Miguel Hernan, Associate Professor of Epidemiology
Department of Epidemiology
Dates of Research:
August 15, 2007 — July 31, 2009
This study will use several HIV cohort studies to answer key questions regarding the use of HAART in the clinical management of HIV-infected individuals. For example, when to initiate HAART? Which HAART regimes work better as initial regimes? To answer these questions, the study will apply the methods of Robins and collaborators: inverse probability weighting, g-estimation, and the parametric g-formula. Observational cohorts of HIV-infected individuals are more cost-efficient than randomized clinical trials, use less restrictive eligibility criteria than randomized trials, and have so far provided estimates of the efficacy of HAART very close to those obtained from randomized trials. This study will pool information from five ongoing prospective cohort studies that total over 100,000 HIV-infected individuals: the Veterans Aging Cohort Study (USA), the Swiss HIV Cohort Study (Switzerland), the French Hospital Database HIV Study (France), GEMES (Spain), and PISCIS (Catalonia and Balearic Islands, Spain).