 Tables: 5 34 Keywords: STI, Etiology, Syndromic Management, Genital Ulcer Disease 35 36 The Zimbabwe STI Etiology Study was supported by funds from the President’s Emergency 37 Plan for AIDS Relief (PEPFAR) through a cooperative agreement between the U.S. Centers for 38 Disease Control and Prevention and the University of Zimbabwe Department of Community 39 Medicine SEAM Project under the terms of Cooperative Agreement Number: 1U2GGH000315- 40 01. 41 2 42 Brief Summary: Among 200 women and men presenting with genital ulcer disease at a diverse 43 group of Zimbabwe STI clinics, 35% of patients tested positive for herpes simplex virus and 44 11.5% for Treponema pallidum, while none tested positive for Haemophilus ducreyi. 45 46 3 47 Abstract 48 Background: In many countries, sexually transmitted infections (STI) are treated 49 syndromically. Thus, patients diagnosed with genital ulcer disease (GUD) in Zimbabwe receive a 50 combination of antimicrobials to treat syphilis, chancroid, lymphogranuloma venereum (LGV) 51 and genital herpes. Periodic studies are necessary to assess the current etiology of GUD and 52 assure the appropriateness of current treatment guidelines. 53 Materials and Methods: We selected six geographically diverse clinics in Zimbabwe serving 54 high numbers of STI cases to enroll men and women with STI syndromes, including GUD. STI 55 history and risky behavioral data were collected by questionnaire and uploaded to a web-based 56 database. Ulcer specimens were obtained for testing using a validated multiplex polymerase 57 chain reaction (M-PCR) assay for Treponema pallidum (TP; primary syphilis), Haemophilus 58 ducreyi (HD; chancroid), LGV-associated strains of Chlamydia trachomatis (CT-LVG) and 59 herpes simplex virus types 1 and 2.). Blood samples were collected for testing with HIV, 60 treponemal and non-treponemal serological assays. 61 Results: Among 200 patients, 77 (38.5%) were positive for HSV, 32 (16%) were positive for 62 TP, and 2 (1%) positive for CT-LGV. No HD infections were detected. No organism was found 63 in 98 (49%) of participants. The overall HIV positivity rate was 52.2% for all GUD patients with 64 higher rates among women (56% v. 44%, p<0.05) and among patients with HSV (68.6% vs. 65 41.8%, p<0.0001). Among patients with GUD, 54 (28%) had gonorrhea and/or chlamydia 66 infection. However, in this latter group, 66.7% of women and 70% of men did not have abnormal 67 vaginal or urethral discharge on examination. 4 68 Conclusions: HSV is the most common cause of GUD in our survey, followed by TP. No cases 69 of chancroid were detected. The association of HIV infections with HSV may suggest high risk 70 for co-transmission, however some HSV ulcerations may be due to HSV reactivation among 71 immunocompromised patients. The overall prevalence of gonorrhea and chlamydia was high 72 among patients with GUD and the majority of them did not meet criteria for concomitant 73 syndromic management covering these infections. 74 5 75 Introduction 76 Sexually transmitted infections (STIs) associated with genital ulcer disease (GUD) continue to be 77 an important cause of morbidity and mortality worldwide. While the etiology of GUD varies in 78 different parts of the world, herpes simplex virus is considered the leading cause of this 79 syndrome globally. In a recently published study1, the World Health Organization estimates that 80 in 2012 19.2 million people aged 15-49 were newly infected with herpes simplex virus type 2 81 (HSV-2) resulting in 417 million people aged 15-49 years living with HSV-2 infection, for a 82 global prevalence of 11.3%. Adding an increasing number of genital herpes caused by herpes 83 simplex type 1 virus (HSV-1), it is estimates that well over 40-80 million people suffer from 84 recurrent genital herpes in any given year. In addition to this morbidity, HSV-2 infection is an 85 important risk factor for HIV acquisition2,3 and can cause debilitating disease in the maternally 86 exposed fetus and neonate.4,5 The global HSV burden disproportionally affects Africa with over 87 30% of people aged 15-49 years chronically infected.1 88 Likewise, Africa, and specifically sub-Saharan Africa is also disproportionally affected by 89 syphilis. While the incidence of syphilis is substantially lower and has been declining over recent 90 years, it is estimated that in 2012, 5.6 million incident syphilis cases occurred worldwide of 91 which 32% in Sub-Saharan Africa.6 Syphilis continues to be a major global public health 92 problem. In 2012, an estimated 930,000 maternal syphilis infections caused 350,000 adverse 93 pregnancy outcomes, including 143,000 early fetal deaths and stillbirths, 62,000 neonatal deaths, 94 44,000 preterm or low weight births, and 102,000 infected infants worldwide.7 95 6 96 The relative proportion of GUD caused by chancroid and lymphogranuloma venereum (LGV) 97 appears to be decreasing. For example, in a study from Botswana in 2002, 2% of patients with 98 GUD were infected with Treponema pallidum and 1% with Haemophilus ducreyi, compared to 99 58% with HSV-2.8 In a study from Malawi conducted between 2004 and 2006, the proportion of 100 GUD caused by H. ducreyi was higher (15%) compared to T. pallidum and LGV strains of 101 Chlamydia trachomatis (both at 6%) while 67% tested positive for HSV-2.9 In a more recent 102 study from Zambia, none of the patients with GUD had H. ducreyi detected and 3% were 103 positive for LGV strains of C. trachomatis.10 Changes in the underlying epidemiology of GUD 104 have been linked to the change in management when the WHO introduced syndromic 105 management for the treatment of GUD after 2000.11 106 In the absence of etiologic testing, STI reporting and management in many countries, including 107 Zimbabwe, is based on presenting symptoms. Thus, in 2015 a total of 252,406 adult STI cases 108 were reported syndromically to the Zimbabwe Ministry of Health and Child Care, comprising 109 20,063 men and 18,996 women with genital ulcer disease, in addition to 56,540 men with 110 urethral discharge, 79,371 women with vaginal discharge, 27,782 women with PID, and 21,489 111 men and 26,165 women with other STI syndromes, including genital warts.12 112 Syndromic treatment of genital GUD, as recommended by the Zimbabwe Ministry of Health 113 and Child Care (MOHCC), includes the combined use of benzathine penicillin, erythromycin, 114 and acyclovir to treat potential infections with T. pallidum, H. ducreyi, LGV-associated C. 115 trachomatis strains, and HSV respectively.13 However, no recent studies have been undertaken in 116 Zimbabwe in recent years to determine the underlying etiology of GUD and thus assess the 117 adequacy of current treatment guidelines. 7 118 As part of a recently completed STI etiology study in Zimbabwe, the primary aim of this analysis 119 was to determine the prevalence of syphilis, chancroid, lymphogranuloma venereum, and genital 120 herpes infections among women and men presenting with GUD. Secondary aims were to 121 determine the prevalence of vaginal or urethral Neisseria gonorrhoeae and C. trachomatis as 122 well as the prevalence and association of human immunodeficiency virus infections (HIV) with 123 GUD in this population. 124 125 Materials and Methods 126 The Zimbabwe STI Etiology Study was conducted in 2014-2015 to determine the causation of 127 the most important STI-associated syndromes in Zimbabwe: male urethral discharge syndrome, 128 vaginal discharge syndrome, and genital ulcer disease. A full description of study methods is 129 reported elsewhere14 and a copy of the full study protocol is available online.15 The most relevant 130 details are summarized below. Results of the female and male genital discharge syndromes will 131 be reported elsewhere. 132 Clinic Selection 133 Using 2012 surveillance statistics, we identified a regionally diverse sample of six clinics with 134 high numbers of reported STI syndromes. Specifically, we selected two clinics from Harare 135 (Mbare and Budiriru), the country’s capital and largest city in the north-eastern part of the 136 country; two clinics in Bulawayo (Nkulumane and Khami Road) the second largest city located 137 southwest and ethnically distinct from Harare, one clinic from Beitbridge (Dulibadzimu) on the 138 southern border with South Africa, and finally one clinic from rural Gutu (Gutu Road Hospital), 139 located near the country center. These clinics were deemed to be sufficiently representative of 8 140 risky populations regionally to yield meaningful and generalizable results. More details on clinic 141 selection can be found elsewehere.14,15 142 For purposes of the analysis in this manuscript, clinics were combined into 3 regions: Harare, 143 Bulawayo, and Beitbridge/Gutu. 144 Sample Size 145 We aimed to enroll 200 subjects in each of the STI syndromes categories, including 100 men and 146 100 women presenting with GUD. This sample size was deemed to yield sufficiently robust 147 prevalence estimates of the major GUD causes and at the same time be practical from the 148 perspective of the project’s limited budget and enrollment window. Since the study population 149 was, in essence, a convenience sample, we did not intend to apply sample weights.15 150 Patient Selection 151 During the enrolment period at each clinic, all sexually active women and men aged 18-55 years 152 presenting with genital ulcer disease were eligible for the study. Excluded from the study were 153 those who did not speak English, Shona or Ndebele (the 3 major languages in Zimbabwe), those 154 unable to provide consent, those who had received antibiotics for STI treatment in the previous 4 155 weeks, or those previously enrolled in the study. 156 Study Procedures 157 A team of 3 nurses was trained in study procedures and deployed sequentially for a period of 10- 158 17 weeks to each of the 6 study sites, starting in the Harare clinics, then moving to Bulawayo and 159 finally to Beitbridge and Gutu. The start of enrollment at each site was preceded by at least two 9 160 site visits involving the study leadership of the team lead (VK), the lead consultant (CAR) and 161 senior researchers representing the MOHCC (AM and MM). 162 Patients with GUD were enrolled after obtaining informed consent and a paper-copy 163 questionnaire was completed by the study nurse that included demographic information, sexual 164 history, description of symptoms, history of STI/HIV, and current use of medications. GUD 165 patients had swabs taken from the ulcer bases. Vaginal swabs for women and urine samples for 166 men were obtained for chlamydia and gonorrhea testing. Blood specimens were collected from 167 patients for syphilis serology and HIV testing. A separate consent for HIV testing was obtained. 168 Refusing a blood draw was not a reason from exclusion from the study. 169 All patients were treated for their presenting STI syndrome according to the 2013 Zimbabwe STI 170 treatment guidelines.16 After completion of the visit, paper data were reviewed by the lead study 171 nurse and then transcribed into a computer-based data system on handheld devices. Data were 172 uploaded daily from study sites to an online secure central database. 173 Laboratory procedures 174 All specimens were kept refrigerated after collection and shipped in a cooler box with cooling 175 packs by courier overnight to the receiving laboratory at Wilkins Hospital in Harare, where all 176 samples were kept refrigerated until further processing. 177 A number of tests were conducted at the receiving laboratory, including rapid HIV serologic 178 testing using the standard testing algorithm in ZimbabweI, treponemal testing by SD Bioline 1. I HIV testing in Zimbabwe follows a standard algorithm of the following HIV rapid tests: 1) initial test by First Response HIV1-2-O; 2) confirmatory test by Alere Determine HIV1/2 if 10 179 DUO rapid test (Standard Diagnostics Inc, Gyeonggi-do, Republic of Korea) and TPHA, and 180 non-treponemal testing by RPR. These tests were all conducted according to test package inserts. 181 All ulcer samples were stored in a -70 oF freezer and batched for shipment to the STI reference 182 laboratory at the National Institute of Communicable Diseases (NICD) in Johannesburg, South 183 Africa. Using an in-house developed multiplex polymerase chain reaction (M-PCR), ulcer 184 specimens were tested for T. pallidum, H. ducreyi, LGV-associated strains of C. trachomatis, 185 and HSV-1/HSV-2. Test details are described by Lewis et al. elsewhere.17 186 Vaginal and urine samples were shipped to two local laboratories and tested for N. gonorrhoeae 187 and C. trachomatis by nucleic acid amplification testing (NAAT) using Becton Dickenson 188 ProbeTec (BD Molecular Diagnostics, Franklin Lakes, NJ, USA) at the University of 189 Zimbabwe/University of California San Francisco (UZ-UCSF) laboratory in the Obstetrics and 190 Gynaecology Department of the University of Zimbabwe School of Medicine and using 191 GeneXpert (Cepheid, Sunnyvale, CA, USA) at the Flowcytometry laboratory in Harare. All 192 testing was done according to test package inserts under standard operating procedures. More 193 detail on laboratory procedures can be found in the online supplement.15 194 195 196 the initial test is positive, and 3) CHEMBIO HIV1/HIV2 as a tie breaker if the initial and confirmatory tests are discrepant. 11 197 Statistical Methods 198 Data on participant demographics, sexual health and STI history were analyzed using SAS 199 software (Cary, NC, USA). Tests for statistical significance included the Chi-Square test for 200 categorical variables and Student’s T-test for continuous variables. 201 Institutional Review 202 The protocol, including consent forms and questionnaires, was reviewed and approved by the 203 Joint Research and Ethics Committee of Parirenyatwa Central Hospital, the Zimbabwe Medical 204 Research Council and the U.S. Centers for Disease Control and Prevention. 205 206 Results 207 Participant Recruitment 208 Of the 100 women and 100 men with GUD, m-PCR results were available for all ulcer 209 specimens; however, urine samples taken from two men were inadequate for 210 gonorrhea/chlamydia NAAT. Enrollment data by study site are summarized in Table 1. Only 4 211 women and 1 man were recruited at Gutu Road Hospital. The low enrollment at the Gutu site 212 was the result of low STI patient volume and a pragmatic decision to focus our limited resources 213 on sites where patient volume was higher and recruitment more productive. Nonetheless, we did 214 not see any reason to exclude these persons from our analyses. For statistical purposes we 215 included them with the Beitbridge sample. 216 217 12 218 Demographic characteristics 219 As expected, a higher proportion of men and women enrolled at the Bulawayo clinics reported 220 Ndebele ethnicity, as this is the capital of Matabeleland with a prominent Ndebele population. In 221 addition, there was a statistically significant difference across regions in the number of 222 participants who reported more than one sex partner in the previous 3 months, with participants 223 from Beitbridge/Gutu reporting the highest number of partners (p<0.05). There were no 224 significant differences between regions with regards to age, condom use with main or non-main 225 partners, self-reported HIV status, and a history of STD (Table 1). 226 Men in the study were significantly older (mean 29.6; median 28.5) than women (mean 27.9; 227 median 26, p<0.05) and men were also significantly more likely to be unmarried (68% vs. 51%, 228 p<0.05) and less likely to be unemployed (59% vs. %, p<0001) when compared to women. 229 However, there were no differences between men and women with respect to number of sex 230 partners in the past 3 months, condom use with main or non-main partners, self-perceived HIV 231 status, or history of STI (data not shown). 232 Etiology of GUD 233 Overall 77/200 (38.5%, 95% C.I.: 32.0% - 45.5%) of patients with GUD were infected with HSV 234 (all but one typed as HSV-2), and 32/200 (16%, 95% C.I.: 11.6% - 21.7%) had T. pallidum 235 infection detected by m-PCR. Nine patients (4.5%) were dually infected with HSV and T. 236 pallidum. Only 2 patients (1%, 95% C.I.: 0.2% - 3.5%) had C. trachomatis-LGV serovars 237 detected, and none (95% C.I.: 0% - 1.9%) had H. ducreyi infection identified. No organism was 238 found in 49.5% (95% C.I. 42.6% - 56.4%) of participants (Table 2). Of those with any pathogen 239 detected (N=101), 76.2% had HSV, 31.7% had T. pallidum infection and 1.9% had chlamydia13 240 LGV serovars. HSV infection was the most common cause of GUD in both women (39%) and 241 men (38%) followed by T. pallidum infection (13% of women and 19% of men). There was no 242 statistically significant difference between men and women with respect to causative organisms 243 (P=0.88). 244 Multiplex PCR results varied by study site with Harare sites having significantly higher 245 proportions of patients with no pathogen identified (65.3% for Harare sites vs. 37.8% for 246 Bulawayo sites vs 44.4% for Beitbridge, p<0.05). However, the proportion of patients diagnosed 247 with HSV and treponemal infections were not different across sites when excluding patients who 248 had no pathogens identified (data not shown). 249 In a further analysis of patients either infected with HSV or T. pallidum (thus excluding those 250 with mixed infections), we found that shorter duration of symptoms (1-7 vs. >7 days) was 251 significantly associated with detection of HSV, but not T. pallidum. Significantly lower rates of 252 HSV were found in patients concomitantly infected with C. trachomatis. Rates were also lower 253 among patients infected with N. gonorrhoeae, but this difference was not statistically significant. 254 Finally, higher HSV rates were also observed among patients who self-reported as HIV infected 255 (Table 3). 256 Treponemal and Non-Treponemal Test Results 257 Blood specimens from 181 patients were available (19 patients declined to submit a blood 258 sample). Valid RPR results were available for all 181 patients, of which 37 (20.4% were 259 positive) and valid TPHA results for 165 patients, of which 42 (25.4%) were positive. Both RPR 260 and TPHA were positive for 26/165 (15.8%) patients; 15/27 (55.5%) of those testing positive for 261 T. pallidum and 11/138 (7.9%) testing negative (Table 4). 14 262 263 HIV Co-infection 264 HIV test results were available for all 181 patients who submitted a blood sample. Overall 59.8% 265 of women and 45.2% of men with GUD tested positive for HIV (p<0.05). There was no 266 variation by region. Testing results for the causative agent by m-PCR varied significantly by 267 HIV status. When excluding mixed infections (N=170), HIV-positive participants had a higher 268 prevalence of HSV infection compared to HIV-negative participants (53.9% vs. 28.4%, 269 p<0.0001), but similar prevalence of T. pallidum (20.5% vs. 17.2%, Table 3), with no gender 270 differences observed (data not shown). 271 Chlamydia and Gonorrhea Infection 272 Results of genital C. trachomatis and N. gonorrhoeae testing are summarized in Table 5. 273 Overall, 31/100 (31%) of women and 23/98 men (23.5%) of men were NAAT positive for either 274 infection. After excluding those with missing observations, 20 of 30 (67%) women and 14 of 20 275 (70%) men with either infection had no discharge observed. Thus of 50 men and women with 276 GUD and concurrent chlamydia and/or gonorrhea infection, 34 (68%) were not eligible for 277 concomitant genital discharge syndrome management, representing 17% of all patients with 278 GUD in our study. 279 Discussion 280 Our study results confirm the continued high prevalence of HSV infections as the probable cause 281 of genital ulcer disease. They also confirm the decline of chancroid as a cause of GUD as shown 282 previously in other African and non-African countries8,9. In 1979, chancroid accounted for 38% 283 of STI cases in Salisbury (now Harare) but declined to just 3% among men with GUD in 199518, 15 284 and, in the current study, no cases were found. This decrease may be due to either syndromic 285 management used in treating STIs or widespread use of antibiotics treating other infections. The 286 apparently sustained decline in chancroid in Zimbabwe and the southern African region, has 287 called into question whether syndromic treatment guidelines should continue to make specific 288 provisions for treating this disease.10 289 HSV prevalence in our study (38.5%) was lower than found in recent surveys in the southern 290 African region; 60.5% in Botswana (2002)8, 62% in Mozambique (2005)19, 67% in Malawi 291 (2004-2006)9, and 73.6% in South Africa (2005-2006)17, but higher than a more contemporary 292 study in Zambia, Zimbabwe’s neighbor to the north: 28% (2010)10. A common finding among 293 these and our studies is the significant association between HSV and HIV infection; close to 54% 294 of persons with HSV in our study were HIV-infected, compared to 28% without HIV. In our 295 study, we could not distinguish primary HSV infections from recurrences, but in the Malawi 296 study, 75% of HSV infections were recurrences.9 Thus, the prevalence of HSV infection among 297 persons presenting with genital ulcer disease appears to be influenced by recurrent HSV 298 infection among HIV-infected patients, especially those experiencing immunosuppression. As 299 access to HIV antiretroviral therapy is rapidly increasing in many sub-Saharan countries, 300 including Zimbabwe, the epidemiology and relative prevalence of etiologic determinants of 301 GUD may change. 302 While HSV infections may be declining, the relative prevalence of syphilis as a cause of GUD 303 may be increasing. In our study, 16% of patients with GUD tested positive for T. pallidum and 304 an additional 8% of patients submitting blood specimens had negative T. pallidum results but 305 positive RPR and TPHA tests suggesting recent syphilis. Thus, overall, up to 24% of patients 306 with GUD in our study had evidence of active syphilis. 16 307 T. pallidum rates in our study were considerably higher than rates found in the Botswana 308 (5.1%)8, Mozambique (0%)19, Malawi (6%)9 and South African (4.9%)17 studies. However, these 309 studies were conducted 10 to 14 years ago, a timeframe that saw rapid increases in early syphilis 310 cases in the U.S. and Europe. While the recent epidemiology of syphilis in the U.S. and Europe 311 is largely driven by syphilis increases among men who have sex with men, there are now 312 indications that syphilis is also increasing among heterosexuals with subsequent rises in 313 congenital syphilis rates20, and there is no reason the assume that the epidemiology in other 314 world regions, including Africa might be different. Indeed, the more recent study from Zambia 315 also saw higher T. pallidum rates (10%).10 Of course, a study of a highly selective group of 316 patients presenting with STI, cannot be generalized and, as discussed above, the relative 317 proportion of syphilis cases may also be a function of a decline in genital herpes cases in our 318 study. Nonetheless, our study raises the possibility of re-emerging syphilis in Zimbabwe and the 319 southern African region that should be further studied and, in the meantime, should reinforce 320 efforts to screen pregnant women to prevent congenital syphilis. 321 Compared to previous studies in the region, we found a higher proportion of GUD cases with 322 unknown etiology, i.e., 49.5% compared to 39.4% in Botswana8, 30% in Mozambique19, 20% in 323 Malawi9, and 20.6% in South Africa.17 By contrast, the more recent study in Zambia found a 324 higher proportion: 55%10. Reasons for failure to find etiologic pathogens include patient 325 selection, staff training and quality of samples, and test performance. We found significantly 326 lower rates of any pathogens detected in the Harare clinics, and since these were the first clinics 327 to enroll patients, we cannot rule out that staff training, differences in patient selection and 328 quality of ulcer sampling could have played a role. As has been shown in other studies, retrieving 329 etiologic agents, especially HSV, is dependent on how long lesions have been present.10 In our 17 330 study, we confirmed that significantly fewer cases of HSV infection were found in persons with 331 lesions 7 days or older. No such association was found for T. pallidum. 332 Finally, we documented a high prevalence of chlamydia and gonorrhea co-morbidity among our 333 patients with GUD, with 31% of women and 23% of men co-infected. Further analysis indicated 334 that 68% of men and women with co-infections did not have vaginal or urethral discharge and 335 would thus not receive optimal treatment for these infections. These represent 17% of all patients 336 with GUD in this study. This finding suggests that current syndromic management guidelines 337 may be inadequate for women and men with GUD with concurrent gonorrhea and should be 338 considered in the development of syndromic management guidelines.17 339 The findings of our study are subject to a number of limitations. First, we included clinics that 340 also provide care to HIV-infected people. Thus, we may have oversampled HIV-infected people 341 for this study and, because HIV infection was closely related to HSV infections, may have 342 caused a bias in study results. Second, the variation in etiologic patterns by clinic location 343 suggests potential limitations in generalizability. Third, as has already been mentioned, the 344 relatively large proportion of GUD patients with no pathogens identified may indicate biases in 345 patient selection and consistency of study and testing procedures. Finally, by its nature, only 346 symptomatic patients were enrolled in the study, and generalizations outside this group to the 347 general population cannot be made. 348 In conclusion, HSV-2 infection was the most common cause of genital ulcer disease in our study, 349 followed by syphilis. No cases of chancroid were found. The strong association of concurrent 350 HIV and HSV infections may in part be explained by recurrent genital herpes in the 351 immunocompromised patient. Nonetheless, the high co-occurrence of HIV infection in this 18 352 group of patients suggests very high risks of HIV transmission. Patients with genital ulcer 353 disease should therefore receive the highest priority for HIV prevention. 354 19 355 References 356 1. 357 358 Looker KJ, Magaret AS, May MT, et al. Global and Regional Estimates of Prevalent and Incident Herpes Simplex Virus Type 1 Infections in 2012. PLoS One. 2015;10(10):e0140765. 2. Freeman EE, Weiss HA, Glynn JR, Cross PL, Whitworth JA, Hayes RJ. Herpes simplex virus 2 359 infection increases HIV acquisition in men and women: systematic review and meta-analysis of 360 longitudinal studies. AIDS. 2006;20(1):73-83. 361 3. 362 Chen CY, Ballard RC, Beck-Sague CM, et al. Human immunodeficiency virus infection and genital ulcer disease in South Africa: the herpetic connection. Sex Transm Dis. 2000;27(1):21-29. 363 4. Whitley RJ. Neonatal herpes simplex virus infections. J Med Virol. 1993;Suppl 1:13-21. 364 5. Kimberlin DW, Whitley RJ. Neonatal herpes: what have we learned. Semin Pediatr Infect Dis. 365 366 2005;16(1):7-16. 6. Newman L, Rowley J, Vander Hoorn S, et al. Global Estimates of the Prevalence and Incidence of 367 Four Curable Sexually Transmitted Infections in 2012 Based on Systematic Review and Global 368 Reporting. PLoS One. 2015;10(12):e0143304. 369 7. 370 371 Wijesooriya NS, Rochat RW, Kamb ML, et al. Global burden of maternal and congenital syphilis in 2008 and 2012: a health systems modelling study. Lancet Glob Health. 2016;4(8):e525-533. 8. Paz-Bailey G, Rahman M, Chen C, et al. Changes in the etiology of sexually transmitted diseases 372 in Botswana between 1993 and 2002: implications for the clinical management of genital ulcer 373 disease. Clin Infect Dis. 2005;41(9):1304-1312. 374 9. 375 376 377 Phiri S, Zadrozny S, Weiss HA, et al. Etiology of genital ulcer disease and association with HIV infection in Malawi. Sex Transm Dis. 2013;40(12):923-928. 10. Makasa M, Buve A, Sandøy IF. Etiologic pattern of genital ulcers in Lusaka, Zambia: has chancroid been eliminated? Sex Transm Dis. 2012;39(10):787-791. 20 378 11. 379 González-Beiras C, Marks M, Chen CY, Roberts S, Mitjà O. Epidemiology of Haemophilus ducreyi Infections. Emerg Infect Dis. 2016;22(1):1-8. 380 12. Zimbabwe Ministry of Health and Child Care. STI Surveillance Data 2015. Vol Unpublished2016. 381 13. Zimbabwe Ministry of Health and Child Welfare. Management of Sexually Transmitted Infections and 382 Reproductive Tract Infections in Zimbabwe . 383 http://nac.org.zw/sites/default/files/STI%20Guidelines%20MoHCW%20Final%20Draft_05Dec20 384 12.pdf. 385 14. 386 387 2012; Rietmeijer C, Mungati M, Machiha A, et al. The Zimbabwe STI etiology study: design, methods, study population. Submitted. 2016. 15. The Zimbabwe STI Aetiology Study Group. The Aetiology of Sexually Transmitted Infections in 388 Zimbabwe - Study Protocol. 2014; 389 http://www.stdpreventiononline.org/index.php/resources/detail/2039. 390 16. 391 392 Ministry of Health and Child Welfare - AIDS and TB Unit. Management of Sexually Transmitted Infections and Reproductive Tract Infections in Zimbabwe. 2013. 17. Lewis DA, Müller E, Steele L, et al. Prevalence and associations of genital ulcer and urethral 393 pathogens in men presenting with genital ulcer syndrome to primary health care clinics in South 394 Africa. Sex Transm Dis. 2012;39(11):880-885. 395 18. 396 397 affected by HIV infection. Bulletin of the Whorl Health Organization. 2001;79:569-577. 19. 398 399 400 O'Farrell N. Targeted interventions required against genital ulcers in African countries worst Zimba TF, Apalata T, Sturm WA, Moodley P. Aetiology of sexually transmitted infections in Maputo, Mozambique. J Infect Dev Ctries. 2011;5(1):41-47. 20. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance 2015. Atlanta: U.S. Department of Health and Human Services;2016. 401 21 402 403 22 404 405 406 23 407 24 408 409 25 410 26