Mortality in an Urban Cohort of HIV-Infected and At-Risk Drug Users in the Era of Highly Active Antiretroviral Therapy

Abstract

Mortality trends among drug users in the era of highly active antiretroviral therapy (HAART) remain unclear. We examined mortality rates, causes of death, and predictors of mortality in 398 human immunodeficiency virus (HIV)-infected and 656 at-risk drug users for the period of 1996-2001. National death index reports were used to confirm deaths, and causes of death were derived from medical records. Cox proportional hazards models were used to determine factors associated with mortality. During 1996-2001, mortality rates in HIV-infected and HIV-uninfected participants were 7.3 and 1.5 deaths per 100 person-years, respectively (P<.001). The mean age at the time of death was 43.6 years for HIV-infected subjects and 47.7 years in HIV-uninfected subjects (P<.001). For 398 HIV-infected participants who were observed for 1443 person-years, death rates decreased from 11.4 to 5.4 deaths per 100 person-years over the 6-year period (P=.04). Among all participants, causes of death were as follows: HIV/AIDS, 27% of subjects; substance abuse, 31%; bacterial infection, 25%; other medical illness, 14%; and violence, 3%. Persons who initiated HAART at a CD4+ lymphocyte count of 201-350 cells/mm3 experienced improved survival, compared with those who initiated it at a CD4+ lymphocyte count of < or =200 cells/mm3 (P=.01). In a multivariate Cox model of HIV-infected subjects, factors independently associated with mortality included receipt of HAART (adjusted hazard ratio [HR(adj)], 0.44; 95% confidence interval [CI], 0.28-0.68) and CD4+ lymphocyte count of < or =200 cells/mm3 (HR(adj), 4.23; 95% CI, 2.24-7.60). Use of methadone or illicit drugs did not predict mortality. To further reduce mortality among drug users, interventions aimed at improving HAART use are warranted. Preventive health and timely management of treatable conditions, such as bacterial infections, also needs emphasis.