The management of sexually transmitted infections (STIs), ulcerative or not, among sex workers in Africa was early identified as a cost-effective intervention to reduce the HIV burden in this high-risk group and possibly in the general population.1
In the early 1990s, the implementation in Côte d'Ivoire2 and Benin3 of an HIV and STI prevention program through dedicated “sex workers clinics” yielded excellent results in terms of increased condom use and decreased incidence of HIV and other STIs. As a result, a decade later, the ulcerative diseases most strongly correlated with HIV infection and other bacterial STIs are rare in this group from both settings.2,3
In Bobo-Dioulasso (the second city of Burkina Faso), we showed that commercial sex was mainly driven by local nonprofessional sex workers with a high HIV prevalence (34%)4 in sharp contrast to a relatively stable HIV prevalence of 5.2% (n = 2364) in the general population of the same age categories in 2000.5 In this West African context, what is the current role of bacterial and parasitic STIs in the spread of HIV through high-risk groups? To address this issue, we present the results of a prospective study that assessed STI and HIV incidence among sex workers in Bobo-Dioulasso.
PATIENTS AND METHODS
The study population and the inclusion process were described previously.4 In brief, we identified through a socioanthropologic survey nonprofessional sex workers, who could be bar waitresses, mobile fruit sellers, women making and selling local beer, or students. None of these women self-identified as a proper sex worker, although reporting an average of 2 to 3 clients (men who pay cash for sex) per week. In addition to this first group, professional sex workers consisted of “seaters” (women waiting for clients sitting on a stool in front of their home) and “roamers” (women roaming for clients in streets and nightclubs), with an average of 28 and 18 clients per week, respectively. For the sake of simplicity in this study, we consider sex workers as belonging to the nonprofessional or professional group, as defined by a mean weekly number of clients greater than 5 or 5 or less, respectively. Although the cohort included HIV-positive and HIV-negative women, we only consider the latter for the purpose of this study.
The study design was an open cohort with follow-up visits every 3 months. The study site was based within an open public service facility for medical consultation, environmental hygiene, and vaccinations. Participants were approached by peer educators in their working places and asked to visit the study site the next day. After obtaining informed signed consent, participants were included in the cohort and administered a face-to-face behavioral questionnaire; in particular, the participants were asked whether they felt at risk for HIV infection based on their past sexual behavior. HIV test counseling was then provided before the participants underwent a medical and gynecologic examination. Finally, genital and blood samples were taken for STI and HIV tests. The same procedure was used for follow-up visits.
All participants benefited from information and education sessions on condom use (including negotiation skills) and HIV and STI infections at each follow-up visit at the clinic. These sessions were performed by peer educators recruited and trained for this purpose by the study staff.
The STI laboratory diagnoses were made at Centre Muraz, an international biomedical research center with extensive experience in bacteriology and HIV testing. Cultures were performed on specific media for Neisseria gonorrhoeae (NG), Candida albicans (CA), and Haemophilus ducreyi (HD) detection, although only in case of ulceration for the latter. Identification of Chlamydia trachomatis (CT) was performed by a direct immunofluorescence (DIF) test (Biomérieux SA, France). Diagnosis of Trichomonas vaginalis (TV) was made by means of wet mount direct microscopy by observation of motile flagellated organisms, whereas bacterial vaginosis (BV) was diagnosed on the basis of a Gram-stained vaginal smear using the Nugent method (score of 7 or more).6 Active or recent syphilis was defined as dual seroreactivity to rapid plasma reagin (RPR; Human GmbH, Wiesbaden, Germany) and Treponema pallidum hemaglutination assay (TPHA; Newmarket Laboratories, Kentford, United Kingdom). HIV-1 infection diagnosis was based on 2 complementary enzyme-lined immunoassays (ELISAs) as described and validated earlier.7 Finally, serologic herpes simplex virus-2 (HSV-2) infection was diagnosed by means of a type-specific HSV-2 IgG ELISA test (Kalon Laboratories, United Kingdom). This test showed the best performance from a panel of 12 HSV-2-specific ELISAs tested on African serum.8
The participants' follow-up was censored at the time of HIV seroconversion. The crude HIV incidence was calculated by dividing the total number of seroconverters (between 1998 and 2002) by the total number of person-years of follow-up.
HSV-2 infection was assessed during follow-up. We considered a woman as HSV-2-positive if she was ELISA-positive at baseline or during follow-up.
A proportional hazard model set to control for age was used to assess the factors associated with HIV acquisition. Our main interest was to evaluate the role of STIs as well as to know whether the risk was different among professionals and nonprofessionals. We first performed a univariate analysis that included all biologic and behavioral factors potentially involved in HIV transmission. Only factors showing a probability value below or close to 0.15 were further entered into a multivariate model.
Changes in sexual behaviors were compared at baseline and at 6-month follow-up visits among HIV-negative participants recruited before 2002. Indeed, the women enrolled afterward benefited from several information and education sessions (led by our project staff) before their entry into the cohort, which may have influenced their sexual behaviors.
Data analysis was performed using EPIINFO, version 6.04a, and STATA (Statacorp, College Station, Texas, USA), version 8.
A total of 377 HIV-negative women were included in the cohort between October 1998 and October 2002. A total of 71 women did not come back to the study site after enrollment. The overall median follow-up was 23.4 months, and the mean number of follow-up visits was 7.2 over the study period.
Women lost to follow-up were younger than others (24 vs. 26 years, respectively; P = 0.04). There was no statistical difference at baseline between women lost to follow-up, however, when considering other key variables, such as the weekly number of clients, condom use, length of prostitution, or having a steady partner (data not shown).
Sexually Transmitted and HIV Infections
Overall, bacterial and parasitic STI prevalence was low at baseline (Fig. 1). No case of NG or HD infection was found over the whole study period, and only 1 woman was diagnosed with active syphilis during follow-up.
Because of technical issues, CT infection diagnosis was only available for a period that included baseline and the first 2 follow-up visits. We observed a decline in this infection, reaching a prevalence of 2.7% at the second follow-up visit (see Fig. 1). Similarly, TV infection declined progressively below a prevalence of 4%.
This low baseline STI prevalence remained stable over the study period; there was no difference between the women enrolled in 1998 and 1999 and the women enrolled later on (data not shown). To confirm our results with a more sensitive method, 100 consecutive cervical samples from various follow-up visits were tested by polymerase chain reaction (PCR NG/CT; Roche Amplicor, Germany) in Canada, showing an identical prevalence of 2% for NG and CT infection, indicating that the prevalence of these infections remained low during the course of the study. Figure 1 displays the results of the non-PCR diagnostic tests. The prevalence of HSV-2 infection was 54.7% (n = 329) at baseline.
The comparison of sexual behaviors was available for 253 women seen at enrollment and 6 months later. We observed an increase in the declaration of condom use with first-time or regular clients among nonprofessionals (Fig. 2). At baseline, 98% of professionals already declared that they used condoms consistently with the same type of clients. Condom use with steady partners (boyfriends) improved only among nonprofessionals, although its systematic use was reported by only 48% of women at 6 months.
Factors Associated With HIV Acquisition
A total of 19 women seroconverted to HIV infection over the study period (17 nonprofessionals and 2 professionals), providing a crude HIV incidence of 3.2 per 100 person-years (Poisson 95% confidence interval [CI]: 1.9-4.9).
On the univariate analysis (Table 1), only self-assessment of high risk taken in the past showed a significant association with HIV acquisition. No expected determinants, such as length of prostitution or education, were related to HIV acquisition. Conversely, age less than 17 years at first sex tended to be associated with the outcome, without reaching statistical significance. Similarly, no bacterial or parasitic genital infection was involved in HIV acquisition, which is not surprising considering their relatively low prevalence. Clinical cervicitis (as defined by cervical discharge and/or an inflammatory cervix) and HSV-2 infection were most frequently associated with HIV acquisition with an odds ratio (OR) just less than 2.
In the multivariate analysis, having on average of more than 5 clients per week, self-assessment of high HIV risk in the past, and not having changed a steady partner were independently associated with HIV acquisition (Table 2). HSV-2 infection, clinical cervicitis, and age less than 17 years at first sex were also likely to be linked to HIV seroconversion in the model, without reaching statistical significance.
We observed an unexpectedly low prevalence of bacterial and parasitic STIs among this high-risk group, with little further decrease after regular screening and treatment. Not surprisingly, acquisition of HIV infection was not associated with any bacterial or parasitic STI. Rather, women infected with HSV-2 were more than 2 times as likely to acquire HIV compared with others, although this association did not reach statistical significance. We found a high HIV incidence in this group despite a successful intervention with respect to reported behavioral changes.
Our results are subject to limitations inherent in the study population. Like other studies targeting sex workers in Africa,9 we experienced a relatively high rate of loss to follow-up. Women who left the cohort were not dramatically different from others with respect to baseline social and demographic characteristics, thereby reducing the risk of a major results distortion, although their future risk of HIV infection cannot be predicted. This relatively high rate of loss to follow-up was partly responsible for our small sample size with respect to the determinants of HIV acquisition, which may raise concerns regarding interpretation of the results. We did not find any tendency at all toward an association between bacterial or parasitic STIs and HIV seroconversion, however.
We expected a much higher rate of bacterial STIs among this high-risk group, although our STI diagnostic tests could lack sensitivity: culture methods and PCR techniques to diagnose TV and HD infection, respectively, stand now as the “gold standards.” The recent use of a culture method (InPouch TV; Biomed Diagnostics, White City, Oregon, USA) for other ongoing studies in the same population did not increase the prevalence of TV detection, however, and no genital ulcerations suggesting HD, which is now a common finding in West Africa, were recorded during our study period.2,3
In addition to the low prevalence of bacterial and parasitic STIs, the HIV epidemic is now stabilized in Burkina Faso. This situation is probably less prone to show a marked impact of STI management on HIV incidence, as was suggested to explain, in part, the different results of the 3 main STI community-based intervention trials carried out in Tanzania and Uganda.10,11
We observed a trend toward an independent association between HSV-2 infection and HIV acquisition, although it did not meet statistical significance, probably because of a sample size issue. This finding is consistent with data from other studies performed in East and southern Africa,12-15 where HIV and HSV-2 prevalence in the general population is high; to our knowledge, no such data have been reported to date from West or Central Africa.
The tendency for an association between clinical cervicitis and HIV acquisition is not surprising, because such an association was already described in the early 1990s as a consequence of NG or CT infection. It is more challenging in our context, because these STIs became rare and cannot be responsible for these cases of cervicitis. Other causes, such as human papillomavirus, TV, or mycoplasma infections, could be involved.
Nonprofessional sex workers, as defined by their weekly number of clients, are at a much greater risk of HIV acquisition than professionals after controlling for other factors and despite our relatively small sample size. This is a new finding in Africa, which is largely explained by the high level of condom use of professionals, who adopted a “no condom no sex” policy before enrollment in the cohort. This result emphasizes the need to consider the whole spectrum of commercial sex in programs targeting high-risk women for HIV prevention.
A high HSV-2 prevalence and high HIV incidence revealed the frequency of unsafe sex among our study population. Because core groups tend to “concentrate” STIs, the relative absence of the latter may be explained by a low level of STIs among the general population. This hypothesis was confirmed by a study performed by Centre Muraz in 2000 among the general population of women in Bobo-Dioulasso, which showed a prevalence of syphilis, NG, and CT of 0.1%, 0.5%, and 2% (n = 1004), respectively (unpublished data). A recent review of STI and HIV epidemiologic data provided further arguments for a decline of STIs over the last decade in this city.16 Although data reporting STI trends over time are scarce in West Africa, this situation may well be generalized to other countries from this region. In Côte d'Ivoire2 and Benin,3 recent publications reported a decreased prevalence of STIs among sex workers over time. This downward trend was thought to result from a long-lasting STI and HIV prevention intervention dedicated to this high-risk group, however. Other data come from East and South Africa, where ulcerative STIs also seemed to decline markedly, although remaining at a high level of prevalence.17-19
STI infections were already low at baseline and did not decrease much further over time despite regular treatment and better condom use. Regular reinfection from steady partners may be suggested as a plausible explanation, considering the low frequency of condom use with this kind of sexual partner. Indeed, the latter proved to be difficult to reach during the STI partner notification process.
Study participants declared safer sexual behaviors with sexual partners after information and education sessions. This was more frequently reported by nonprofessionals, whereas professionals had already adopted safe behaviors with their clients at baseline.
The association of not changing a steady partner with HIV acquisition is surprising, and we have few arguments to explain it. Some further work to describe the different types of steady partners (eg, boyfriend, protector) and to understand the kind of relationship the women develop with them is needed to explore this finding.
HIV incidence remained at a high level in the cohort despite increased knowledge of HIV prevention measures and better reported sexual behaviors. The independent and strong association of risk self-assessment with HIV acquisition highlights the vulnerability of these women. Our participants were fully aware that they were likely to become infected with HIV. We are currently performing a qualitative study to explore the social and psychologic mechanisms involved in the self-perception of HIV vulnerability.
Bacterial and parasitic STIs play a minor role in HIV transmission in the current Burkina Faso epidemiologic context. While maintaining interventions to control STIs in all highly exposed women, an additional focus should be on strategies to reach local nonprofessional sex workers and to deliver adequate information and services for prevention to them.
Such programs should include novel intensive peer-delivered HIV and STI prevention activities (including much focused on issues of condom negotiation with sexual partners), education on genital herpes, open access to general medical practice comprising STI management through a classic syndromic approach (adapted to the local STI epidemiology) without further regular screening, and the development of strategies to reach steady partners.
Finally, our data confirm that research on HSV-2 infection should be a priority for Africa.
The authors thank all the Service d'Hygiene team that actively worked on this study. They also thank Michel Alary for performing the NG/CT PCR tests in Quebec, which were financed by the Department for International Development Knowledge Programme of the London School of Hygiene and Tropical Medicine. They express gratitude to Philippe Mayaud for his critical comments on this paper. Finally, they thank all the study participants.
1. Laga M, Alary M, Nzila N, et al. Condom promotion, sexually transmitted diseases treatment, and declining incidence of HIV-1 infection in female Zairian sex workers. Lancet
2. Ghys PD, Diallo MO, Ettiegne-Traore V, et al. Increase in condom use and decline in HIV and sexually transmitted diseases among female sex workers in Abidjan, Cote d'Ivoire, 1991-1998. AIDS
3. Alary M, Mukenge-Tshibaka L, Bernier F, et al. Decline in the prevalence of HIV and sexually transmitted diseases among female sex workers in Cotonou, Benin, 1993-1999. AIDS
4. Nagot N, Ouangre A, Ouedraogo A, et al. Spectrum of commercial sex activity in Burkina Faso: classification model and risk of exposure to HIV. J Acquir Immune Defic Syndr
5. Lagarde E, Congo Z, Meda N, et al. Epidemiology of HIV infection in urban Burkina Faso. Int J STD AIDS
6. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol
7. Meda N, Gautier-Charpentier L, Soudre RB, et al. Serological diagnosis of human immuno-deficiency virus in Burkina Faso: reliable, practical strategies using less expensive commercial test kits. Bull World Health Organ
8. Van Dyck E, Buve A, Weiss H, et al. Performance of 12 commercially available enzyme immunoassays for the detection of antibodies against herpes simplex 2 in African populations. J Clin Microbiol
9. Ghys PD, Diallo MO, Ettiegne-Traore V, et al. Effect of interventions to control sexually transmitted disease on the incidence of HIV infection in female sex workers. AIDS
10. Grosskurth H, Gray R, Hayes R, et al. Control of sexually transmitted diseases for HIV-1 prevention: understanding the implications of the Mwanza and Rakai trials. Lancet
11. Sangani P, Rutherford G, Wilkinson D, eds. Population-Based Interventions for Reducing Sexually Transmitted Infections, Including HIV infection (Cochrane Review)
. The Cochrane Library, no. 4. Chichester, UK: John Wiley & Sons; 2004.
12. Del Mar Pujades Rodriguez M, Obasi A, Mosha F, et al. Herpes simplex virus type 2 infection increases HIV incidence: a prospective study in rural Tanzania. AIDS
13. Wald A, Link K. Risk of human immunodeficiency virus infection in herpes simplex virus type 2-seropositive persons: a meta-analysis. J Infect Dis
14. McFarland W, Gwanzura L, Bassett MT, et al. Prevalence and incidence of herpes simplex virus type 2 infection among male Zimbabwean factory workers. J Infect Dis
15. Kamali A, Nunn AJ, Mulder DW, et al. Seroprevalence and incidence of genital ulcer infections in a rural Ugandan population. Sex Transm Infect
16. Nagot N, Meda N, Ouangre A, et al. Review of STI and HIV epidemiological data from 1990 to 2001 in urban Burkina Faso: implications for STI and HIV control. Sex Transm Infect
17. Kharsany A. Changing aetiology of genital ulcer disease in STD clinic attenders with a risking HIV prevalence. Presented at: XIII International Conference on AIDS; 2000; Durban, South Africa.
18. 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
19. Temmerman M, Fonck K, Bashir F, et al. Declining syphilis prevalence in pregnant women in Nairobi since 1995: another success story in the STD field? Int J STD AIDS