PubMedDate not supplied
PURPOSE: To assess variables predictive of nonadherence persisting over time in HIV-infected people treated with highly active antiretroviral therapy. METHOD: Prospective study of consecutive HIV-infected patients who were prescribed ritonavir- or indinavir-containing regimens in a university-based HIV clinic in Rome. A patient questionnaire assessing knowledge of treatment regimen, adherence behavior, reasons for taking and missing therapy, factors influencing adherence, and health behaviors was administered at baseline and 1 year later. A predose protease inhibitor plasma level was measured concurrently. Persistent nonadherence was defined as patient self-reported nonadherence both at enrollment and at follow-up questionnaires. RESULTS: From April 1998 to July 1998, 140 patients were enrolled into the study. At follow-up, 10% remained persistently nonadherent, and 15% of the previously adherent patients became nonadherent. On bivariate analysis, being less than 35 years old (odds ratio [OR] 8.9; 95% CI 1.8-43.1; p =.002), self-reporting nonadherence at enrollment (OR 14.5; 95% CI 3.5-5.8; p <.001), and having experienced "a fair amount" or "a lot" of vomiting (OR 11.1;95% CI 1.6-74.7; p =.02) or pruritus (OR 16.4; 95% CI 2.6-102.8; p =.004) during the 4 weeks before enrollment were significantly correlated to persistent nonadherence. CONCLUSION: Previous self-reported nonadherence was a strong predictor of persistent nonadherence during follow-up. Moreover, being of younger age and self-reporting vomiting or pruritus were also associated with a higher risk of nonadherence persisting over time.