Effects of Alcohol Use on Patient Retention in HIV Care in East Africa.

ABSTRACT

We sought to investigate the association between hazardous alcohol use and gaps in care for people living with HIV over a long-term follow-up period. Adults who had participated in our previously published Phase I study of hazardous alcohol use at HIV programs in Kenya and Uganda were eligible at their 42 to 48 month follow-up visit. Those who re-enrolled were followed for an additional ~ 12 months. Hazardous alcohol use behavior was measured using the Alcohol Use Disorders Identification Test (AUDIT) tool. Deidentified clinical data were used to assess gaps in care (defined as failure to return to clinic within 60 days after a missed visit). The proportion of patients experiencing a gap in care at a specific time point was based on a nonparametric moment-based estimator. A semiparametric Cox proportional hazard model was used to determine the association between hazardous alcohol use at enrollment in Phase I (AUDIT score ≥ 8) and gaps in care. Of the 731 study-eligible participants from Phase I, 5.5% had died, 10.1% were lost to follow-up, 39.5% transferred, 7.5% declined/not approached, and 37.3% were enrolled. Phase II participants were older, had less hazardous drinking and had a lower WHO clinical stage than those not re-enrolled. Hazardous drinking in the re-enrolled was associated with a Hazard Ratio (HR) of 1.88 [p-value = 0.016] for a gap in care. Thus, hazardous alcohol use at baseline was associated with an increased risk of experiencing a gap in care and presents an early target for intervention.

RESUMEN

RESUMEN Buscamos investigar la asociación entre el uso riesgoso de alcohol y retención en programas de VIH a largo plazo. Todo adulto que participó en nuestro estudio previamente publicado sobre el uso riesgoso de alcohol en programas de VIH en Kenia y Uganda era elegible a los 42 a 48 meses de seguimiento. Los adultos reinscritos en la fueron seguidos por ~ 12 meses adicionales. Usamos el "Alcohol Use Disorders Identification Test" (AUDIT) para medir uso de alcohol. Usamos datos clínicos anonimizados para evaluar interrupciones en cuidado (definido como falta de regresar a clínica 60 días después de faltar a una cita). Basamos la proporción de pacientes con una interrupción en cuidado clínico en un estimador momentáneo y no-paramétrico. Determinamos la asociación entre el uso riesgoso de alcohol al inicio de la primera fase (puntuación AUDIT ≥8) con retención en servicios clínicos usando un modelo de riesgo Cox semiparamétrico. De los 731 participantes elegibles, 5.5% habían muerto, 10.1% fueron perdidos a seguimiento clínico, 39.5% se transfirieron a otro programa, 7.5% declinaron participación o no fueron reclutados y 37.3% fueron reinscritos en la segunda fase. Los participantes reinscritos eran mayores, tenían menos uso riesgoso de alcohol y tenían VIH menos avanzado. El uso peligroso del alcohol se vio asociado con el riesgo de tener una interrupción en cuidado clínico [Proporción de Riesgo (Hazard Ratio, HR) PR=1.88, valor-p = 0.016]. Por lo tanto, el uso peligroso del alcohol incrementa el riesgo de perder seguimiento clínico y presenta una oportunidad para intervención.