Adherence to highly active antiretroviral therapies (HAART) in HIV-infected patients: from a predictive to a dynamic approach

To-date, most socio-behavioural research about HIV-infected patients' adherence to highly active antiretroviral therapies (HAART) has been based on cross-sectional studies. The French APROCO cohort gave us the opportunity to conjointly analyse the relationships between short-term adherence to HAART and HIV-infected patients' characteristics before initiation of treatment on the one hand, factors related to patients' subjective experience with HAART on the other hand. At the fourth-month follow-up visit (M4) after first prescription of HAART (MO), 26.71% of our sample of 445 patients self-reported non-adherence behaviour. Some patients' characteristics at MO (younger age, poor housing conditions, lack of social support, and problems of adherence with previous antiretroviral regimens) were related to non-adherence at M4 in multivariate analysis. Non-adherence at M4 was, however, also related to the evolutions that affected a number of factors between MO and M4: levels of depression, symptoms associated with treatment side effects, perception of individual state of health, beliefs towards effectiveness and toxicity of HAART, increases in alcohol and tobacco consumption, as well as contacts with other physicians than hospital HAART prescribers. Our prospective study brings additional evidence that even short-term non-adherence cannot be reliably predicted on the sole basis, of a few a priori patient characteristics that clinicians could easily identify before initiation of HAART. It suggests that a dynamic approach to adherence, continuously monitoring the impact of experience with HAART on patients' daily lives, is needed for improving management of HIV/AIDS care. (C) 2002 Elsevier Science Ltd. All rights reserved.