AIDS RESEARCH AND HUMAN RETROVIRUSES Volume 29, Number 3, 2013 ª Mary Ann Liebert, Inc. DOI: 10.1089/aid.2012.0032
EPIDEMIOLOGY
Short Communication HIV Incidence Among Vulnerable Populations in Honduras: Results from an Integrated Behavioral and Biological Survey Among Female Sex Workers, Men Who Have Sex with Men, and Garifuna in Honduras, 2006 Andrea A. Kim,1 Sonia Morales,2 Ivette Lorenzana de Rivera,3 Mayte Paredes,3 Sandra Juarez,4 Berta Alvarez,2 Xin Liu,1 Bharat Parekh,1 Edgar Monterroso,1 and Gabriela Paz-Bailey 2,5
Abstract
Honduras has one of the highest HIV prevalence rates in Central America. Data on HIV incidence are needed to identify groups at greatest need of prevention interventions to inform the national HIV response. We applied a test for recent infection to HIV-positive specimens from a biological and behavioral survey to estimate assayderived incidence among men who have sex with men (MSM), female sex workers (FSW), and the Garifuna population in Honduras. Assay-derived estimates were compared to the mathematically modeled estimates in the same populations to assess plausibility of the assay-based estimates. Assay-derived incidence was 1.1% (95% CI 0.2–2.0) among MSM, 0.4% (95% CI 0.1–0.8) among the Garifuna, and 0% (95% CI 0–0.01) among FSWs. The modeled incidence estimates were similar at 1.03% among MSM, 0.30% among the Garifuna, and 0.23% among FSWs. HIV incidence based on the assay was highest among MSM in Honduras, lowest among FSWs, and similar to modeled incidence in these groups. Targeted programs on HIV prevention, care, and treatment are urgently needed for the MSM population. Continued support for existing prevention programs for FSWs and Garifuna are recommended.
A
mong the seven countries in Central America, Honduras is one of the most affected by HIV with over 1% of the adult population living with HIV infection in the mid-2000s.1 The HIV epidemic in Honduras is concentrated in populations that engage in high-risk behaviors, such as transactional sex and male-to-male sexual behavior. According to the 2001/2002 Estudio Multice´ntrico—a survey of HIV and sexually transmitted infections (STI) among high-risk groups in Central America—in Honduras, men who have sex with men (MSM) had the highest level of HIV infection, ranging from 8% to 16%, followed by female sex workers (FSW), ranging from 7.5% to 13%.2 The epidemic has also been particularly severe in the Garifuna population,3 an indigenous ethnic minority population in Honduras, where reported HIV prevalence rates were as high as 8.4% in 1998 and 4.5% in 2006.4,5
HIV incidence, or the rate of new infections in a population over time, is a critical component of a comprehensive HIV surveillance system. The most common approaches for estimating HIV incidence in populations are through laboratory-based methods and mathematical models of incidence.6 Due to limitations in these approaches, it is recommended that multiple methods for estimating incidence are applied and interpreted in combination to arrive at a best-supported estimate of incidence in the population of interest.7 In 2006, the Ministry of Health of Honduras conducted a biological and behavioral survey among MSM, FSW, and the Garifuna in Honduras. This study collected standardized indicators on behavior, measured the prevalence of HIV and other STI, and estimated HIV incidence using an incidence
1
HHS-Centers for Disease Control and Prevention/Division of Global HIV/AIDS, Atlanta, Georgia. Universidad del Valle de Guatemala, Centro de Estudios en Salud, Guatemala City, Guatemala. 3 Ministry of Health of Honduras, Tegucigalpa, Honduras. 4 HHS-Centers for Disease Control and Prevention/Division of Global HIV/AIDS/Regional Office for Central America and Panama, Guatemala City, Guatemala. 5 Tephinet, Inc. Atlanta, Georgia. 2
516
HIV INCIDENCE AMONG HIGH-RISK POPULATIONS IN HONDURAS assay. This article presents the results of the HIV incidence analysis for the three study populations. The 2006 Honduras biological and behavioral survey is described in greater detail elsewhere.4 Briefly, the survey was conducted in four large urban cities in Honduras (Tegucigalpa, San Pedro Sula, La Ceiba, and Comayagua) for MSM and FSW and five urban and three rural sites for the Garifuna. The MSM and FSW populations were recruited using the respondent-driven sampling (RDS).8,9 Eligible FSWs were women aged ‡ 18 years who reported currently having vaginal or anal sex in exchange for money or work in the surveyed city. Eligible MSM were defined as men ‡ 18 years who reported anal or oral sex with another man during the past 12 months. A two-stage household cluster sampling method was used to recruit eligible Garifuna participants aged ‡ 18 years residing in the home, who self-identified as Garifuna ethnicity, and reported the most recent birthday among all adults in the household.4 After providing consent, participants completed a behavioral interview using audio computer-assisted interview (ACASI) methods and were asked to provide blood samples for HIV testing. HIV testing was conducted on-site using two consecutive HIV rapid tests [Determine HIV-1/2 (Abbott, IL) and OraQuick Rapid HIV-1 Antibody Test]. An enzyme immunosorbent assay was conducted at the national reference laboratory as a tie breaker for discordant results. Participants received their HIV test results with post-test counseling during the study visit. HIV-infected persons were referred to nearby care and treatment facilities for follow-up services. The BED IgG capture enzyme immunoassay (referred to as the BED assay) was applied to HIV-positive samples from the survey to estimate HIV incidence.10 Assay-derived incidence and 95% confidence intervals (CI) were estimated using recommended formulas.11 We calculated HIV incidence using local parameters required for calculating assay-derived incidence, including the assay’s mean duration of recency and false-recent rate (FRR). The mean duration of recency applied was 162 days (95% CI 146–179) to correspond with clade B infections.12 The FRR applied was 10.7% (95% CI 8.9–11.3), a value generated from a sample of known long-standing HIV infections in El Salvador, a neighboring country with HIV epidemic and HIV-1 subtype distribution similar to Honduras. In 2007, HIV prevalence among adults aged 15–49 years was estimated to be 0.7% (0.4, 1.1) in Honduras and 0.8% (0.6, 1.5) in El Salvador.13 In both countries, the HIV epidemic is concentrated among FSW and MSM, and the vast majority of prevalent HIV-1 subtypes in the Central America region is clade B infection.14–16 Assay-derived incidence estimates were adjusted
517
to account for the impact of missing HIV-positive specimens on the final incidence estimate.17 Descriptive analysis was conducted to assess differences in assay-derived incidence estimates by population group. HIV incidence derived from the assay was compared to mathematically modeled estimates of incidence from the 2007 Modes of Transmission (MOT) model to assess the plausibility of the assay-derived estimates. The MOT model is based on a software package developed by the Joint United Nations Programme on HIV/AIDS to estimate expected incidence in the adult population by mode of exposure using data on HIV prevalence, risk of exposure, and population size estimates in specific populations.18 Standard errors were not available for the MOT estimates. A total of 791 FSWs provided specimens for HIV testing, and 19 (2.4%) were HIV-positive. Among these, 16 (84.2%) were available for incidence testing; none tested recent on the assay. Of 812 Garifuna that provided a specimen for HIV testing, 37 (4.6%) were HIV-positive (Table 1).4 Among 32 (86.5%) HIVpositive specimens available for incidence testing, 5 (15.6%) tested recent on the assay. Among 568 MSM who provided specimens for HIV testing, 42 (7.4%) were HIV-positive. A total of 31 (73.8%) HIV-positive specimens were available for incidence testing, and 6 (19.3%) tested recent on the assay. HIV prevalence was 2.4% (95% CI 1.3–3.5) among FSWs, 4.5% (95% CI 3.1–6.0) among the Garifuna, and 7.4 % (95% CI 5.2–9.5) among MSM. Assay-derived incidence was 0% (95% CI 0–0.01%) among FSWs, 1.1% (95% CI 0.2–2.0) among MSM, 0.4% (95% CI 0.1–0.8) among the Garifuna. The proportion of HIV infections that were presumed to be recent based on the results of the incidence assay was 10% among FSWs, 11%, among the Garifuna, and 15% among MSM. Incidence derived from the MOT model was 0.23% among FSW, 0.30% among the Garifuna, and 1.03% among MSM in 2007 (Fig. 1). The rate-ratio of assay-derived incidence to HIV prevalence was approximately 10% for the Garifuna and FSW populations, corresponding to expected patterns of a stable HIV epidemic in a country.19 Among MSM, the assay-derived incidence rate-ratio was 15% that of the HIV prevalence, suggesting a higher vulnerability of this sub-population. The point estimates for assay-derived incidence among MSM increased over 2-fold from 0.5% in 2001(estimated from previously published data, adjusted using the BED FRR applied in this analysis) to 1.1% in 2006.16,20 While this increase may not be statistically significant, we can state with certainty that the rate of new HIV infections has not decreased among MSM in Honduras. In addition, HIV prevalence has not changed during the same time period suggesting that the number of
Table 1. Assay-Derived HIV Incidence Estimates by Population, Honduras, 2006 Population FSW (n = 790)b Garifuna (n = 817)c MSM (n = 568)c
HIV + (N)
HIV + available N (%)
Recent N (%)
HIV prevalence (95% CI)
Assay-derived HIV incidence (95% CI)a
19 37 42
16 (78.9%) 32 (86.5%) 31 (73.8%)
0 5 (15.6%) 6 (19.3%)
2.4 (1.3, 3.5) 4.5 (3.1, 6.0) 7.4 (5.2, 9.5)
0 (0, 0.01) 0.4 (0.1, 0.8) 1.1 (0.2, 2.0)
a Assay-derived HIV incidence estimates calculated using a mean duration of recency of 162 days for HIV-1 subtype B and a regional FRR of 10.7%. Estimates were also adjusted for the impact of missing HIV-positive specimens. b Sites included San Pedro Sula, Tegucigalpa, La Ceiba, and Comayagua for FSW and San Pedro Sula, Tegucigalpa, and La Ceiba for MSM. c Sites included Bajamar, Tornabe´, Triunfo de la Cruz, La Ceiba, Sambo Creek, Corozal, Santa Rosa y la del Aguan, and Limon for the Garifunas. FSW, female sex workers; MSM, men who have sex with men.
518
FIG. 1. Comparison of assay-derived HIV incidence to modeled HIV incidence by population, Honduras, 2006– 2007.a,b new infections and HIV-related deaths may have balanced out. Moreover, reported epidemiologic data on risk behaviors in this population suggest that indicators of high-risk behavior, such as inconsistent condom use among male partners, have not improved since 2001.21,22 In combination, these data demonstrate that new infections are likely to be persisting in the MSM population and fueling the epidemic in this vulnerable group. In contrast to MSM, FSWs appear to be experiencing encouraging declines in the HIV epidemic. Since 2001 both assay-derived incidence and HIV prevalence have declined substantially.23 Declines are corroborated by reports of a declining prevalence of STI and high-risk sexual behavior among the surveyed population during the same time period.22–24 These promising trends point to a potential prevention success in this population. Targeted programs on HIV and STI prevention among FSWs have been in place in the largest cities in Honduras and supported by governmental bodies in the country since the early 2000s. Participation in these programs has remained high among FSWs and has demonstrated substantial declines in HIV and STI rates in this population.25 Though HIV incidence rates were low among the Garifuna, levels of STI and inconsistent condom use have continued to be high in this group, suggesting that current prevention interventions also need to be enhanced for this group.4 These programs should include key prevention messages such as knowing the HIV status of you and your partners, HIV disclosure to partners, promotion of condom use, STI services, and linkages to care and treatment services. Interventions for the Garifuna should be culturally sensitive and provided in the local indigenous language for greatest impact. This analysis was subject to the following limitations. Mathematically modeled incidence was used as an independent measure of HIV incidence to assess the plausibility of the assay-derived estimates. However, because the inputs for the model were derived from limited surveillance and programmatic data available for FSWs, MSM, and Garifuna in 2006, errors for the modeled estimates are likely to be large. Due to low levels of HIV-1 antibody, persons on antiretroviral (ARV) therapy are prone to test as recent on the BED assay, which may result in an overestimation of HIV
KIM ET AL. incidence. In this survey, we were unable to measure ARV use among participants. However, the level of ARV use among persons with advanced HIV infection in Honduras was low in 2006 at approximately 20% of those in need of HIV treatment.26 Given this, it is assumed that the level of misclassification due to ARV use was negligible. A large number of HIV-positive specimens were missing and not tested by the incidence assay. Though we accounted for the impact of missing specimens in the analysis, the adjustment was based on an assumption that specimens were missing at random. No significant differences in demographics and sexual behavior were observed among HIV-positive specimens tested and not tested by the incidence assay (data not shown). The variance in the assay-derived incidence estimates is high due to the small sample sizes of surveyed groups and uncertainty in the FRR applied. In addition, though a degree of uncertainty exists in population estimates derived from RDS and two-stage cluster sampling surveys, the incidence estimates presented in this analysis did not account for the added uncertainty incurred due to the sampling methodology applied. Incidence estimates, therefore, are not generalizable beyond the samples tested. Given the statistical challenges in obtaining reliable estimates of assay-derived HIV incidence in surveys of populations at high risk for HIV infection, the application of incidence assays to newly diagnosed HIV cases in a robust HIV case-based surveillance system may serve as an alternative source of incidence surveillance data for countries with HIV epidemics similar to Honduras. HIV case-based surveillance data provide the opportunity to describe patterns of recently acquired HIV infection among newly diagnosed cases by geography, age, gender, and behavioral indicators, which can directly inform programs on where and among whom to target their HIV prevention efforts. If HIV testing patterns can be quantified accurately in the country, these systems can also be used estimate HIV incidence on a population level.27,28 Our findings highlight the importance of collecting data on recent HIV infections to monitor the epidemic among populations, evaluate interventions, and inform the national HIV response. These data confirm that MSM continue to be a highly vulnerable population in Honduras and also highlight successes in targeted prevention strategies among FSW. The results of this analysis led to increased donor funding for MSM programs and the expansion of sentinel surveillance for sexually transmitted infections to include MSM-friendly sites in the two largest cities in the country: San Pedro Sula and Tegucigalpa. Continued national and international support for organizations working with MSM populations is needed to improve the reach and effectiveness of current programs for men and to increase the awareness of MSM vulnerability both within the MSM community and the public health community. Established programs for FSWs and Garifuna should continue at the same funding and priority levels to maintain low levels of HIV infection in this group. In addition, investments made by local and external funds for the national HIV response should be evaluated periodically to ensure the investment is appropriate and prioritized for groups at highest risk for HIV infection. It is recommended that these results be used for advocacy, policy, and development of HIV strategic plans for most-at-risk populations in Honduras.
HIV INCIDENCE AMONG HIGH-RISK POPULATIONS IN HONDURAS Acknowledgments This work was funded by the U.S. President’s Emergency Plan for AIDS Relief. The findings and conclusions in this report are those of the authors and do not necessarily represent the official views of the U.S. Centers for Disease Control and Prevention (CDC). Author Disclosure Statement No competing financial interests exist. References 1. UNAIDS: 2005 Report on the Global AIDS Epidemic, July 2005. Edited by UNAIDS, Geneva, 2005. 2. Soto RJ, Ghee AE, Nunez CA, et al.: Sentinel surveillance of sexually transmitted infections/HIV and risk behaviors in vulnerable populations in 5 Central American countries. J Acquir Immune Defic Syndr 2007;46(1):101–111. 3. Escure G (Ed.): Garifuna in Belize and Honduras. In: Creoles, Contact, and Language Change: Linguistics and Social Implications. John Benjamins Publishing Company, Philadelphia, 2004. 4. Paz-Bailey G, Morales MS, and Jacobson J: High rates of STD and sexual risk behaviors among Garifunas in Honduras. J Acquir Immune Defic Syndr 2009;15(S1):S26–34. 5. Secretaria de Salud de Honduras: Estudio seroepidemiologico de Sı´filis, He´patitis B y VIH en poblacion Garı´funa de El Triunfo de la Cruz, Bajamar, Sambo Creek y Corozal. Tegucigalpa: Secretaria de Salud de Honduras, 1998. 6. Mastro T, Kim A, Hallett T, et al.: Estimating HIV incidence in populations using tests for recent infection: Issues, challenges and the way forward. J HIV/AIDS Surveill Epidemiol 2010;2(1):7. 7. Kim A, Hallett T, Stover J, et al.: Estimating HIV incidence among adults in Kenya and Uganda: A systematic comparison of multiple methods. PLoS One 2011;6(3):e17535. 8. Heckathorn D: Respondent driven sampling II: Deriving valid population estimates from chain-referral samples of hidden populations. Social Prob 2002;49(1):11–34. 9. Heckathorn D: Respondent driven sampling: A new approach to the study of hidden populations. Social Prob 1997;44(2):174–199. 10. Parekh BS, Kennedy MS, Dobbs T, et al.: Quantitative detection of increasing HIV type 1 antibodies after seroconversion: A simple assay for detecting recent HIV infection and estimating incidence. AIDS Res Hum Retroviruses 2002;18(4):295–307. 11. McWalter T and Welte A: Relating recent infection prevalence to incidence with a sub-population of non-progressors. J Math Biol 2010;60(5):687–710. 12. Parekh B, Hanson D, Hargrove J, et al.: Determination of mean recency period for estimation of incidence with the BED-capture EIA in persons infected with diverse HIV-1 subtypes AIDS Res Hum Retroviruses 2011;27(3):265–273. 13. UNAIDS: 2008 Report on the Global AIDS Epidemic. Edited by UNAIDS. UNAIDS, Geneva, 2008. 14. Hemelaar J, Gouws E, Ghys P, and Osmanov S: Global and regional distribution of HIV-1 genetic subtypes and recombinants in 2004. AIDS 2006;20:W13–23.
519
15. Murillo W, Paz-Bailey G, Morales S, et al.: Transmitted drug resistance and recency of infection in newly diagnosed HIV1 individuals in Honduras. J Clin Virol. 2010;49(4):239–44. 16. Soto RJ, Ghee AE, Nunez CA, et al.: Sentinel surveillance of sexually transmitted infections/HIV and risk behaviors in vulnerable populations in 5 Central American countries. J Acquir Immune Defic Syndr 2007;46(1):101–111. 17. World Health Organization (WHO): When and how to use assays for recent infection to estimate HIV incidence at a population level. Edited by Organization WHO. Geneva, 2011. 18. UNAIDS: Epidemiologic software and tools: Modes of transmission (MoT) spreadsheets (Excel files) and manuals. UNAIDS, Geneva, 2009. 19. Garnett GP, Gregson S, Stanecki KA, et al.: Criteria for detecting and understanding changes in the risk of HIV infection at a national level in generalized epidemics. Sexually Transmitt Infect 2006;82(s1):i48–51. 20. El Salvador Ministry of Health: Final Report for the Central American Survey of Sexual Behavior and HIV/STI Prevalence among Vulnerable Populations in El Salvador. Ministry of Health of El Salvador, San Salvador, 2009. 21. Padilla I and Soto R: Estudio multice´ntrico centroamericano de prevalencia de VIH/ITS y comportamientos en hombres que tienen sexo con otros hombres en Honduras. In AIDS Action Project for Central America-PASCA, Tegucigalpa, 2003. 22. Secretaria de Salud De Honduras: Encuuesta Centroamericana de Vigilancia de Comportanmiento Sexual y Prevalencia del VIH/ITS en Poblaciones Volnerables. Tegucigalpa, 2008. 23. Soto R, Ghee A, Nunez C, et al.: Sentinel surveillance of sexually transmitted infections/HIV and risk behaviors in vulnerable populations in 5 Central American countries. J Acquir Immune Defic Syndr 2007;46(1):101–111. 24. Paredes M and Soto R et al.: Estudio multice´ntrico centroamericano de prevalencia de VIH/ITS y comportamientos en mujeres trabajadoras comerciales del sexo en Honduras. Edited by America-PASCA AAPfC. Tegucigalpa, 2003. 25. Tinajeros F, Miller W, Castro L, et al.: Declining sexually transmitted infections among female sex workers: The results of a HIV and STI prevention strategy in Honduras, 2006–2008. International Journal of STD and AIDS. 2012;23(2):88–93. 26. United National Data Retrieval System. http://data.un.org/ Data.aspx?d = MDG&f = seriesRowID%3A765. 27. Hall HI, Song R, Rhodes P, et al.: Estimation of HIV incidence in the United States. JAMA 2008;300(5):520–529. 28. Le Vu S, Meyer L, Cazein F, et al.: Performance of an immunoassay at detecting recent infection among reported HIV diagnoses. AIDS 2009;24(23):1773–1779.
Address correspondence to: Andrea A. Kim Centers for Disease Control and Prevention National Center for Global Health Division of Global HIV/AIDS 1600 Clifton Road Atlanta, Georgia 30333 E-mail:
[email protected]
