Allen, Caroline F. MA, PhD*; Edwards, Morris MB ChB, MSc†; Williamson, Lisa M. BA, MPhil*; Kitson-Piggott, Wendy MSc‡; Wagner, Hans-Ulrich MD, MTropMed§; Camara, Bilali MPH∥; Hospedales, C. James MBBS, MFPHM‡
Guyana has a population of 767,000, with more than 56% of the population in the age group 15 to 49 years. Gross domestic product per capita was US $1093 in 2001, making this the poorest country in the Caribbean, after Haiti.1 More than a third of the population lives in the urbanized coastal region, mostly in the capital city Georgetown. Although located on the northeast coast of South America, Guyana is generally considered part of the Caribbean region, sharing a similar colonial and political history and being a member of the Caribbean Community. South America differs from the Caribbean in characteristics of the HIV epidemic, having lower HIV prevalence and a concentrated epidemic, notably among intravenous drug users.2 Twelve Caribbean countries have adult HIV prevalence exceeding 1%, indicating a generalized epidemic.3 Surveys among men who have sex with men and sex workers (SW) show substantially higher prevalence. Adult HIV prevalence in Guyana in 2003 was estimated to be 2.5%, with a range between the "low" and "high" estimates of 0.8% to 7.7%.2 In 1996, HIV prevalence among pregnant women was found to be 7.1%, illustrating the Caribbean pattern of vulnerability among sexually active women.4
Although several Caribbean governments are directing HIV prevention interventions toward groups generally recognized as high risk, opportunities to improve effectiveness may have been missed because data are lacking on risk factors and to monitor achievements. There have been few studies to identify factors associated with HIV status in the Caribbean.5 Exceptions are SW surveys in Montego Bay, Jamaica,6 and in Georgetown, Guyana, where SW seem to be highly vulnerable to HIV, with surveys showing prevalence of 25% in 19937 and 46% in 1997.8,9 Our study extends the analysis of the Guyana surveys by studying SW across the whole of Georgetown, including a larger number of possible explanatory variables, and examining synergies between variables found to be associated with HIV in bivariate and multivariate analyses.
Since 1996, the Guyana National AIDS Programme Secretariat (NAPS) has implemented a risk reduction strategy: the Georgetown SW Project. A network was established for the distribution of condoms via brothel owners, street distributors, the genitourinary medicine (GUM) clinic, and health centers. The NAPS and GUM clinic collaborated to improve access to sexually transmitted infection (STI) screening, voluntary counseling and testing for HIV (VCT), and STI/HIV treatment services for SW. Fifteen SW were trained to conduct outreach, during which they distributed condoms, educated SW about HIV/STIs and condom negotiation and other safer sex skills, and referred SW to health services. We report on a survey that aimed to identify risk factors for HIV among SW in Georgetown and assist with the orientation of the SW Project.
In July 2000, respondents participated in anonymous structured interviews and provided oral fluid samples for HIV testing. Eight fieldworkers were recruited, comprising sex workers who had received peer educator training and members of a women's organization with experience of working with SW.
The NAPS and fieldwork team mapped locations of sex work by visiting sites and talking to people in the locality. Aiming to cover all main sex work locations, fieldwork included more prosperous areas in the north of Georgetown as well as the poorer, "downtown" areas where the 1997 survey was conducted.8 Respondents were accessed by snowballing starting by including SW at known sex work sites, with researchers seeking information on other SW from the workers themselves.10,11 The survey took place over 3 weeks and was conducted at night and during the day to enhance coverage of the sex worker population.
The major risk from this research was of social harm: that HIV-positive status would be revealed with resulting discrimination. Respecting the choices of research subjects, we consulted sex workers in the process of research design. Although they were first informed of standard procedures planned to safeguard anonymity, they maintained that they would not accept research that involved collection of personal identifiers that may have enabled HIV status to be revealed. They advised that research participants be counseled to seek VCT and be guided on how to access it and other sexual and reproductive health services. Fieldworkers provided this counseling, thus enhancing access to publicly available services, and provided condoms. Each potential respondent was informed of the nature of the study and the procedures to safeguard confidentiality; consent was obtained for all participants. Respondents were paid a fee of around US $8 as compensation for opportunity cost of participating in the survey.
Oral fluid samples were collected from each study participant with the FDA-licensed Orasure sampling device manufactured by Epitope Inc (Beaverton, OR). Procedures for collection of gingival/oral fluid followed manufacturer's instructions. Laboratory work was conducted at the Caribbean Epidemiology Centre (CAREC). Oral fluid extract was screened with the Organon Teknika (Boxtel, the Netherlands) Vironostika Uniform 11 + O HIV 1 & 2 enzyme-linked immunosorbent assay. Reactive specimens were confirmed with the Epitope oral fluid Western Blot confirmatory assay. A system of consecutive study numbers enabled prevalence and interview data to be matched by respondent while maintaining respondent anonymity.
Data analysis was conducted using Epi Info 6 (Centers for Disease Control, Atlanta, GA) and SPSS version 9 software (SPSS Inc, Chicago, IL). The Pearson χ2 test was used for bivariate comparisons of categorical data. Means of continuous data were compared using t tests. The statistical significance level was set at 5%. Multivariate logistic regression was used to produce adjusted odds ratios (ORs) and to assess their significance, where numbers permitted (some variables were excluded where numbers were too small for statistical analysis).
Ethical approval for the study was obtained from the Ministry of Health, Guyana, and the CAREC Ethics Review Committee.
All women approached for interview agreed; there were no refusals. The volume of the oral fluid samples from 47 women (of sample N = 299) was inadequate for testing. Of those who were tested (n = 252), 30.6% were HIV positive [95% confidence interval (CI) 24.9-36.3], 65.1% HIV negative (CI 59.2-71.0), and 4.4% indeterminate (CI 1.9-6.9).
Table 1 provides a profile of characteristics and behavior of respondents for selected categorical variables. Reported condom use was high; but with the exception of condom use during group sex, variables indicating condom use and attitudes toward condoms by self or clients were not associated with HIV status.
Respondents' ages ranged from 16 to 50 years, with median age of 29. HIV-positive respondents were significantly older, at average age of 31, than HIV-negative women, at average age of 28. The length of time that respondents reported doing sex work ranged from 1 to 20 years. The average number of years women had been doing sex work was 5.2 among those who were HIV positive and 4.2 among those who were HIV negative; the difference was not significant (P = 0.08). Most women interviewed in the poorer downtown area usually found their clients on the streets or in hotels/brothels (83%) (as opposed to in discos, on ships, via referral, or other), whereas 48% of uptown women found clients in these locations.
Unadjusted ORs were calculated for factors possibly associated with HIV status. Variables found to be significant at the bivariate level were entered into 2 multivariate models. Table 2 shows ORs for HIV risk factors. Table 3 shows ORs for health service use factors associated with HIV status. Age and location of interview (an indicator of socioeconomic status) were entered into both multivariate models because it was hypothesized that they may be associated both with other risk factors and with service use patterns.
In the multivariate risk factor model, only having a vaginal ulcer or sore in the last 12 months remained significant when combined with other variables. In the health service use model, last having an HIV test more than 6 months ago; not going back for the results of the last HIV test; and getting condoms from the GUM clinic, health center, NAPS, street distributor or a brothel (ie, from public sector STI services) remained significant.
We explored interaction effects on HIV status between these variables and those found significant in bivariate analyses, but confidence intervals were too wide to draw firm conclusions given the small number of cases. Tests of association between variables found significant in the multivariate model and the bivariate analyses revealed the following as significant. Having had an ulcer or sore on the vagina in the last 12 months was associated with having ever tried cocaine or crack and having had treatment for syphilis in the past 12 months. Ever having been tested for HIV (whether within the last 6 months or longer) was associated with having visited the GUM clinic in the past 12 months. Getting condoms from public sector STI services was associated with being interviewed in the downtown (poorer) area, currently using nothing to prevent pregnancy, ever having tried cocaine or crack, having visited the GUM clinic in the past 12 months, having been treated for syphilis in the last 12 months, and having larger numbers of clients per day. Not knowing the results of the last HIV test was associated with having larger numbers of partners per day or night and never using contraception (P < 0.05).
Findings from Guyana, including our own, suggest that the profile of the HIV/AIDS epidemic is similar to that in West Africa, where prevalence in the general adult population is lower than that in East and Southern Africa (≤7%), transmission is largely heterosexual, and high prevalence (>30%) is found among SW. In such circumstances, SW constitute a "core group" for HIV transmission.12,13 At this stage in the epidemic, HIV prevention interventions with SW are particularly important to public health.
In contrast with other studies,14-16 SW mobility was not a risk factor for HIV. International mobility in our sample was within the Caribbean and other non-Spanish-speaking American countries, reflecting the social links of Guyana and implying links of the HIV epidemic with these countries rather than with Spanish Latin America.
Our bivariate analyses identified a number of factors such as age, number of clients, anal sex, low socioeconomic status, and STI, which other studies have also identified as risk factors for HIV among SW.6,12,14,16-27 Condom use was not related to HIV status, in contrast with other studies14,16 and suggesting some overreporting. In our study, only genital ulceration was independently associated with HIV status in multivariate analysis. The cross-sectional design of this study prevents a judgement of whether vaginal ulceration was a risk factor for HIV or a result of HIV immunosuppression.28 Either way, it indicates the need for an active program of screening and treatment of ulcerative STIs.29-31
Sexual behavioral risk factors (nonuse of contraception, anal sex, alcoholic inebriation, cocaine use, not changing condoms during group sex) applied to minorities of the women studied. Nevertheless, they suggest areas where health education messages should focus and that it is important to identify and reach highly vulnerable minorities. Nonintravenous drug use has rarely been identified as a risk factor, but Caribbean SW studies confirm associations between crack cocaine use and risky sexual practices,32,33 and that cocaine use is a risk factor for HIV.6,8 We also found that 34% of respondents had ever used marijuana; and in the last year, 7% had exchanged sex for drugs and 1% had injected themselves with drugs. Locations of cocaine and other illicit drug trade/consumption should be targeted, as should sex workers with alcohol or cocaine addiction. The association between reported vaginal ulceration and crack cocaine use emphasizes the need for the development of joint strategies between substance abuse and STI services targeting SW.34 The importance of drug use in the HIV epidemic may increase with the expansion of the drug trade in the Americas.
The inclusion of health service use factors enabled the appraisal of how well services were targeting women at high risk. The association of HIV status with source of condoms indicates that the Georgetown SW Project, which includes condom distribution via the GUM clinic, health centers, brothels, NAPS and street distributors, was successfully reaching women with HIV. The association between where condoms were obtained and factors such as drug use and number of partners shows women with high rates and high risk of infection were apparently heeding health education messages and accessing condoms from the project. Comparison with the 1997 study conducted in the downtown area of the city8 suggests the project may have led to a drop in HIV prevalence downtown from 45.8% to 36.2% (CI 29.8%-43.2%) in our study.
Women with HIV were more likely to report that they did not go back for their HIV test results and that their test was not within the last 6 months. Not going back for results was in turn associated with indicators of vulnerability such as higher numbers of partners and nonuse of contraception. Women who did not seek their results may have feared the consequences of finding out they are HIV positive at a time when antiretroviral treatment was not available to SW. Our results suggested that sensitively designed follow-up services should be strengthened, along with building trust in the confidentiality of services, to encourage women to discover their HIV status and adopt preventive measures.
Initiatives conducted since our study include a program to make antiretroviral treatments available to all people living with HIV/AIDS in Guyana, without discrimination (since April 2002). In 2003, the NAPS initiated the training of doctors throughout Guyana in HIV/AIDS care and support. The Pan-Caribbean Partnership on HIV/AIDS and the Caribbean Community have established a program on HIV/AIDS law, ethics, and human rights, leading to national policies and proposals for law reform in Guyana and other countries.35 These initiatives may assist in reducing discrimination against sex workers, particularly those with HIV, and enhance their access to and use of STI services, including VCT. However, no further measurement of HIV epidemiology among Georgetown sex workers has been conducted since our study. An assessment by the Canadian HIV/AIDS Legal Network reported that there is still a widespread perception of discrimination harbored by persons living with HIV/AIDS, and some health care workers remain reluctant to work with them.36 In these circumstances, initiatives to reduce discrimination and improve care and support may take some time to affect HIV risk behavior and access to services among sex workers and other vulnerable groups.
Thanks to the survey participants, and to Karen De Souza and Dusilley Cannings of Red Thread Women's Development Programme for their contributions to survey design and management. Shirley Goodman, Phil Pilgrim, Olinda Jacobs, Cora Belle, Vanessa Ross, Halima Khan, Jacqueline Delph, and Cheryl Benjamin collected the data. We are grateful to the Ministry of Health, Guyana, and to CAREC for permitting publication and for the contributions of staff to design, data entry, and administration. The research was funded by the CAREC German Technical Cooperation project. The office of the PAHO/WHO representative in Guyana kindly administered and disbursed funds for the local components of the research.
1. ECLAC. Statistical yearbook for Latin America and the Caribbean 2003. Santiago de Chile: Economic Commission for Latin America and the Caribbean; 2004.
2. UNAIDS. 2004 report on the global AIDS epidemic, ed. J.U.N.P.o.H. AIDS. 2004, Geneva.
3. Camara B. Twenty years of the HIV/AIDS epidemic in the Caribbean: a summary. Port of Spain, Trinidad and Tobago: Special Programme on Sexually Transmitted Infections, Caribbean Epidemiology Centre/Pan American Health Organisation/World Health Organisation; 2002.
4. Camara B. Nineteen years of the HIV/AIDS epidemic in the Caribbean: a summary. Port of Spain, Trinidad and Tobago: Special Programme on Sexually Transmitted Infections, Caribbean Epidemiology Centre/Pan American Health Organisation/World Health Organisation; 2001.
5. Allen CF. Gender and the transmission of HIV in the Caribbean. The Society for Caribbean Studies Annual Conference Papers Online, 2002. 3.
6. Douglas KG, et al. Risk factors associated with STD/HIV seropositivity among female commercial sex workers (FCSWs)-implications for an integrated intervention strategy. West Indian Med J. 1996;46(suppl 2):12-13.
7. Carter K, et al. HIV risk perception, risk behavior, and seroprevalence among female commercial sex workers in Georgetown, Guyana. Rev Panam Salud Publica. 1997;1(6):451-459.
8. Persaud NE, et al. Drug use and syphilis. Co-factors for HIV transmission among commercial sex workers in Guyana. West Indian Med J. 1999;48(2):52-56.
9. Persaud NE. Ethnic differences in HIV risk among female commercial sex workers in Guyana: a case-control study. Miami: University of Miami; 2003.
10. Camara B, et al. Guidelines for upgrading of HIV/AIDS/STI surveillance in the Caribbean: the third generation surveillance of HIV/AIDS/STI, linking case-reporting, behavioural and care surveillance. Port of Spain, Trinidad and Tobago: Special Programme on Sexually Transmitted Infections, Caribbean Epidemiology Centre/Pan American Health Organisation/World Health Organisation; 2002.
11. Family Health International. Behavioral surveillance surveys: guidelines for repeated behavioral surveys in populations at risk of HIV, ed. F.H. International. 2000. United States Agency for International Development and UK Department for International Development.
12. Asamoah-Adu C, et al. HIV infection among sex workers in Accra: need to target new recruits entering the trade. J Acquir Immune Defic Syndr. 2001;28(4):358-366.
13. Plummer F, et al. The importance of core groups in the epidemiology and control of HIV-1 infection. AIDS. 1991;5(suppl 1):169-176.
14. Aklilu M, et al. Factors associated with HIV-1 infection among sex workers of Addis Ababa, Ethiopia. AIDS. 2001;15(1):87-96.
15. Brewer TH, et al. Migration, ethnicity and environment: HIV risk factors for women on the sugar cane plantations of the Dominican Republic. AIDS. 1998;12(14):1879-1887.
16. Thuy NTT, et al. HIV infection and risk factors among female sex workers in southern Vietnam. AIDS. 1998;12(4):425-432.
17. Celentano DD, et al. HIV-1 infection among lower class commercial sex workers in Chiang Mai, Thailand. AIDS. 1994;8(4):533-537.
18. Cohen CR, et al. Bacterial vaginosis and HIV seroprevalence among female commercial sex workers in Chiang Mai, Thailand. AIDS. 1995;9(9):1093-1097.
19. Fonck K, et al. Sexually transmitted infections and vaginal douching in a population of female sex workers in Nairobi, Kenya. Sex Transm Infect. 2001;77(4):271-275.
20. Ford K, et al. AIDS and STD knowledge, condom use and HIV/STD infection among female sex workers in Bali, Indonesia. AIDS Care. 2000;12(5):523-534.
21. Laga M, et al. Condom promotion, sexually-transmitted diseases treatment, and declining incidence of HIV-1 infection in female Zairian sex workers. Lancet. 1994;344(8917):246-248.
22. Lankoande S, et al. Prevalence and risk of HIV infection among female sex workers in Burkina Faso. Int J STD AIDS. 1998;9(3):146-150.
23. Limpakarnjanarat K, et al. HIV-1 and other sexually transmitted infections in a cohort of female sex workers in Chiang Rai, Thailand. Sex Transm Infect. 1999;75(1):30-35.
24. Lurie P, et al. Socioeconomic status and risk of HIV-1, syphilis and hepatitis B infection among sex workers in Sao Paulo State, Brazil. AIDS. 1995;9(1):S31-S37.
25. Onorato IM, et al. Prevalence, incidence, and risks for HIV-1 infection in female sex workers in Miami, Florida. J Acquir Immune Defic Syndr Hum Retrovirol. 1995;9(4):395-400.
26. Siraprapasiri T, et al. Risk factors for HIV among prostitutes in Chiangmai, Thailand. AIDS. 1991;5(5):579-582.
27. van Griensven GJP, et al. Socioeconomic and demographic characteristics and HIV-1 infection among commercial sex workers in Thailand. AIDS Care. 1995;7(5):557-565.
28. Dickerson MC, et al. The causal role for genital ulcer disease as a risk factor for transmission of human immunodeficiency virus. An application of the Bradford Hill criteria. Sex Transm Dis. 1996; 23(5):429-440.
29. Kaul R, et al. Monthly antibiotic chemoprophylaxis and incidence of sexually transmitted infections and HIV-1 infection in Kenyan sex workers: a randomized controlled trial. JAMA. 2004;291(21):2555-2562.
30. Orroth K, et al. Syndromic treatment of sexually transmitted diseases reduces the proportion of incident HIV infections attributable to these diseases in rural Tanzania. AIDS. 2000;14(10):1429-1437.
31. Steen R, Dallabetta G. Sexually transmitted infection control with sex workers: regular screening and presumptive treatment augment efforts to reduce risk and vulnerability. Reprod Health Matters. 2003;11(22):74-90.
32. Hope Enterprises Ltd. Report of a knowledge, attitudes, behaviour and practices study among female commercial sex workers, conducted as part of the evaluation of the USAID/AIDSCAP/Jamaica project. Kingston: FHI/AIDSCAP in collaboration with the Epidemiology Unit, Ministry of Health; 1996.
33. Persaud NE, et al. Sexually transmitted infections, drug use, and risky sex among female sex workers in Guyana. Sex Transm Infect. 2000;76:318.
34. Reid SD. HIV seroprevalence and risk factors in female substance abusers seeking rehabilitation in Trinidad and Tobago. West Indian Med J. 2004;53(3):155-158.
35. Patterson D. Caribbean: regional support for HIV/AIDS law and policy reform. Canadian HIV/AIDS Policy & Law Review. 2004;9(1):29-31.
36. Bulkan A. HIV/AIDS, law, and discrimination in Guyana. HIV/AIDS Policy & Law Review/Canadian HIV/AIDS Legal Network. 2004;9(3):32-34.
© 2006 Lippincott Williams & Wilkins, Inc.