Introduction: This is an amended application to develop a cognitive-behavioral therapy for HIV medication adherence and major depression. Depression is prevalent in HIV and is associated with poor self-care behaviors including poor adherence to antiretroviral medications. Patients with HIV and depression are at risk for poor health outcomes and possibly increased morbidity. Cognitive-behavioral therapy is the most widely studied and efficacious psychosocial intervention for depression. Overview of project goals and conceptual model of intervention: The main goal of this project is to complete NIH-defined stage 1 activities in developing a cognitive behavioral intervention for depression and ART medication adherence. We propose to estimate the effect size of the intervention on improved depression, improved adherence to medications, and improved health status as defined by a clinically significant reduction in HIV viral load. Following the goals of the R21 mechanism, this will allow for the collection of the necessary pilot data to conduct a full-scale intervention study. We hypothesize that the psychosocial intervention will achieve improved health status in two ways: by directly increasing adherence to antiretroviral medications using the adherence skills training, and by treating the depression which otherwise makes it difficult for patients to acquire or use these adherence skills. Overview of research plan: Patients with a detectable viral load who have a diagnosis of major depressive disorder will be randomized into either: (1) """"""""CBT,"""""""" the combination of CBT for depression and HIV medication adherence or (2)""""""""Enhanced Clinical Management,"""""""" a single-session adherence intervention (Safren et al., 1999, 2001). Those who are assigned to Enhanced Clinical Management will be re-assigned to CBT after the acute phase of the study (4 months) if they have not improved on key outcome variables. Eligibility requirements at this stage of treatment development were selected to maximize the chances of finding an effect with a circumscribed sample (and minimizing the possibility of a type-II error). This will allow for an adequate power analysis for a full-scale intervention trial.
