Rates of hospitalizations and associated diagnoses in a large multisite cohort of HIV patients in the United States, 1994-2005.

Objectives: To assess temporal trends in the rates of hospitalizations and associated diagnoses among HIV-infected patients before and during the era of highly active antiretroviral therapy.
Design: A prospective cohort study of 7155 patients enrolled in the HIV Outpatient Study at 10 US HIV clinics.
Methods: We evaluated rates of hospitalizations for major categories of medical conditions during 1994-2005 and modeled trends in these rates using multivariable Poisson regression models for repeated observations. We assessed patient characteristics associated with hospitalization using multiple logistic regression.
Results: The rates of hospitalizations (per 100 person-years) fell from 24.6 in 1994 to 11.8 in 2005 (P < 0.0001). The rates of hospitalizations for AIDS opportunistic infections decreased from 7.6 in 1994-1996 to 1.0 in 2003-2005 (P < 0.0001). AIDS opportunistic infections were present at 31% of hospitalizations in 1994-1996 versus 9.5% in 2003-2005, and chronic end-organ disease conditions were present at 7.2% of such hospitalizations in 1994-1996 versus 14.3% in 2003-2005. Mean CD4+ cell count at hospitalization increased from 115 cells/mul in 1994 to 310 cells/mul in 2005. Factors independently associated with hospitalization in the highly active antiretroviral therapy era (1997-2005) included older age, history of substance abuse, lower CD4+ cell count, history of AIDS, and public health insurance.
Conclusion: The rates of hospitalizations for HIV-infected patients declined substantially during 1994-2005, due mainly to reductions in the AIDS opportunistic infections. Compared with the period 1994-1997, patients in the highly active antiretroviral therapy era were hospitalized with higher CD4+ cell counts and more frequently for chronic end-organ conditions.