 |
|
|
| |
In-Hospital Death Risk 4 Times Higher in Blacks Than Whites With HIV
|
| |
| |
48th ICAAC
October 25-28, 2008
Washington, DC
Mark Mascolini
HIV-infected blacks admitted to the hospital had a 4 times higher risk or dying there than whites in a 1996-2005 US study, independently of opportunistic infection, hepatitis C virus (HCV) coinfection, cocaine use, or age over 49 years [1]. Black race also independently predicted a longer hospital stay.
Researchers analyzed the National Hospital Discharge Survey from 1996 through 2005 to track death and hospital-stay rates in 14,153 adults with an HIV diagnosis. They excluded anyone who left the hospital against medical advice. The study group included 7270 blacks and 3840 whites with a median age of 41 years. A high number of people in this database, 4229, got admitted via the emergency room.
Higher proportions of whites than blacks were men (82% versus 62%) and had HCV coinfection (10% versus 8%). And a larger share of blacks than whites had a record of substance abuse (25% versus 20%) and had an opportunistic infection (39% versus 37%). Although these differences were statistically significant, the last three comparisons may have reached significance only because of the large size of the cohort; they do not seem clinically meaningful.
A multivariate model considering race, cocaine use, HCV coinfection, opportunistic infection, and age over 49 years found that all variables except HCV independently raised the risk of dying in the hospital. Blacks had more than a 4 times higher risk of death (likelihood by chi-squared ratio test 4.25, P = 0.0391). An opportunistic infection independently raised the risk of death almost 14 times (likelihood 13.92, P = 0.0002). Age above 49 made death almost 20 times more likely (likelihood 19.54, P < 0.0001). And cocaine use upped the risk of death 45 times (likelihood 45.27, P < 0.0001).
Being black independently lengthened the hospital stay (likelihood 3.86, P = 0.0496), as did age over 49 (likelihood 29.35, P < 0.0001) and opportunistic infection (likelihood 345.24, P < 0.0001). Cocaine use more than tripled the chance of a longer hospital stay, but that impact fell short of statistical significance (likelihood 3.47, P = 0.06). HCV coinfection did not make a longer hospital stay more likely.
The database did not include other variables that can have a profound impact on mortality risk, such as length of HIV infection, CD4 count, and rates or severity of non-AIDS diseases. The database did include insurance status, but the investigators did not factor insurance into their prediction models. Nor did they report causes of death.
Reference
1. Oramasionwu CU, Ryan L, Frei CR. Disparities in comorbid conditions among white and black HIV/AIDS patients in the United States National Hospital Discharge Survey. 48th Annual International Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 25-28, 2008. Washington, DC. Abstract H-445.
|
| |
|
|
|
|
|
