THE PRESENCE OF sexually transmitted infections (STIs) is a risk factor for HIV transmission, not least because of its association with increased viral shedding. 1 Studies 1,2 have demonstrated that treatment of gonorrhea in women is associated with decreased HIV shedding in cervicovaginal secretions. In addition, improved treatment of STIs was associated with a 40% decrease in the incidence of HIV infections. 3 A number of investigators have reported on the high rates of STIs in the sex worker population and the role of sex work in the spread of HIV infections. 1,4,5 Consequently, treatment of STIs is critical to the prevention and control of HIV infection.
In the KwaZulu-Natal province of South Africa, antenatal surveys have found that the prevalence of HIV in the year 2000 was 36%. 6 Among sex workers in the region, the HIV prevalence was already over 50% in 1998, and the prevalences of STIs due to specific pathogens were as follows: 41.3% for Trichomonas vaginalis, 14.3% for Neisseria gonorrhoeae, and 16.4% for Chlamydia trachomatis infections. 7 Distinctly lacking, however, are data on the success of STI prevention efforts among HIV-positive sex workers. We therefore measured the incidence of STI infection in HIV-positive individuals. The objective of this study was to determine the incidence of three STIs (T vaginalis, N gonorrhoeae, and C trachomatis infections) in a cohort of HIV-positive sex workers, recruited from truck stops in the KwaZulu-Natal region, and to determine whether risk factors commonly associated with highly prevalent STIs are also independent risk factors for incidence cases.
During 1996 to 1999, sex workers working at truck stops in the KwaZulu-Natal midlands were screened for participation in a phase III efficacy trial of a vaginal microbicide, COL-1492. Women were recruited and screened for HIV and other exclusion criteria before admission in the study. Details of the recruitment strategies are described elsewhere. 7 Of the 472 women screened, 209 were HIV-negative and were excluded from this study. Of the 263 HIV-positive women, 88 returned for at least 1 follow-up visit. For the purposes of this study, women who missed more than two consecutive visits were excluded. Seventy-seven women were included in the final analysis. Compared with those excluded, they were slightly older (P < 0.001). Otherwise, characteristics such as duration of sex work, number of partners per week, practice of anal and oral sex, and baseline prevalence of STIs were similar to those of excluded women.
In the study group, condom use among women at screening was <50%. Baseline prevalences of STIs were high:T vaginalis, 25%;C trachomatis, 5%; syphilis, 68%; and N gonorrhoeae, 10%. Women who were HIV-positive were recruited to ascertain the incidence of STIs in this group. The women who had STI at baseline screening were treated for infection. Thereafter, they were observed monthly. At each visit, women received health education and free condoms. In addition, they were counseled regarding their HIV status and urged to stop sex work. In addition to the extensive counseling provided in the study, the women were referred to HIV counseling clinics in the vicinity of the truck stops. After obtaining informed consent, the trained research staff conducted structured interviews, and clinical examinations were performed by a clinician. All STIs diagnosed at baseline or on follow-up visits were treated according to South African guidelines for syndromic management. Women were counseled to discuss infections with their partners, but steady partners were not treated, as the men were not aware that their partners were sex workers. The study was approved by the University of Natal Ethics Committee.
T vaginalis was identified from wet smear microscopy specimens inoculated into modified Diamond's medium for culture. N gonorrhoeae was identified in endocervical swabs with use of modified New York City medium, and C trachomatis was detected with direct immunofluorescence. Antibody to HIV was detected by means of a recombinant HIV-1/HIV-2 ELISA (Abbott, Chicago). The second ELISA used was the Vironostika HIV Uniform II Micro-ELISA 4 system (Omnimed, United Kingdom).
Definition of a New Case
The distinction between a new case and treatment failure is not clear-cut, particularly in a high-risk cohort such as commercial sex workers. In order to avoid overestimating or underestimating incidence, the analysis was performed by first counting all positive tests as new cases (maximum) and subsequently counting a positive test as a new case only if preceded by a negative test (minimum).
Significant risk factors associated with the prevalence at baseline of three STIs (T vaginalis, N gonorrhoeae, and C trachomatis infections) were identified with chi-square statistical tests. Condom use was recorded at baseline and at each monthly visit. Due to the variability in monthly condom use, an estimate of median use during the whole follow-up period was used in the analysis.
Incidence rates per 100 women-years were calculated for all newly diagnosed episodes with use of the maximum and minimum criteria specified above. Confidence intervals for incidence rates were based on the Poisson assumption. Time until each infection, measured from enrollment, was analyzed with a Cox proportional hazards model, stratified by number of previous infections. 8 Three-month incidence rates were also calculated to compare the results with those of other studies. Data were analyzed with use of Stata 6.
Demographic and Behavioral Risk Factors
The 77 female sex workers included in the analysis were followed up for 58.1 women-years (median duration per woman, 8.1 months; range, 1–20.7 months). The mean age of the women was 27.9 years (SD, 6.5) within a range of 19 to 50 years. The median duration of sex work was 30 months (range, 1–105 months). The median number of commercial partners per week was 20, within a range of 5 to 40. At baseline, 30% of the women reported using condoms more than 50% of the time. This increased to 80% during the follow-up period (P < 0.001). Women were also tested for syphilis: 18% tested negative; 50% had a titer between 1:1 and 1:4; 18% had a titer of 1:8; and 13% had a titer ≥1:16.
Incidence and Prevalence of T vaginalis,C trachomatis, and N gonorrhoeae
At enrollment, the most common STI of the three studied was T vaginalis infection (25%). None of the risk factors examined (education, use of condoms, duration of sex work, number of partners, and age) were significantly associated with the prevalence of T vaginalis, C trachomatis, or N gonorrhoeae. However, women with a baseline rapid plasma reagin titer ≥1:8 had a greater risk of T vaginalis infection (RR: 3; CI, 1.4–6.6) than women with a low titer.
The numbers of new infections, as defined on the basis of the maximum and minimum criteria, are presented in Table 1. There were 89 new T vaginalis infections in the study period, giving an incidence rate of 150 per 100 women-years (CI: 120–180). On the basis of the minimum criteria the incidence would be 130 (CI: 100–160). Thirty women (39%) remained negative throughout the study period. As expected, reinfection was strongly associated with length of time in the study. The 3-month incident rates show a sharp increase for nine months and then an equally sharp decline (Figure 1). In a Cox proportional hazards model, none of the measured risk factors were significantly associated with an increased risk of new infection of T vaginalis. When condom use was examined during each 3-month period, the incidence of T vaginalis infection among women who reported 100% condom usage was similar to that among women reporting less than 100% condom usage for each period.
At enrollment, only four women (5%) had C trachomatis infection. The incidence rate was 30 per 100 women-years (CI: 14–42; with minimum estimate of 24, 95% CI = 11–37). Sixty-four women (83%) remained negative throughout. Three-month incidence rates increased until 9 months, followed by a sharp decline Figure 1. None of the measured risk factors were associated with an increased risk of acquiring a new infection.
The prevalence of N gonorrhoeae at baseline was 10% (8 women). The incidence rate was 66 per 100 women-years (CI: 45–87; with minimum estimate of 56, 95% CI = 37–75). Fifty-seven women (74%) remained negative throughout the study period. Three-month incidence rates showed a steady increase until 15 months, followed by a decline.
Women younger than 30 years of age (RH: 3.2; CI: 1.4–8) showed a greater risk of reinfection, as did women with infrequent condom use at baseline (RH: 2.7; CI: 1–7).
Acquisition of Any STI
During follow-up, 54 of 77 women (70.1%) acquired at least one STI (T vaginalis, N. gonorrhoeae, or C trachomatis infection). Of the 54 infected women, 17 were infected once, 12 were infected twice, 12 were infected 3 times, 6 were infected 4 times, and 7 were infected 5 or more times. In total, 143 new infections occurred among the 77 women during the 58.1 person-years of follow-up. With use of the minimum criteria, 122 new infections would have been diagnosed. The incidence of newly diagnosed STIs was 244 per 100 women-years (CI: 200–280; with minimum estimate of 210, CI = 170–250).
This longitudinal study of a group of HIV-positive women provides important data pointing to the difficulty in reducing the risk of STIs in HIV-positive sex workers.
This study shows a high incidence of STIs among HIV-positive sex workers despite monthly STI diagnosis and treatment that included health education and intensive condom promotion. In addition, the study raises concern that incidence actually increased during the first 9 months of the study. A similar study, of HIV-negative sex workers in Zaire, found a decreased incidence of N gonorrhoeae and T vaginalis infections but not C trachomatis infections over a 3-year period. 7 A number of factors could explain the different findings. The frequency of visits was monthly in this study, as opposed to every 3 months. Thus, new infections were detected and treated within a shorter time interval, and in a cohort with high exposure due to multiple partners, the probability of detecting new infections is greater. Finally, sex workers in this study worked at truck stops located between two large cities, unlike those in Zaire, who were working in the city of Kinshasa. This may have had an influence on the type and regularity of their clients and their exposure to new infections.
Detection and treatment of STIs in commercial sex workers as well as the continued promotion of condom use have been shown to reduce HIV risk. 9 However, the continued high incidence of STIs in spite of a substantial intervention strategy raises concerns about appropriate interventions for the control of HIV and STI transmission in this high-risk population. While the increased use of condoms during follow-up and the ultimate decline in incidence suggest the intervention may have had an impact on behavior, the high incidence of STIs suggests that unprotected sex is still occurring and that these sex workers are not only highly vulnerable to new infections but also at increased risk of transmitting HIV and STI infections to their sexual partners.
Furthermore, the findings of a study conducted by Morar et al 10 of a sex workers cohort suggested that women who were HIV-positive were not willing to give up sex work after knowing their status and were not willing to insist on condom use with their clients, despite intensive counseling. The women believed that sex work was essential for their economic survival. Sex with condoms fetched a lower price than sex without condoms. In addition, the same author showed that intravaginal douching with harmful substances was high among this group of women. 10 Previous studies have shown that vaginal douching increases risk of pelvic inflammatory diseases and STIs. 11
A recent study conducted by Ramjee and Gouws 12 showed that HIV prevalence among truck drivers frequenting sex workers at the same truck stops was 56%. It appears that both sex workers and truck drivers who may be HIV-negative have a high risk of acquiring HIV infections. These individuals travel to neighboring provinces and countries, and the resultant sexual networking may be facilitating the spread of HIV infections in the country.
The findings of this study document the high incidence of STIs among HIV-positive sex workers. The rate of rapid reinfection in the short term emphasizes the urgent need for an innovative program of service and prevention directed at sex workers based in areas of sex work such as truck stops and at times of peak activity. Different approaches are also needed to empower women to seek other lines of work.
1. Ghys PD, Fransen K, Diallo MO, et al. The associations between cervicovaginal HIV shedding, sexually transmitted diseases and immunosuppression in female sex workers in Abidjan, Cote d'Ivoire. AIDS 1997; 11: F85–F93.
2. Mostad SB, Overbaugh J, DeVange DM, et al. Hormonal contraception, vitamin A deficiency and other risk factors for shedding of HIV-1 infected cells from the cervix and vagina. Lancet 1997; 350: 922–927.
3. Grosskurth H, Mosha F, Todd J, et al. Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomised controlled trial. Lancet 1995; 346: 530–536.
4. Morison L, Weiss A, Buve A, et al. Commercial sex and the spread of HIV in four cities in sub-Saharan Africa. AIDS 2000; 15 (suppl 4): S61–S69.
5. Kilmarx PH, Kimpakarnjanarat K, Mastro TD, et al. HIV-1 seroconversion in a prospective study of female sex workers in Northern Thailand: continued high incidence among brothel-based women. AIDS 1998; 12: 1889–1898.
6. Department of Health. National HIV, syphilis sero-prevalence survey of women attending public antenatal clinics in South Africa. Available at: http://www.doh.gov.za
7. Ramjee G, Abdool Karim SS, Sturm AW. Sexually transmitted infections among sex workers in KwaZulu-Natal, South Africa. Sex Transm Dis 1998; 25: 346–349.
8. Prentice RL, Williams BJ, Peterson AV. On the regression analysis of multivariate failure time data. Biometrika 1981; 68: 373–379.
9. 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 1994; 344: 246–248.
10. Morar NS, Ramjee G, Abdool Karim SS. Vaginal insertion and douching practices among sex workers at truck stops in KwaZulu-Natal. S Afr Med J 1998; 88: 4.
11. Foxman B, Aral SO, Holmes KK. Interrelationships among douching practices, risky sexual practices, and history of self-reported sexually transmitted diseases in an urban population. Sex Transm Dis 1998; 25: 90–99.
12. Ramjee G, Gouws E Prevalence of HIV among truck drivers visiting sex workers in Kwazulu-Natal, South Africa. Sex Transm Dis 2002; 29: 44–49