ABSTRACT Objective Early, continued engagement with the HIV treatment continuum can help achieve viral suppression, though few studies have explored how risk factors for delays differ across the continuum. The objective of this study was to identify predictors of delayed diagnosis, delayed linkage to care, and nonadherence to treatment in the city of Salvador, Bahia, Brazil. Methods Data were collected during 2010 in a cross-sectional study with a sample (n = 1 970) of HIV-infected individuals enrolled in care. Multiple logistic regression analyses identified sociodemographic variables, behaviors, and measures of health service quality that were associated with delayed diagnosis, delayed linkage to care, and treatment nonadherence. Results For delayed diagnosis, male gender (adjusted odds ratio (AOR), 3.02; 95% confidence interval (CI), 2.0–4.6); age 45 years and older (AOR, 1.67; 95% CI, 1.1–2.5); and provider-initiated testing (AOR, 3.00; 95% CI, 2.1–4.4) increased odds, while drug use (AOR, 0.29; 95% CI, 0.2–0.5) and receiving results in a private space (AOR, 0.37; 95% CI, 0.2–0.8) decreased odds. For delayed linkage to care, unemployment (AOR, 1.42; 95% CI, 1.07–1.9) and difficulty understanding or speaking with a health care worker (AOR, 1.61; 95% CI, 1.2–2.1) increased odds, while posttest counseling (AOR, 0.49; 95% CI, 0.3–0.7) decreased odds. For nonadherence, experiencing verbal or physical discrimination related to HIV (AOR, 1.94; 95% CI, 1.3–3.0) and feeling mistreated or not properly attended to at HIV care (AOR, 1.60; 95% CI, 1.0–2.5) increased odds, while posttest counseling (AOR, 0.34; 95% CI, 0.2–0.6) decreased odds. Conclusions More attention is needed on how policies, programs, and research can provide tailored support across the treatment continuum.

RESUMEN Objetivo La participación temprana y continua en el continuo de tratamiento de la infección por el VIH puede ayudar a lograr la supresión viral, aunque pocos estudios han explorado la manera en que los factores de riesgo debidos a los retrasos difieren a lo largo del proceso continuo. El objetivo de este estudio fue determinar los factores predictivos de un diagnóstico tardío, la demora en la vinculación con la atención y el incumplimiento del régimen terapéutico en la ciudad de Salvador de Bahía, en Brasil. Métodos Se recopilaron datos en un estudio transversal realizado durante el 2010, con una muestra (n = 1 970) de personas con la infección por el VIH que recibían atención. Por medio de múltiples análisis de regresión logística se determinaron variables sociodemográficas, comportamientos y mediciones de la calidad del servicio de salud que estaban asociadas a un diagnóstico tardío, así como a la demora en la vinculación con la atención y el incumplimiento del régimen terapéutico. Resultados Con relación a los diagnósticos tardíos, ser de sexo masculino (razón de posibilidades ajustada (AOR) 3,02; intervalo de confianza (IC) de 95%, 2,0-4,6), tener más de 45 años (AOR, 1,67; IC de 95%, 1,1-2,5) y que las pruebas hayan sido iniciadas por el proveedor (AOR, 3,00; IC de 95%, 2,1-4,4) aumentaron las probabilidades, mientras que el consumo de drogas (AOR, 0,29; IC de 95%, 0.2-0.5) y la recepción de los resultados en un espacio privado (AOR, 0,37; IC de 95%, 0,2-0,8) disminuyeron las probabilidades. Con respecto a la vinculación tardía a la atención, el desempleo (AOR, 1,42; IC de 95%, 1,07-1,9) y las dificultades para comprender al trabajador de salud o hablar con él (AOR, 1,61; IC de 95%, 1,2-2,1) aumentaron las probabilidades, mientras que la orientación posterior a las pruebas (AOR, 0,49; IC de 95%, 0,3-0,7) disminuyó las probabilidades. Con respecto al incumplimiento del régimen terapéutico, la discriminación verbal o física relacionada con la infección por el VIH (AOR, 1,94; IC de 95%, 1,3-3,0) y el sentirse maltratado o no atendido adecuadamente en los establecimientos de atención (AOR, 1,60; IC de 95%, 1,0-2,5) aumentaron las probabilidades, mientras que la orientación posterior a las pruebas (AOR, 0,34; IC de 95%, 0,2-0,6) disminuyó las probabilidades. Conclusiones Debe prestarse mayor atención a la forma en que las políticas, los programas y la investigación pueden prestar apoyo personalizado a lo largo del proceso continuo de tratamiento.

Resumo: Estratégias de amostragem como respondent-driven sampling (RDS) e time-location sampling (TLS) são importantes opções metodológicas para acessar populações chaves para a epidemia de HIV/AIDS como homens que fazem sexo com homens (HSH) e mulheres transexuais. Porém são poucos os estudos que avaliam desafios de implementação desses métodos. Objetivamos apresentar os principais desafios na implementação dos RDS no Brasil e TLS no Peru. No Brasil, o local da pesquisa, a seleção das sementes, e o tipo de ressarcimento foram desafios na implementação do RDS-HSH. E no Peru, as questões geográficas, preocupações com segurança, e o tempo necessário para participar no estudo complicaram o TLS com HSH e mulheres transexuais. Pesquisa formativa, participação das populações chaves em todas as fases da pesquisa, e transparência no desenvolvimento do estudo são necessários para utilização prática dos resultados. Melhorar a implementação dessas pesquisas pode contribuir para aumentar o acesso aos serviços de prevenção e controle do HIV/AIDS entre aqueles que mais precisam.

Abstract: Sampling strategies such as respondent-driven sampling (RDS) and time-location sampling (TLS) offer unique opportunities to access key populations such as men who have sex with men (MSM) and transgender women. Limited work has assessed implementation challenges of these methods. Overcoming implementation challenges can improve research quality and increase uptake of HIV services among key populations. Drawing from studies using RDS in Brazil and TLS in Peru, we summarize challenges encountered in the field and potential strategies to address them. In Brazil, study site selection, cash incentives, and seed selection challenged RDS implementation with MSM. In Peru, expansive geography, safety concerns, and time required for study participation complicated TLS implementation with MSM and transgender women. Formative research, meaningful participation of key populations across stages of research, and transparency in study design are needed to link HIV/AIDS research and practice. Addressing implementation challenges can close gaps in accessing services among those most burdened by the epidemic.

Resumen: Estrategias de muestra como respondent-driven sampling (RDS) y time-location sampling (TLS) son importantes opciones metodológicas para acceder a poblaciones claves para la epidemia de VIH/SIDA como los hombres que tienen sexo con hombres (HSH) y mujeres transexuales. No obstante, son pocos los estudios que evalúan los desafíos de implementación de esos métodos. Tenemos como objetivo presentar los principales desafíos en la implementación de los RDS en Brasil y TLS en Perú. En Brasil, el lugar de la investigación, la selección de las muestras, y el tipo de compensación fueron desafíos en la implementación del RDS-HSH. Y en Perú, las cuestiones geográficas, preocupaciones con la seguridad, y el tiempo necesario para participar en el estudio complicaron el TLS con HSH y mujeres transexuales. Se trata de una investigación formativa, con participación de las poblaciones claves en todas las fases de la investigación, y transparencia en el desarrollo del estudio, todo ello necesario para la utilización práctica de los resultados. Mejorar la implementación de esas investigaciones puede contribuir a aumentar el acceso a los servicios de prevención y control del VIH/SIDA entre aquellos que más lo necesitan.

Resumo: Estratégias de amostragem como respondent-driven sampling (RDS) e time-location sampling (TLS) são importantes opções metodológicas para acessar populações chaves para a epidemia de HIV/AIDS como homens que fazem sexo com homens (HSH) e mulheres transexuais. Porém são poucos os estudos que avaliam desafios de implementação desses métodos. Objetivamos apresentar os principais desafios na implementação dos RDS no Brasil e TLS no Peru. No Brasil, o local da pesquisa, a seleção das sementes, e o tipo de ressarcimento foram desafios na implementação do RDS-HSH. E no Peru, as questões geográficas, preocupações com segurança, e o tempo necessário para participar no estudo complicaram o TLS com HSH e mulheres transexuais. Pesquisa formativa, participação das populações chaves em todas as fases da pesquisa, e transparência no desenvolvimento do estudo são necessários para utilização prática dos resultados. Melhorar a implementação dessas pesquisas pode contribuir para aumentar o acesso aos serviços de prevenção e controle do HIV/AIDS entre aqueles que mais precisam.

Abstract: Sampling strategies such as respondent-driven sampling (RDS) and time-location sampling (TLS) offer unique opportunities to access key populations such as men who have sex with men (MSM) and transgender women. Limited work has assessed implementation challenges of these methods. Overcoming implementation challenges can improve research quality and increase uptake of HIV services among key populations. Drawing from studies using RDS in Brazil and TLS in Peru, we summarize challenges encountered in the field and potential strategies to address them. In Brazil, study site selection, cash incentives, and seed selection challenged RDS implementation with MSM. In Peru, expansive geography, safety concerns, and time required for study participation complicated TLS implementation with MSM and transgender women. Formative research, meaningful participation of key populations across stages of research, and transparency in study design are needed to link HIV/AIDS research and practice. Addressing implementation challenges can close gaps in accessing services among those most burdened by the epidemic.

Resumen: Estrategias de muestra como respondent-driven sampling (RDS) y time-location sampling (TLS) son importantes opciones metodológicas para acceder a poblaciones claves para la epidemia de VIH/SIDA como los hombres que tienen sexo con hombres (HSH) y mujeres transexuales. No obstante, son pocos los estudios que evalúan los desafíos de implementación de esos métodos. Tenemos como objetivo presentar los principales desafíos en la implementación de los RDS en Brasil y TLS en Perú. En Brasil, el lugar de la investigación, la selección de las muestras, y el tipo de compensación fueron desafíos en la implementación del RDS-HSH. Y en Perú, las cuestiones geográficas, preocupaciones con la seguridad, y el tiempo necesario para participar en el estudio complicaron el TLS con HSH y mujeres transexuales. Se trata de una investigación formativa, con participación de las poblaciones claves en todas las fases de la investigación, y transparencia en el desarrollo del estudio, todo ello necesario para la utilización práctica de los resultados. Mejorar la implementación de esas investigaciones puede contribuir a aumentar el acceso a los servicios de prevención y control del VIH/SIDA entre aquellos que más lo necesitan.

ABSTRACT OBJECTIVE To analyze the factors associated with late presentation to HIV/AIDS services among heterosexual men. METHODS Men infected by HIV who self-identified as heterosexual (n = 543) were included in the study. Descriptive, biivariate and logistic regression analyses were performed to evaluate the factors associated with late presentation (defined as individuals whose first CD4 count was <350 cells/mm3) in the study population. RESULTS The prevalence of late presentation was 69.8%. The multivariate logistic analysis showed testing initiated by the provider (ORadjusted 3.75; 95%CI 2.45–5.63) increased the odds of late presentation. History of drug use (ORadjusted 0.59; 95%CI 0.38–0.91), history of having sexually transmitted infections (ORadjusted 0.64; 95%CI 0.42–0.97), and having less education (ORadjusted 0.63; 95%CI 0.41–0.97) were associated with a decreased odds of LP. CONCLUSIONS Provider initiated testing was the only variable to increase the odds of late presentation. Since the patients in this sample all self-identified as heterosexual, it appears that providers are not requesting they be tested for HIV until the patients are already presenting symptoms of AIDS. The high prevalence of late presentation provides additional evidence to shift towards routine testing and linkage to care, rather than risk-based strategies that may not effectively or efficiently engage individuals infected with HIV.

ABSTRACT OBJECTIVE To analyze the factors associated with late presentation to HIV/AIDS services among heterosexual men. METHODS Men infected by HIV who self-identified as heterosexual (n = 543) were included in the study. Descriptive, biivariate and logistic regression analyses were performed to evaluate the factors associated with late presentation (defined as individuals whose first CD4 count was <350 cells/mm3) in the study population. RESULTS The prevalence of late presentation was 69.8%. The multivariate logistic analysis showed testing initiated by the provider (ORadjusted 3.75; 95%CI 2.45–5.63) increased the odds of late presentation. History of drug use (ORadjusted 0.59; 95%CI 0.38–0.91), history of having sexually transmitted infections (ORadjusted 0.64; 95%CI 0.42–0.97), and having less education (ORadjusted 0.63; 95%CI 0.41–0.97) were associated with a decreased odds of LP. CONCLUSIONS Provider initiated testing was the only variable to increase the odds of late presentation. Since the patients in this sample all self-identified as heterosexual, it appears that providers are not requesting they be tested for HIV until the patients are already presenting symptoms of AIDS. The high prevalence of late presentation provides additional evidence to shift towards routine testing and linkage to care, rather than risk-based strategies that may not effectively or efficiently engage individuals infected with HIV.

O artigo avalia como a descentralização dos recursos e ações do Programa Nacional de DST e AIDS influenciou as transferências de recursos para os programas de prevenção ao HIV/AIDS entre usuários de drogas injetáveis no Rio de Janeiro, Brasil (1999-2006). Foram avaliados os efeitos da política de descentralização sobre as transferências de recursos para os Programas de Redução de Danos (PRD) no Rio de Janeiro. A descentralização das transferências para os PRD no Rio de Janeiro se efetivou em 2006, com a virtual supressão do financiamento federal direto, obrigando estes programas a buscarem fontes alternativas de financiamento. Constata-se uma fragilidade dos PRD no Estado, agravada pela redução recente do volume de recursos. Dos 22 programas, existentes em 2002, o Estado passou a contar, em 2006, com apenas dois programas, financiados pelo Estado e por um município. A descontinuidade dessas ações pode vir a favorecer a reemergência da epidemia de AIDS nesta população, o que reclama um processo de descentralização mais gradual e melhor pactuado dessas iniciativas.

This paper assesses how decentralization of resources and initiatives by the Brazilian National SDT/AIDS Program has impacted the transfer of funds for programs to prevent HIV/AIDS among injecting drug users in Rio de Janeiro, Brazil (1999-2006). The effects of the decentralization policy on Rio de Janeiro's Syringe Exchange Programs (SEPs) are assessed in detail. Decentralization effectively took place in Rio de Janeiro in 2006, with the virtual elimination of any direct transfer from the Federal government. The elimination of direct transfers forced SEPs to seek alternative funding sources. The structure of local SEPs appears to be weak and has been further undermined by current funding constraints. Of 22 SEPs operating in 2002, only two are still operational in 2006, basically funded by the State Health Secretariat and one municipal government. The current discontinuity of SEP operations may favor the resurgence of AIDS in the IDU population. A more uniform, regulated decentralization process is thus needed.

O artigo avalia como a descentralização dos recursos e ações do Programa Nacional de DST e AIDS influenciou as transferências de recursos para os programas de prevenção ao HIV/AIDS entre usuários de drogas injetáveis no Rio de Janeiro, Brasil (1999-2006). Foram avaliados os efeitos da política de descentralização sobre as transferências de recursos para os Programas de Redução de Danos (PRD) no Rio de Janeiro. A descentralização das transferências para os PRD no Rio de Janeiro se efetivou em 2006, com a virtual supressão do financiamento federal direto, obrigando estes programas a buscarem fontes alternativas de financiamento. Constata-se uma fragilidade dos PRD no Estado, agravada pela redução recente do volume de recursos. Dos 22 programas, existentes em 2002, o Estado passou a contar, em 2006, com apenas dois programas, financiados pelo Estado e por um município. A descontinuidade dessas ações pode vir a favorecer a reemergência da epidemia de AIDS nesta população, o que reclama um processo de descentralização mais gradual e melhor pactuado dessas iniciativas.

This paper assesses how decentralization of resources and initiatives by the Brazilian National SDT/AIDS Program has impacted the transfer of funds for programs to prevent HIV/AIDS among injecting drug users in Rio de Janeiro, Brazil (1999-2006). The effects of the decentralization policy on Rio de Janeiro's Syringe Exchange Programs (SEPs) are assessed in detail. Decentralization effectively took place in Rio de Janeiro in 2006, with the virtual elimination of any direct transfer from the Federal government. The elimination of direct transfers forced SEPs to seek alternative funding sources. The structure of local SEPs appears to be weak and has been further undermined by current funding constraints. Of 22 SEPs operating in 2002, only two are still operational in 2006, basically funded by the State Health Secretariat and one municipal government. The current discontinuity of SEP operations may favor the resurgence of AIDS in the IDU population. A more uniform, regulated decentralization process is thus needed.