Abstract
BACKGROUND: The impact of adherence enhancing interventions on the relationship between active drug use and adherence is largely unknown. METHODS: We conducted a 24-week randomized controlled trial of antiretroviral directly observed therapy (DOT) vs. treatment as usual (TAU) among HIV-infected methadone patients. Our outcome measure was pill count antiretroviral adherence, and our major independent variables were treatment arm (DOT vs. TAU) and active drug use (opiates, cocaine, or both opiates and cocaine). We defined any drug use as ≥ 1 positive urine toxicology result, and frequent drug use as ≥ 50% tested urines positive. We used mixed-effects linear models to evaluate associations between adherence and drug use, and included a treatment arm-by-drug use interaction term to evaluate whether DOT moderates associations between drug use and adherence. RESULTS: 39 participants were randomized to DOT and 38 to TAU. We observed significant associations between adherence and active drug use, but these were limited to TAU participants. Adherence was worse in TAU participants with any opiate use than in TAU participants without (63% vs. 75%, p<0.01); and worse among those with any polysubstance (both opiate and cocaine) use than without (60% vs. 73%, p=0.01). We also observed significant decreases in adherence among TAU participants with frequent opiate or frequent polysubstance use, compared to no drug use. Among DOT participants, active drug use was not associated with worse adherence. CONCLUSIONS: Active opiate or polysubstance use decreases antiretroviral adherence, but the negative impact of drug use on adherence is eliminated by antiretroviral DOT.