2011 HSR&D National Meeting Abstract
1078 — Ten-Year Trends in the Quality of Care and Racial Disparities after the Veterans Affairs Organizational Transformation
Trivedi AN (Providence VA Medical Center), Grebla RC
(Providence VA Medical Center), Wright SM
(VA Office of Quality and Performance), Washington DL
(Greater Los Angeles VA Medical Center)
We assessed trends in overall quality and racial disparities in quality during the 10-year period following the VA’s organizational transformation.
We linked individual-level data from the VA’s External Peer Review Program (EPRP) on adherence to quality of care indicators from 2000 to 2009 with Medical SAS datasets which provided sociodemographic characteristics. We supplemented race data from Medical SAS with Medicare enrollment data, which reduced missing race data to < 3%. For each quality indicator, we used generalized linear regression to assess the independent effect of race, year, and a race-year interaction on achievement of that indicator, adjusting for demographic characteristics, Census region, and a VAMC-level fixed effect. The sample included 1,446,604 white and 235,294 black VA enrollees.
Black enrollees were younger, more likely to be residing in the South, and had lower area-level income and education than white enrollees. With the exception of breast cancer screening, aggregate performance improved over time for all indicators. Absolute differences in performance rates between white and black enrollees were less than 2 percentage points for 5 of 6 process of care measures during each study year. However, disparities for the four intermediate outcomes indicators ranged from 5.5 percentage points for HbA1c control in diabetes to 8.0 percentage points for cholesterol control following an acute coronary event (p < 0.01 for white-black comparisons). There were modest declines in racial disparity for blood pressure control (7.7 to 4.9 percentage points; p < 0.01 for race-year interaction) and cholesterol control after an acute myocardial infarction (9.5 to 7.4 percentage points; p < 0.01 for race-year interaction). Racial disparities were statistically unchanged for HbA1c control and cholesterol control in diabetes. Adjustments for VAMC, Census region, and area-level socioeconomic status produced minimal change in these disparities.
The quality of care improved and racial disparities were minimal for most measures of the process of care from 2000-2009. However, these improvements were not accompanied by meaningful reductions in racial disparity for important clinical outcomes.
VA clinicians and administrators should focus efforts on understanding mediators of poor control of intermediate outcomes among black veterans and consider assessing racial disparities in clinical performance at the level of the VAMC.