2006 HSR&D National Meeting Abstract
3040 — Risk Adjusted Mortality as an Indicator of Outcomes
Kazis LE (CHQOER)
Rogers W (Health Institute)
Qian S (CHQOER)
Rothendler J (CHQOER)
Lee A (CHQOER)
Ren X (CHQOER)
Haffer S (Centers of Medicare and Medicaid)
Miller D (CHQOER)
Spiro A (CHQOER)
Fincke G (CHQOER)
The Medicare Advantage Program (MAP) and the Veterans Health Administration (VHA) currently provide many services that benefit the elderly, and a comparative study of their risk-adjusted mortality rates has the potential to provide important information regarding these two systems of care. We compared mortality rates between the MAP and the VHA after controlling for case-mix differences.
This is a retrospective study of 584,294 MAP patients from the Medicare Health Outcomes Survey and 420,514 VHA patients from the 1999 Large Health Survey of Veteran Enrollees. We used the Death Master File to ascertain the vital status of each study subject over approximately 4 years. We used Cox regression models to estimate hazard ratios (HR) with 95% confidence intervals (CI) for the MAP compared with VHA patients.
The average age for male MAP patients was 73.8 (±5.6) and for male VHA patients was 74.05 (±6.3). Unadjusted mortality rates of males for VHA and MAP were 25.7% and 22.8% over approximately 4 years (p<0.0001), respectively. The case-mix of VHA patients, however, was sicker than those from MAP. After adjusting for case-mix, the HR for mortality in the MAP was significantly higher than that in the VHA (HR, 1.404 [95% CI 1.383-1.426]). We obtained similar results when we compared the mortality rates of females for VHA and MAP.
After adjusting for their higher prevalence of chronic disease and worse self-reported health, mortality rates were lower for patients cared for in the VHA compared to those in the MAP.
Further studies should examine what differences in care structures and processes contribute to lower mortality in the VHA.