Soto, Ramón J MD, MPH*; Ghee, Annette E PhD, MPH†‡; Nuñez, Cesar A MD, MPH§; Mayorga, Ruben MD∥; Tapia, Kenneth A MS†; Astete, Sabina G PhD†; Hughes, James P PhD†¶; Buffardi, Anne L MSW†; Holte, Sarah E PhD†¶#; Holmes, King K MD, MPH†‡**; and the Estudio Multicéntrico Study Team
The Latin American region is the third most affected by HIV after sub-Saharan Africa and Asia. An estimated 179,200 persons aged 16 years or older live with HIV infection in El Salvador, Guatemala, Honduras, Nicaragua, and Panama; 29% of these are women.1 HIV infection in Central America is predominantly sexually transmitted and concentrated in a few populations.1 Information regarding men who have sex with men (MSM), the group with the greatest burden of disease elsewhere in the region,2 is particularly scant.
The Estudio Multicéntrico (EMC) in female sex workers (FSWs) and MSM constituted “second-generation surveillance” of HIV/sexually transmitted infection (STI) prevalence3 and linked risk behaviors enhanced to include HIV seroincidence estimation. Second-generation surveillance has been advocated to replace HIV surveillance implemented separately from behavioral surveillance.4-6 In low-prevalence settings, this surveillance approach for the highest risk populations is considered cost-effective7 despite its challenges.8-10 Recently proposed enhancements would include HIV seroincidence estimation and ongoing monitoring of HIV resistance patterns.11 The EMC was framed as a technology-transfer operations research activity and policy enhancement tool.
We report seroprevalence and seroincidence of HIV infection, prevalence of other STIs, and sociobehavioral characteristics for MSM and FSWs in El Salvador, Guatemala, Honduras, Nicaragua, and Panama. Associations of sociobehavioral and workplace-level risk factors (eg, access to condoms in the workplace) with STI/HIV prevalences are described separately (unpublished data).
Study Sites and Eligibility
This cross-sectional study involved 2466 FSWs and 1418 MSM recruited between May 2001, and April 2002. Implementation in each country started on independent time lines and lasted between 2 and 6 months, with most participants in each study population (>95%) enrolled within a 4-month period. Self-identified FSWs and MSM at least 18 years of age were recruited in the 6 largest cities of the region (San Salvador, Guatemala City, Tegucigalpa, Managua, Panama City, and San Pedro Sula). FSWs also were recruited in 8 port communities (Acajutla, El Salvador; Puerto Barrios and San José, Guatemala; Puerto Cortés and the San Lorenzo area, Honduras; Bluefields and Corinto, Nicaragua; and Colón, Panama). National epidemiologists selected communities based on HIV infection levels or AIDS cases and STI cases reported through surveillance systems. In major cities, between 1 and 4 government STI clinics served as screening sites for FSW participants; for MSM participants, 1 or 2 clinics at community-based organizations served as screening sites. In port communities, 1 government STI clinic served as the screening site for FSWs. A total of 23 screening sites were used. All screening sites were selected for ease of access and for the clinic team's experience in working with MSM and FSW populations.
Institutional review boards (IRBs) at the University of Washington and in each country approved the study. All participants gave informed consent; written consent was obtained in Honduras, the first country in which the EMC was implemented, and oral consent was obtained in all remaining countries. In Panama City, MSM participants were initially told they could participate in the behavioral interview or the clinical examination with or without specimen collection; midway through the study, we encouraged participation in all procedures. Specimens and data collection documents had only study identification numbers; participants obtained test results in sealed envelopes labeled solely with identification numbers. Participants were invited to return to their screening site at 2 weeks for results of confirmatory testing for HIV and again 2 months after screening for other pending STI test results. The Ministry of Health in each country provided free follow-up care at all collaborating screening sites. In accordance with national guidelines, individuals were treated syndromically for curable STIs. At the time of screening, FSWs were treated for syphilis and bacterial vaginosis (BV) based on rapid test results (see section on laboratory methods) and for Neisseria gonorrhoeae or Chlamydia trachomatis based on signs of cervical inflammation. MSM were treated at the time of screening for syphilis or for N. gonorrhoeae or C. trachomatis based on the results of on-site rapid plasma reagin (RPR) tests and the presence of urethral discharge, respectively. Additional treatment for syphilis or referral to HIV care was given at 2-week follow-up, and treatment for N. gonorrhoeae or C. trachomatis, based on results of polymerase chain reaction (PCR) testing, was offered at 2-month follow-up. Treatment for HIV infection was limited but expanding when this study was done.
Clinicians, laboratory specialists, and counselors of each country's Ministry of Health AIDS/STI program staffed the study. For the MSM survey, MSM counselors from collaborating community-based organizations (CBOs) served as counselor/interviewers; for the FSW survey, counselor/interviewers were from government health facilities or collaborating CBOs. We trained study teams at national workshops covering the protocol (including clinical, laboratory, and interviewing techniques), ethical considerations, and study coordination. Teams reported to National AIDS/STI Program directors; an international study coordinator (RJS) provided technical support. Study teams provided HIV pre- and posttest counseling, free treatment for any curable STI detected, and educational materials and condoms, referring HIV-positive participants to available HIV treatment centers.
Target sample sizes were 300 FSWs and MSM per city and 150 FSWs per port community. Large city recruitment targets were based on estimated HIV seroprevalences of 5% in MSM and 2% in FSWs so as to achieve prevalence estimates precise to ±2.5% and ±1.9%, respectively.
For FSWs in the 6 cities, we used 2-stage cluster sampling; all mapped commercial sex workplaces constituted independent sampling frames for women working in “fixed” establishments (eg, brothels, bars, nightclubs) and in informal “ambulatory” settings (eg, streets, parks, hostels, hotels) and then considered strata for the analysis. Experienced CBO outreach workers mapped workplaces from prior knowledge and by using key informants. Workplaces were visited at least twice to enumerate potentially eligible FSWs. Settings posing security risks were excluded. Within each sampling frame, workplaces were selected up to 3 times with probability proportional to estimated numbers of eligible women. All FSWs in workplaces with ≤10 FSWs were invited to participate. For larger workplaces, 7 FSWs were randomly selected and invited to participate each time a workplace was selected.
Mapping showed smaller populations of FSWs in port communities than in cities. We invited all self-identified FSWs in port communities to participate.
For MSM, community mapping identified sites frequented by MSM. Study team members and outreach workers from partnering CBOs repeatedly visited these sites to describe study procedures to opinion leaders and potential participants. Convenience sampling emphasized recruitment of initial participants with diverse sexual identities and encouraged referral of friends, a common sampling technique in qualitative research.12,13
Clinicians obtained sexual histories (and reproductive histories from women) and performed physical examinations, including visual inspection of the external genitalia. Clinicians collected urine from all MSM and urethral swab specimens from those with signs of urethral exudate. For FSWs, clinicians performed speculum and bimanual examinations and collected cervical and vaginal specimens; those without a uterus and cervix were not eligible for participation. In all countries except Nicaragua, clinicians from collaborating CBOs, themselves MSM, examined male participants.
Sexually Transmitted Infection Testing
After giving informed consent, undergoing interviews, and receiving HIV/STI pretest counseling, participants gave 10 mL of venous blood. RPR testing (Arlington Scientific, Springville, UT) was performed on-site, and specimens were transported on cold packs to national reference laboratories and stored at −20°C before confirmatory Treponema pallidum particle agglutination assay (TP-PA) testing (Serodia; Fujirebio, Malvern, PA). Aliquots were prepared for herpes simplex virus (HSV) type-specific antibody testing (HerpeSelect; Focus Diagnostics, Cypress, CA) and HIV antibody testing (Abbott Laboratories, Abbott Park, IL). Dual enzyme-linked immunosorbent assay (ELISA) testing for HIV was performed at national reference laboratories; dual positive sera were shipped on dry ice to the US Centers for Disease Control and Prevention for confirmatory Western blot (WB) and BED-capture enzyme immunoassay (referred to as LS-EIA)14 to identify likely recent infections. In Honduras, Guatemala, and Panama, ELISAs for serum HSV-2 antibody were performed at national reference laboratories; sera from El Salvador and Nicaragua were shipped on cold packs to the Gorgas Institute Reference Laboratory in Panama City for serologic testing. Index values >1.1 at 450 nm were classified as HSV-2-seropositive. A random subsample of all sera underwent HSV quality control WB testing15,16 at the University of Washington Virology Laboratory; overall concordance was 94%, and WB assays were positive for HSV-2 antibody for 50/50 sera with Focus Diagnostic ELISA index values ≥ 3.5, 3/6 with index values from 1.1 to 3.4, and 1/17 with index values <1.1.
Specimens were evaluated for N. gonorrhoeae and C. trachomatis using the COBAS PCR automated system (reagents provided by Roche Molecular Diagnostics). For MSM, urethral exudate swab specimens and leukocyte esterase (LE) strip-positive or trace-positive urine samples plus all cervical samples from FSWs were tested by PCR. Swab specimens were stored in 2SP medium at −20°C. In Honduras, Guatemala, and Panama, specimens were tested at the reference laboratory, whereas specimens from El Salvador and Nicaragua were transported on cold packs for testing at the Gorgas Institute in Panama. At the University of Washington, we performed quality control PCR testing for all positive N. gonorrhoeae specimens and random 20% and 5% subsamples of C. trachomatis-negative and N. gonorrhoeae-negative specimens, respectively.
Vaginal fluid specimens were Gram stained and interpreted by a technician at the reference laboratory in Honduras who was trained in the Nugent system17 for diagnosis of BV. For quality control, 100 slides were sent to the University of Washington Research Microbiology Laboratory for verification (weighted κ = 0.84). FSWs with vaginal fluid positive for amines and pH ≥4.7 by FemExam test card (Litmus Concepts, Santa Clara, CA) were treated for BV according to national guidelines.
Trichomonas vaginalis was detected using the InPouch TV culture system (BioMed Diagnostics, White City, OR) incubated at 37°C and examined daily for 1 week at local laboratories. Quality control was not performed because of fragility of the parasite during transportation to national laboratories.
Demographic and Behavioral Characteristics
Counselor-interviewers used questionnaires based on those used previously in the Latin American region18-20 and internationally;3 questionnaires measured sociodemographic characteristics, sexual behaviors by different sex partner types, STI/HIV-related health care-seeking behaviors, knowledge and attitudes regarding STI/HIV, drug or alcohol use, attitudes toward HIV-positive persons, and exposure to HIV/STI prevention programs. Questionnaires were field tested in 30 pilot interviews in Honduras; subsequently, we revised wording minimally to ensure comprehension in other countries. Interviews took 25 to 40 minutes and were conducted before pretest counseling.
We report a sociodemographic profile of participants, STI/HIV prevalences, and programmatic and behavioral indicators using descriptive statistics, including proportions, means, and standard deviations.
We computed programmatic and behavioral indicators to approximate published definitions.3 The indicator for knowledge of HIV prevention methods was the proportion indicating that HIV could be prevented by using condoms for each sex act (anal sex for MSM) among all participants with valid responses. We defined the percent reporting incorrect beliefs or myths regarding HIV transmission separately for MSM and FSWs as follows: for the 2 most common incorrect beliefs, we computed the proportion that did not hold these beliefs and agreed that a healthy-looking person can be HIV-positive. For MSM and FSWs, respectively, the most common incorrect beliefs were that HIV can be transmitted by insect bites or sharing eating utensils and by insect bites or sharing a toilet with an infected person. Exposure to behavior-change interventions was the proportion reporting behavior change in response to radio, counseling, printed messages, counseling, or educational lecture sessions (charlas) among those with valid responses. Consistent condom use was defined as always versus less than always for the time frame defined for each partner type.
Among MSM, sexual bridgers were those reporting having had sex with at least 1 man and 1 woman in the prior 6 months. Unprotected “bridging” sex was defined as inconsistent condom use with men and women in the prior 30 days among bridgers who provided valid condom use information for both partner types.
Selected descriptive statistics were compared using χ2 tests. Exact binomial 95% confidence intervals (CIs) around estimates of STI/HIV prevalence were computed. HIV seroincidence estimates and CIs were computed according to previously described procedures (see Appendix 1 of the study by Janssen et al21) using exact methods.
For FSWs, we computed sampling probability based on sampling frames, 2 for each large city, corresponding to women in “fixed” versus “ambulatory” commercial sex workplaces, and on the 2-stage sampling procedure described previously. We derived weights reflecting the probability of an individual falling into the sample and applied them to all prevalence estimates for STI/HIV in FSWs (see Appendix).
Study Population Characteristics
Although records of individuals refusing participation were not systematically kept, national coordinators indicated that fewer than 5% of FSWs and 20% of MSM refused, consistent with direct observation of enrollment activities by the international study coordinator (RJS).
Men Who Have Sex With Men
Participants averaged 25.6 years of age; 77.5% had a secondary or university education (83.9% of Panamanian MSM reported a university education; Table 1). Most were single, although 4.6% were cohabitating with a woman. A few (11%) reported performing sex work. Most (53.4%) self-identified as gay or homosexual, 28.9% as bisexual, 10.6% as heterosexual, and 7.2% as transvestite.
Female Sex Workers
Participants averaged 27.2 years of age (see Table 1). Most had a primary or no education, half were married or cohabitating with 1 or more men, and 36.2% were divorced or widowed. Most worked in brothels, appointment houses, bars, or nightclubs. Combined payment from the last new plus last repeat client averaged US $26, ranging from $17 in El Salvador to $50 in Panama. In El Salvador, Nicaragua, and Honduras, <2.0% were foreign-born, whereas 59.0% in Guatemala and 68.1% in Panama were foreign-born.
Prevalences of Sexually Transmitted Infection/HIV and HIV Seroincidence Estimates
Comparing Men Who Have Sex With Men and Female Sex Workers
The regional HIV seroprevalence of 11.7% for MSM was substantially higher than for FSWs (3.7%) (Fig. 1), with the highest seroprevalence seen among MSM in El Salvador (15.3%) and the lowest among FSWs in Nicaragua and Panama (both 0.2%). Similarly, the estimated HIV seroincidence per 100 person-years was higher for MSM (5.1) than for FSW (1.5) (data not shown). Conversely, HSV-2 seroprevalence was considerably higher for FSWs (85.3%) than for MSM (48.2%) (Fig. 2A). After adjustment for income, venue type, lubricant use, and country, HSV-2 infection was significantly associated with HIV seropositivity in FSWs (odds ratio [OR] = 11.0, 95% CI: 1.4 to 83.7, data not shown). Syphilis seropositivity prevalence was similar for MSM (8.3%) and FSWs (9.6%) (see Fig. 2B). For MSM, HSV-2 and syphilis seropositivity were associated with HIV seropositivity (OR = 5.0, 95% CI: 2.6 to 9.6 and OR = 3.7, 95% CI: 2.2 to 6.2, respectively; each model adjusted for city, age, education, sex work, and numbers of male partners; data not shown).
Men Who Have Sex With Men
The seroincidence of HIV infection per 100 person-years by LS-EIA assay for 975 MSM was 5.1 (95% CI: 2.9 to 7.9) and ranged from 2.1 in Guatemala to 2.7 in Panama, 3.6 in El Salvador, 4.9 in Honduras, and 14.4 in Nicaragua (95% CI: 5.4 to 29.7) based on only 7 ELISA-positive LS-EIA-seronegative individuals (data not shown). HSV-2 seroprevalence varied little by countries (see Fig. 2A), whereas syphilis seropositivity (RPR-positive confirmed by TP-PA) and potential early syphilis (titer ≥1:16) were higher in El Salvador, Guatemala, and Nicaragua than in Honduras or Panama (see Fig. 2B). In MSM tested after symptoms or signs of urethral infection or a positive urine LE test result, 2.4% had gonococcal infection and 7.2% had chlamydial infection, with the highest levels seen in Honduras and the lowest in El Salvador (data not shown and not available for Nicaragua).
Female Sex Workers
HIV seroprevalences were highest among FSWs in Honduras, Guatemala, and El Salvador (9.6%, 4.3%, and 3.2%, respectively; see Fig. 1). Seroincidence estimates ranged from 0 in Panama or Nicaragua to 1.8 in El Salvador, 3.0 in Guatemala, and 3.2 in Honduras (data not shown). FSWs in El Salvador, Guatemala, and Honduras also had the highest HSV-2 seroprevalences (see Fig. 2A). As with MSM, syphilis seropositivity in FSWs was highest in El Salvador, Guatemala, and Nicaragua (15.0%, 11.8%, and 8.4%, respectively; see Fig. 2B).
Overall prevalences of endocervical infections were 20.1% for C. trachomatis and 8.1% for N. gonorrhoeae (see Fig. 2C). C. trachomatis prevalence was lowest in Honduras, and N. gonorrhoeae prevalence was lowest in Nicaragua and Panama.
Prevalence of T. vaginalis infection was 11.0% overall and highest in Honduras (13.1%), followed by El Salvador and Nicaragua (both 11.7%), Guatemala (10.3%), and Panama (5.9%); the prevalence of BV among 1591 FSWs tested was 54.8% and was highest in El Salvador (63.9%), followed by Guatemala (59.8%), Honduras (59.2%), Nicaragua (48.7%), and Panama (39.0%; data not shown).
Programmatic and Behavioral Risk Indicators
The proportion with correct knowledge of HIV prevention methods was higher than 95% in FSWs and nearly 90% in MSM, although lowest for MSM in El Salvador and Nicaragua (Table 2). More than half of all FSWs held incorrect beliefs regarding HIV transmission, as did MSM in Nicaragua.
Only 50.1% of FSWs and 42.1% of MSM had previously undergone voluntary HIV testing and received results; for FSWs, the proportion was highest in Honduras and lowest in Panama, and for MSM, it was highest in Panama and Honduras and lowest in El Salvador and Guatemala. The proportion reporting exposure to behavior change interventions was 44% for FSW and was lowest in Guatemala (33.5%); it was even lower for MSM (28.8%), especially for those in El Salvador (15.9%).
MSM reported consistent condom use in the past month more often with casual (61.2%) than with stable (46.7%) male partners; 77% of MSM sex workers reported consistent condom use with clients in the past month. MSM in Nicaragua reported the lowest levels of consistent condom use for all partner types. FSWs most often reported consistent condom use with new clients (77.2%), less often with regular or repeat clients (71.9%), and infrequently with stable partners (7.9%). FSWs in Nicaragua and El Salvador reported the least frequent condom use with clients.
Bridging behavior was reported by 22.2% of MSM, ranging from 8.9% in Panama to 34.0% in El Salvador. Among bridgers who provided condom use information, 36.6% reported having had unprotected sex with men and women during the past month. Injecting drug use was rare for MSM (0.7%) and FSWs (2.1%).
The EMC study demonstrated the feasibility of standardized sentinel surveillance, enhanced to assess several STIs in addition to HIV infection, with sociobehavioral instruments adapted to vulnerable populations in resource-limited settings.
Concurrent estimates of HIV prevalence in the general population were highest in Honduras (1.8%), followed by Guatemala (1.1%), Panama (0.9%), El Salvador (0.7%), and Nicaragua (0.2%).22 Consistently, we found substantially higher HIV prevalences in MSM, ranging from 6.9 times higher in Honduras with the longest standing epidemic to approximately 10-fold higher in Guatemala and Panama and to 21.8 and 38 times higher in El Salvador and Nicaragua, respectively. Although HIV prevalences in FSWs were low in Panama and Nicaragua (0.2% in both countries), in Honduras, Guatemala, and El Salvador, they were 4- to 5-fold higher than corresponding general population estimates.
Sentinel surveys can galvanize commitments to HIV/STI prevention for stigmatized vulnerable populations and help to overcome any reluctance to acknowledge and support those needs.23 Despite limitations in laboratory capacity and minor variations in protocol implementation, we transferred technology to teams of laboratory, clinical, and counseling/interviewer staff; reinforced collaborative relations between government programs and CBOs; and made preliminary findings available to policymakers, resulting in increased commitments to Central American intervention programs for at-risk populations (culminating session with Ministers of Health at the Third Central American HIV/AIDS Congress [CONCASIDA], Panama City, October 2003).24,25
STI/HIV prevalences, HIV seroincidence, behavior patterns, and programmatic indicators all highlight the sexual health needs of Central American MSM. Shortfalls in these services are seen in other Latin American settings.26-29 We found HIV seroprevalence to be highest in El Salvador, Guatemala, and Honduras, and seroincidence estimates were highest in Nicaragua (based on small numbers of cases) and Honduras. We detected the highest syphilis seroreactivity levels, including TP-PA titers ≥1:16, in El Salvador, Guatemala, and Nicaragua. HSV-2 and syphilis seropositivity were associated with HIV infection. MSM bridging, including unprotected sex, is consistent with ethnographic and sociologic research in Latin America.30-32 These findings, combined with evidence of modest levels of previous voluntary HIV testing, exposure to behavioral interventions, and condom use despite relatively high levels of education and HIV infection compared with the FSW population, compel the expansion of outreach and integrated clinical and behavior-change interventions acceptable and accessible to MSM.
In FSWs, we report higher HIV seroprevalence and seroincidence than in Mexican FSWs33 or Central American immigrant FSWs at the Mexican-Guatemalan border.34 We also found higher HIV seroprevalence and lower syphilis seropositivity than found a decade earlier among FSWs in Tegucigalpa.35 We found higher HSV-2 seroprevalence in FSWs than seen in Asian or African FSWs36 or in Mexican and other Latin American at-risk populations,37,38 with the exception of a single Mexican study,34 and we documented an association between HSV-2 and HIV infection. HSV-2 infection may increase HIV transmission and acquisition.39C. trachomatis infection seems to be poorly controlled by existing clinical services, and N. gonorrhoeae prevalence seemed to be high in 3 countries, based on the PCR assay used. Consistent condom use levels were higher for FSWs than for MSM but suggest gaps in condom promotion for both populations, particularly in Nicaragua, Guatemala, and El Salvador. High levels of knowledge regarding HIV prevention methods are tempered by incorrect beliefs regarding transmission, indicating an ongoing need for educational interventions.
Several country-level characteristics have been associated with national HIV seroprevalences40,41 and vary among Central American countries (Table 3). In our study, small numbers (5 countries) precluded formal ecologic analyses, yet several differences emerged. Panama has the smallest proportion of young people and the greatest economic development, with the highest gross domestic product (GDP), the lowest GINI index of economic inequity, and the highest literacy rate, all of which are associated with a lower prevalence of HIV.40,41 Although Nicaragua, with the lowest GDP, has the lowest HIV seroprevalence in MSM and FSWs, estimated HIV seroincidence in MSM was highest. Panama and Guatemala have the highest ratio of kilometers of paved highways per land area, possibly affecting the relative role of migration and commerce in STI transmission. We did not systematically compare STI services across countries, but striking differences in the prevalence of curable STIs, lowest in Panama and generally highest in El Salvador, Guatemala, and Honduras, coupled with low levels of reported exposure to behavioral interventions (especially among MSM in El Salvador and Guatemala) and condom use (especially in Nicaragua), all highlight needs for improved treatment and prevention services.
For FSWs, national differences likely cannot be attributable to sampling, because we used standardized probabilistic sampling methods. Results for Panama may be affected by participation bias to an unknown degree, because foreign FSWs are given 6-month work permits. For MSM, sampling differences among study sites could bias intracountry differences in HIV/STI prevalence and behavioral patterns. Our CBO partners recommended against using time-location or other sampling methods, because MSM are heavily stigmatized. Difficulties in reaching target sample sizes for MSM probably reflect this reality and do not suggest small MSM communities in the 6 survey cities. Comparing the utility and effectiveness of time-location42,43 versus respondent-driven44,45 sampling for both populations seems warranted.
In perhaps the most comprehensive regional sentinel surveillance for HIV infection and other STIs yet undertaken, the study offers a unique regional epidemiologic profile providing prevalences of STI, including HIV infection, estimates of HIV seroincidence, and profiles of selected risk behaviors and programmatic indicators in Central American MSM and FSWs. Repeat surveys monitor trends in HIV and STI prevalence over time in vulnerable and lower risk populations and can monitor antiretroviral resistance patterns as treatment programs expand. To inform the design of targeted interventions for FSWs and MSM, associations of sociobehavioral and demographic factors with prevalences of STI and HIV infection have been analyzed separately (unpublished data). Despite varying patterns between countries requiring customized planning priorities, concentrated HIV epidemics exist among MSM in all 5 countries; high levels of HSV-2 infection and an association with HIV seropositivity were documented among FSWs; HSV-2 and syphilis seropositivity were associated with HIV infection in MSM; and the greatest gap in addressing basic sexual health needs of MSM and FSWs apparently exists in El Salvador, Guatemala, and Nicaragua.
The Estudio Multicéntrico was implemented by the National AIDS and STI Control Programs of El Salvador, Guatemala, Honduras, Nicaragua, and Panama; the Central American AIDS Action Project (Proyecto de Acción en SIDA Centroamericano), a project of the Futures Group International; and the University of Washington.
The authors thank the men and women who participated in this study as well as the Ministry of Health clinical, counseling, interview, and laboratory teams in each of the participating study sites and the collaborating CBO partners that made invaluable contributions to recruitment and interviewing activities.
The authors thank Drs. Tim Granade and Bharat Parekh at the US Centers for Disease Control and Prevention for conducting confirmatory and LS-EIA testing of HIV specimens for our study. They recognize Kara Richmond and Mary Catlin (both of the University of Washington's Center for AIDS and STD) for their valuable assistance with study implementation. They also greatly appreciate comments on prior versions of this manuscript from Drs. Michelle A. Williams and Jeanne M. Marrazzo of the Departments of Epidemiology and Medicine, respectively, at the University of Washington.
The national EMC study teams are as follows: El Salvador: Gladys de Bonilla, Luis Palma, Telma de Pineda, Sonia Perez, Marmin Ruth de Sosa, and Zandra Elizabeth de Fuentes; Guatemala: Dory Lucas Alecio, Sergio Aguilar, V. A. Moscoso, Francisco Ardón, Malvina de León, Eyda de Campollo, Rosana de Mazariegos, and Dina Cruz; Honduras: Marco A. Alvarenga, Iris Padilla, Iván Espinoza, Rosalinda Hernández, Victor Solera, Berta Alvarez, Marco Urquía, Suyapa Mendoza, Mirla Rosa, Roque López, and Rita Meza; Nicaragua: Matilde Román, Federico Avilés, Juan José Amador, Alcides González, Lesbia Altamirano, Jean Dominique, René Berríos, Ángel Balmaceda, Yara Saborio, and Justo Reyes; Panama: Gladys Alicia Guerrero, Biverly Grajales, Norma García de Paredes, Itza Barahona de Mosca, Federico Hernández, Zoila Castillo, Elvira de Austin, Dalys de Mojica, Edda Márquez, Cirilo Lyons, Rubén Campo, Raquel Bolaños, Carlos Justo, Markela Quinzada, Rolando Blanco, and Rosmery Birham.
Collaborating community-based partners are as follows: El Salvador: Asociación de Mujeres “Flor de Piedra,” and Asociación “Entre Amigos”; Guatemala: Asociación de Salud Integral, and Asociación de Apoyo a una Sexualidad Integral frente al SIDA; Honduras: Programa para el Desarrollo de la Infancia y la Mujer, Asociación para el Desarrollo Comunitario, Fraternidad Sampedrana de Lucha Contra el SIDA, Programa de Comunicación y Vida de la Municipalidad Sampedrana, Grupo PRISMA, Asociación Violeta, Comunidad Gay Sampedrana, and Fundación Solidaridad y Confraternidad contra el SIDA; Nicaragua: Centro para la Educación y Prevención del SIDA, Trabajadores para la Educación, Salud e Integración Social, Fundación Xochiquetzal, and Cruz Roja Nicaragüense; Panama: Equipo Multidisciplinario en Salud Sexual y Reproductiva and Asociación Hombres y Mujeres Nuevos de Panamá.
2. Montano SM, Sanchez JL, Laguna-Torres A, et al. Prevalences, genotypes, and risk factors for HIV transmission in South America. J Acquir Immune Defic Syndr. 2005;40:57-64.
3. Behavioral Surveillance Surveys: Guidelines for Repeated Behavioral Surveys in Populations at Risk of HIV. Arlington, VA: Family Health International; 2000.
4. Garnett GP, Garcia-Calleja JM, Rehle T, et al. Behavioural data as an adjunct to HIV surveillance data. Sex Transm Infect. 2006;82(Suppl 1):i57-i62.
5. Zaba B, Slaymaker E, Urassa M, et al. The role of behavioral data in HIV surveillance. AIDS. 2005;19(Suppl 2):S39-S52.
6. Rehle T, Lazzari S, Dallabetta G, et al. Second-generation HIV surveillance: better data for decision-making. Bull World Health Organ. 2004;82:121-127.
7. Ainsworth M, Over M. Confronting AIDS: Public Priorities in a Global Epidemic. A World Bank policy research report. Washington, DC: World Bank; 1997.
8. Schwartlander B, Ghys PD, Pisani E, et al. HIV surveillance in hard-to-reach populations. AIDS. 2001;15(Suppl 3):S1-S3.
9. McFarland W, Caceres CF. HIV surveillance among men who have sex with men. AIDS. 2001;15(Suppl 3):S23-S32.
10. Ghys PD, Jenkins C, Pisani E. HIV surveillance among female sex workers. AIDS. 2001;15(Suppl 3):S33-S40.
11. Lama JR, Sanchez J, Suarez L, et al. Linking HIV and antiretroviral drug resistance surveillance in Peru: a model for a third-generation HIV sentinel surveillance. J Acquir Immune Defic Syndr. 2006;42:501-505.
12. Patton MQ. Qualitative Evaluation and Research Methods. 2nd ed. Newbury Park, CA: Sage; 1990.
13. Coyne IT. Sampling in qualitative research. Purposeful and theoretical sampling; merging or clear boundaries? J Adv Nurs. 1997;26:623-630.
14. 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;18:295-307.
15. Ashley-Morrow R, Nollkamper J, Robinson NJ, et al. Performance of focus ELISA tests for herpes simplex virus type 1 (HSV-1) and HSV-2 antibodies among women in ten diverse geographical locations. Clin Microbiol Infect. 2004;10:530-536.
16. Ashley RL, Militoni J, Lee F, et al. Comparison of Western blot (immunoblot) and glycoprotein G-specific immunodot enzyme assay for detecting antibodies to herpes simplex virus types 1 and 2 in human sera. J Clin Microbiol. 1988;26:662-667.
17. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol. 1991;29:297-301.
18. Sanchez J, Campos PE, Courtois B, et al. Prevention of sexually transmitted diseases (STDs) in female sex workers: prospective evaluation of condom promotion and strengthened STD services. Sex Transm Dis. 2003;30:273-279.
19. Sanchez J, Gotuzzo E, Escamilla J, et al. Sexually transmitted infections in female sex workers: reduced by condom use but not by a limited periodic examination program. Sex Transm Dis. 1998;25:82-89.
20. Brewer TH, Hasban J, Ryan CA, et al. Migration, ethnicity and environment: HIV risk factors for women on the sugar cane plantations of the Dominican Republic. AIDS. 1998;12:1879-1887.
21. Janssen RS, Satten GA, Stramer SL, et al. New testing strategy to detect early HIV-1 infection for use in incidence estimates and for clinical and prevention purposes. JAMA. 1998;280:42-48.
24. World Bank. Countries of Central America need to invest more in HIV/AIDS prevention. 2003. Rev Panam Salud Publica. 2003;14:436-439.
25. Abreu AG, Noguer I, Cowgill K. HIV/AIDS in Latin American Countries. Washington, DC: The World Bank; 2003.
26. Bautista CT, Sanchez JL, Montana SM, et al. Seroprevalence of and risk factors for HIV-1 infection among South American men who have sex with men. Sex Transm Infect. 2004;80:498-504.
27. Tabet S, Sanchez J, Lama J, et al. HIV, syphilis and heterosexual bridging among Peruvian men who have sex with men. AIDS. 2002;16:1271-1277.
28. Tabet SR, de Moya EA, Holmes KK, et al. Sexual behaviors and risk factors for HIV infection among men who have sex with men in the Dominican Republic. AIDS. 1996;10:201-206.
29. Hernandez M, Uribe P, Gortmaker S, et al. Sexual behavior and status for human immunodeficiency virus type 1 among homosexual and bisexual males in Mexico City. Am J Epidemiol. 1992;135:883-894.
30. Parker RG. Behaviour in Latin American men: implications for HIV/AIDS interventions. Int J STD AIDS. 1996;7(Suppl 2):62-65.
31. Carrier JM. Cultural factors affecting urban Mexican male homosexual behavior. Arch Sex Behav. 1976;5:103-124.
32. Parker R, Caceres C. Alternative sexualities and changing sexual cultures among Latin American men. Cult Health Sex. 1999;1:201-206.
33. Conde-Glez CJ, Juarez-Figueroa L, Uribe-Salas F, et al. Analysis of herpes simplex virus 1 and 2 infection in women with high risk sexual behaviour in Mexico. Int J Epidemiol. 1999;28:571-576.
34. Uribe-Salas F, Conde-Glez CJ, Juarez-Figueroa L, et al. Sociodemographic dynamics and sexually transmitted infections in female sex workers at the Mexican-Guatemalan border. Sex Transm Dis. 2003;30:266-271.
35. Venegas VS, Madrid JP, Lorenzana I, et al. Human immunodeficiency virus infection and syphilis in Honduran female prostitutes. Int J STD AIDS. 1991;2:110-113.
36. Mbopi-Keou FX, Robinson NJ, Mayaud P, et al. Herpes simplex virus type 2 and heterosexual spread of human immunodeficiency virus infection in developing countries: hypotheses and research priorities. Clin Microbiol Infect. 2003;9:161-171.
37. Uribe-Salas F, Hernandez-Avila M, Conde-Gonzalez CJ, et al. Low prevalences of HIV infection and sexually transmitted disease among female commercial sex workers in Mexico City. Am J Public Health. 1997;87:1012-1015.
38. Weiss H. Epidemiology of herpes simplex virus type 2 infection in the developing world. Herpes. 2004;11(Suppl 1):24A-35A.
39. Corey L, Wald A, Celum CL, et al. The effects of herpes simplex virus-2 on HIV-1 acquisition and transmission: a review of two overlapping epidemics. J Acquir Immune Defic Syndr. 2004;35:435-445.
40. Drain PK, Smith JS, Hughes JP, et al. Correlates of national HIV seroprevalence: an ecologic analysis of 122 developing countries. J Acquir Immune Defic Syndr. 2004;35:407-420.
41. Over M. The effects of societal variables on urban rates of HIV infection in developing countries: an exploratory analysis. In: Ainsworth M, Fransen L, Over M, eds. Confronting AIDS: Evidence from the Developing World. Brussels, Belgium: The European Commission; 1998:39-51.
42. Xia Q, Tholandi M, Osmond DH, et al. The effect of venue sampling on estimates of HIV prevalence and sexual risk behaviors in men who have sex with men. Sex Transm Dis. 2006;33:545-550.
43. Stueve A, O'Donnell LN, Duran R, et al. Time-space sampling in minority communities: results with young Latino men who have sex with men. Am J Public Health. 2001;91:922-926.
44. Magnani R, Sabin K, Saidel T, et al. Review of sampling hard-to-reach and hidden populations for HIV surveillance. AIDS. 2005;19(Suppl 2):S67-S72.
45. Ramirez-Valles J, Heckathorn DD, Vazquez R, et al. From networks to populations: the development and application of respondent-driven sampling among IDUs and Latino gay men. AIDS Behav. 2005;9:387-402.
The probability Pi that individual j in cluster i fell into the sample, and the sampling weight for each observation, is proportional to
Equation (Uncited)Image Tools
where m was the number of clusters chosen within each stratum of the sample based on the sampling frame structure and M was the total number of individuals available for each stratum. Mi was the estimated number of women potentially available to participate from the ith cluster assessed during mapping, ni was the actual number of participants from the ith cluster, and Ni was the observed number of women available for each cluster.1 An inflation factor “I” was used to adjust for the possibility that the same cluster could be selected multiple times. For FSWs recruited in port communities, sampling weights were set to 1. Cited Here...
© 2007 Lippincott Williams & Wilkins, Inc.