In-hospital mortality fell steadily and significantly from 1995 through 2011, according to a large study at a U.S. urban academic center. But during that time the proportion of hospital deaths caused by non-AIDS illnesses rose from under half to almost three-quarters.
Despite steady advances in HIV care over the past few decades, hospital admission rates remain higher for people with HIV than for the general population, observe researchers who conducted the new mortality study. To address limitations in research on hospital mortality in HIV populations, these investigators examined trends in in-hospital death from 1995 to 2011, identified causes of death and attributed deaths to AIDS or non-AIDS illnesses.
The large retrospective analysis involved all HIV patients admitted to the Yale-New Haven Hospital, an urban tertiary care center served by Connecticut's largest ambulatory HIV clinic. The investigators determined which HIV patients died in the hospital from January 1, 1995, to December 31, 2011, and collected relevant data from medical records. They divided the study period into the early antiretroviral therapy (ART) era (1995-2001) and the late-ART era (2002-2011).
Over the study period, 406 of 12,183 HIV-positive inpatients (3.3%) died. Mortality dropped from 6.2% in 1995 to 1.5% in 2011 (P < .0001). Two-thirds of those who died (65.5%) were men, three-quarters (73.3%) were nonwhite and two-thirds (65.9%) were taking ART, but only one-third (31.3%) had a viral load below 400 copies/mL.
Most people (56.3%) died of a non-AIDS cause. But this proportion shifted over the years. In the early-ART era 57% of deaths were AIDS-related, compared with 29.5% in the late-ART era. Conversely, 43% of deaths had non-AIDS causes in the early-ART era, compared with 70.5% in the late-ART era. In the late ART period the 70.5% of non-AIDS deaths consisted mainly of non-AIDS infection (25.4%), cardiovascular disease (15%) and non-AIDS malignancy (13%), followed by liver-related death (8.3%), renal death (5.2%) and other causes (3.6%).
Logistic regression analysis identified four independent predictors of non-AIDS death over the whole study period. The strongest predictor was a last CD4+ cell count above 200 cells/mm3 in the year before death (adjusted odds ratio [aOR] 16.5, 95% confidence interval [CI] 5.3 to 51.4), followed by a viral load at or below 400 copies/mL in the year before death (aOR 7.5, 95% CI 2.3 to 24.2), underlying liver disease (aOR 4.5, 95% CI 2.2 to 9.3) and underlying cardiovascular disease (aOR 4.2, 95% CI 1.8 to 9.9). These four associations grew stronger when the investigators limited the analysis to the late-ART era. Age, gender and race did not predict non-AIDS mortality in these analyses.
"Our findings emphasize that HIV-infected patients remain at high risk for complications from non-AIDS infections, even when their immune system has been restored as measured by the CD4 cell count, and at increased risk of cardiovascular and liver disease, which highlights the need to carefully monitor HIV-positive patients admitted with these conditions," the researchers conclude.
Although much research attention has focused on cardiovascular and liver disease in the current ART era, the authors stress that non-AIDS infection remains the leading non-AIDS cause of death and that the proportion of deaths caused by non-AIDS infection did not decrease significantly over time. The most frequent causes of non-AIDS infection were unspecified sepsis, nonrecurrent bacterial pneumonia and Clostridium difficile infection. The authors urge hospital clinicians to monitor HIV patients for non-AIDS infection and caution that bacterial infection may be a prelude to septic shock.
The number of hospital deaths attributed to non-AIDS malignancy, though small, more than quadrupled from the early ART period to the late ART period. And the reported rate may be an underestimate, the researchers suggest, because of increased use of hospices and community care for people with late-stage cancer.