Excess mortality in patients with AIDS in the era of highly active antiretroviral therapy: temporal changes and risk factors

Abstract

Background: Excess mortality has decreased among human immunodeficiency virus (HIV)-infected patients but without evidence of a decrease among patients with AIDS. We assessed temporal changes in excess mortality and elucidated risk factors for excess mortality in patients with AIDS diagnosed in the era of highly active antiretroviral therapy (HAART).

Methods: We included 1188 patients of the Longitudinal Study of Ocular Complications in AIDS who were aged 25-64 years at enrollment and who received a diagnosis of AIDS after 1995. We calculated excess mortality as the age-, year-, and sex-adjusted difference in mortality rates between patients with AIDS and persons in the US general population during the period 1999-2007. We used a relative survival model to identify risk factors for excess mortality.

Results: There were a mean of 50 excess deaths per 1000 person-years (95% confidence interval [CI], 44-57 excess deaths per 1000 person-years) during 1999-2007. Excess mortality almost halved, with an annual decrease of 8.0% per year (95% CI, 3.0%-12.7%; P = .002) but remained high at 36 excess deaths per 1000 person-years in 2007. Viral load >400 copies/mL (compared with <or= 400 copies/mL; risk ratio, 3.4; 95% CI, 2.3-5.0), CD4(+) count <200 cells/μL (compared with >or= 200 cells/μL; risk ratio, 2.7; 95% CI, 1.9-3.9), and cytomegalovirus retinitis (risk ratio, 1.6; 95% CI, 1.2-2.1) were the strongest risk factors for excess mortality.

Conclusions: Excess mortality among patients with AIDS was nearly halved in the HAART era and most strongly linked to stage of HIV disease. These results reflect the continuing improvements in AIDS management but also highlight that excess mortality remains ∼5 times higher in patients with AIDS than in HIV-infected patients without AIDS.

Figure 1. Decline in excess mortality between 1999 and 2007

The figure shows the decline in excess mortality in the LSOCA AIDS cohort and in the CASCADE HIV cohorts between 1999 and 2007. The upper line in the graph shows the number of excess deaths per 1,000 person years (solid rectangles) in patients with AIDS enrolled in LSOCA and 95% confidence intervals. The dashed line represents the fitted line based on a Poisson regression model. The annual decline in excess mortality averaged 8.0% per year (95% CI 3.0 to 12.7, p=0.002). The lower line in the graph shows the number of excess deaths per 1,000 person years (solid circles) in HIV patients enrolled in the CASCADE cohorts and 95% confidence intervals. The dotted line represents the fitted line based on a Poisson regression model. The annual decline in excess mortality averaged 4.8% (95% CI 3.1 to 6.5, p<0.001).

Figure 2. Decline in Excess Mortality Between 1999 and 2007, Stratified by CD4 Count at Enrollment

The figure shows the decline in excess mortality in the LSOCA AIDS cohort stratified for CD4+ count at study enrollment. The upper line in the graph shows the number of excess deaths per 1,000 person years (solid circles) in patients with low CD4+ counts and 95% confidence intervals. The dashed line represents the fitting line based on a Poisson regression model, which showed an average annual decline in excess mortality of 8.3% (95% CI 2.7 to 13.4, p=0.004). The lower line in the graph shows the number of excess deaths per 1,000 person years (solid rectangles) in patients with moderate to high CD4+ counts at study enrollment and 95% confidence intervals. The dotted line represents the fitting line based on a Poisson regression model. There was no change in excess mortality (1.4%, 95% CI -10.7 to 15.3, p=0.83). The interaction between CD4+ count status and year was not statistically significant (p=0.16).