National Meeting 2008

1066 — Comparison of the Performance of the VHA versus the Medicare Advantage Plans: A National Comparison of two Large Systems Using Health Outcomes

Selim AJ (Center for Health Quality, Outcomes and Economic Research (CHQOER), Bedford VAMC) , Berlowitz D (CHQOER, Bedford VAMC), Kazis LE (CHQOER, Bedford VAMC), Rogers W (Health Institute, New England Medical Center, Boston, MA), Qian S (CHQOER, Bedford VAMC), Rothendler JA (CHQOER, Bedford VAMC), Spiro, III A (MAVERIC, VA Boston Healthcare System), Miller D (CHQOER, Bedford VAMC), Selim B (Boston University School of Medicine), Fincke B (CHQOER, Bedford VAMC)

In 1995, VHA launched a major healthcare transformation by adopting managed care principles, new information technology, measurement and reporting of performance, and integration of services with realignment of payment policies. However, few studies have compared patient outcomes in the VHA with other health care systems using data collected before or early in the VHA transformation. We sought to compare Medicare Advantage plans (MAP) which provide comprehensive health care to 4.6 million Medicare enrollees through contracted managed care organizations across the US with the VHA, the nation's largest integrated health care system.

This was a retrospective cohort design with men 65 years and older receiving care in MAP (n=198,421) who are randomly selected and in the VHA (n=360,316) over 2 years. Comparisons were made between the two systems using the Veterans RAND 12-Item Health Survey (VR-12) physical (PCS) and mental (MCS) summary scores before/early on in the transformation and much later. Case mix adjustments using multivariate models (sociodemographics – age, race/ethnicity, marital status, education and income; co-morbidities and baseline health status) included the adjusted probability of being alive with the same or better physical and mental health at 2 years of follow-up. Outcomes in specific vulnerable patient populations were also examined.

VHA had a higher adjusted probability of being alive with the same or better PCS compared to MAP participants (VHA 69.2% vs. MAP 64.3%, p < 0.001) and for same or better MCS (76.1% vs. 69.6%, p < 0.001) also favoring VHA. We obtained similar findings across vulnerable subpopulations for the old, African Americans, Hispanics, and those with hypertension, diabetes, CAD/MI (p < 0.0025, with Bonferroni correction).

After case-mix adjustment, we found that the VHA had substantially better health outcomes when compared with the MAP for the population as a whole as well as for vulnerable subpopulations. Further studies should examine what differences in organizational structures and processes of care contribute to the better health outcomes in the VHA.

This study provides an important basis for understanding best practices in the VHA and future work that targets health care beneficial to veterans.