icon-folder.gif   Conference Reports for NATAP  
 
  ID Week
October 2-6, 2013
San Francisco
Back grey_arrow_rt.gif
 
 
 
Non-AIDS Death Rate Jumps From Early to
Recent ART Era in Large Single-Site Study

 
 
  IDWeek, October 2-6, 2013, San Francisco
 
Mark Mascolini
 
Non-AIDS illnesses accounted for 43% of deaths in the early combination antiretroviral therapy (ART) era (1995-2001) and for 70.5% in the recent ART era (2002-2011), according to a retrospective analysis of 12,183 HIV-positive people discharged from a New Haven, Connecticut hospital [1]. AIDS explained 57% of deaths in 1995-2001 and 29.5% in 2002-2011. Non-AIDS infections, cardiovascular disease, and cancer accounted for most non-AIDS deaths in the more recent era.
 
AIDS remains a leading cause of death in European and North American HIV populations, but the proportion of AIDS-related deaths fell with the advent of combination ART while the proportion of non-AIDS deaths rose. To examine trends in causes of death at an urban teaching hospital affiliated with Yale University, researchers conducted this retrospective chart review of all HIV-positive patients who died after hospital discharge from January 1995 through December 2011. Two clinicians classified causes of death according to published definitions. The Yale team used univariate and multivariate logistic regression analysis to identify factors predicting non-AIDS deaths.
 
Among 12,183 HIV-positive people discharged in the study period, 406 (3.3%) died. The analysis focused on 400 people with detailed records, including 207 in the early-ART era and 193 in the recent-ART era. Median age was significantly older in the recent-ART era than the early-ART era (49 versus 41, P < 0.0001), the proportion on ART higher (74.5% versus 58.3%, P = 0.0008), median CD4 count higher (153 versus 50, P < 0.0001), and the proportion with a viral load below 400 copies higher (41.7% versus 13.3%, P < 0.0001).
 
The percentage of people who died yearly dropped from about 6% in 1995 to about 3.5% in 2002, then continued to decline to about 1.5% in 2011. In the early-ART era (1995-2001), AIDS explained 57% of deaths, compared with 29.5% in the recent-ART era (2002-2011) (P < 0.0001). In contrast, non-AIDS illnesses accounted for 43% of deaths in the early-ART era and 70.5% in the recent-ART era. From the early to the recent-ART era, proportions of deaths due to malignancy rose significantly from (2.9% to 13.0%, P = 0.013). Proportions of deaths due to other major non-AIDS illnesses rose nonsignificantly (non-AIDS infection 15.5% to 25.4%; cardiovascular disease 7.7% to 15%; renal failure 3.9% to 5.2%) or remained largely unchanged (liver disease 8.7% to 8.3%).
 
Comparing AIDS deaths with non-AIDS deaths, the researchers found these significant differences: younger age with AIDS versus non-AIDS deaths (mean 39.7 versus 46.7, P < 0.0001), lower proportion on ART (55.8% versus 74.1%, P = 0.0001), higher proportion with CD4 count below 200 (97.1% versus 47.2%, P < 0.0001), and lower proportion with viral load below 400 copies (4.1% versus 38.1%, P < 0.0001). Compared with people who died of AIDS, those who died of a non-AIDS illness had significantly higher proportions with chronic conditions including diabetes, chronic kidney disease, hepatitis C infection, cirrhosis, hypertension, coronary artery disease, and chronic obstructive pulmonary disease.
 
Multivariate analysis determined that a viral load above 400 copies independently lowered the odds of non-AIDS death more than 85% (odds ratio [OR] 0.133, 95% confidence interval [CI] 0.041 to 0.434), while three factors independently raised the odds of non-AIDS death: liver disease (OR 4.47, 95% CI 2.15 to 9.30), cardiovascular disease (OR 4.19, 95% CI 1.78 to 9.89), and CD4 count above 200 (OR 16.45, 95% CI 5.27 to 51.38).
 
The Yale team believes their findings underline "the importance of early preventive primary care in the ART era to reduce mortality" in people with HIV. They recommended that clinicians focus on "identifying and reducing cardiac risk factors as well as diagnosing and treating concurrent hepatitis C infection." Future work, they proposed, should examine racial and gender differences in non-AIDS mortality to identify populations most at risk.
 
Reference
 
1. Cowell A, Shenoi S, Kyriakides T, Barakat L. Trends in hospital mortality among HIV-infected patients during the antiretroviral therapy era, 1995-2011. IDWeek 2013. October 2-6, 2013. San Francisco. Abstract 330.