|| Checking for direct PDF access through Ovid
Mortality has declined in most HIV-infected populations yet remains high among those with barriers to accessing antiretroviral (ARV) therapy. We sought to determine predictors of death in a group of HIV-infected homeless persons in San Francisco. Between 1996 and 2002, quarterly interviews and blood draws were conducted. Hazards of death were compared by number of months of the prior 6 months that an individual took any ARV, drug use, hepatitis C virus (HCV) status, and housing status. Among 330 participants, 65% were HCV-seropositive at baseline, 85% received ARV during the study period, and there were 57 deaths (5.3 per 100 person-years). Compared with 0 of the prior 6 months on therapy, the risk of death was not significantly reduced for individuals on 1 to 5 months of therapy (hazard ratio [HR] = 0.82, 95% confidence interval [CI]: 0.43-1.57), but the risk of death was reduced 62% for those on ARV therapy for 6 months (HR = 0.38, CI: 0.19-0.76). Housing status and HCV status were not significant predictors of death. HIV is the major cause of death in this population, whereas the impact of HCV infection seems to be minimal. Sustained ARV treatment significantly reduces the risk of death among the homeless.