IAA 08-087
VA Facility Determinants of Racial-Ethnic Variations in Quality of Care
Donna L Washington, MD MPH VA Greater Los Angeles Healthcare System, West Los Angeles, CA West Los Angeles, CA Funding Period: June 2009 - November 2011 |
BACKGROUND/RATIONALE:
Though the VA is committed to delivering high quality care to all veterans, racial-ethnic disparities remain in VA healthcare quality. Patient factors such as socio-demographic characteristics do not fully account for observed disparities. VA facility characteristics (i.e., healthcare organizational structures and processes) may contribute to the observed relationships between patient race-ethnicity and VA quality of care. OBJECTIVE(S): We aimed to identify VA facility characteristics associated with quality of care for specific racial-ethnic minority groups, VA facility characteristics associated with racial-ethnic disparities in VA quality of care, and area contextual characteristics associated with healthcare quality and disparities. METHODS: We conducted statistical analyses of merged data from multiple existing data sources. All VA healthcare users with quality indicators measured by the External Peer Review Reporting Program (EPRP) in fiscal years 2006, 2007, and 2008, were identified and linked to create composite measures of quality. These global process and global clinical outcome measures represented receipt of all processes of care for which an individual was eligible, and achievement of all intermediate clinical outcomes, respectively. Mean number of process and outcome measures received were also examined. Survey of Healthcare Experiences of Patients (SHEP) and VA data provided self-reported race-ethnicity. VA facility level organizational and practice characteristics were derived from the VHA Clinical Practice Organizational Survey - a nationwide census of VA facilities and large community-based outpatient clinics that was fielded in 2006-07. Area and other facility characteristics were also derived from the Area Resource File, SHEP, and VA administrative data. We conducted multilevel models comparing African-American and Hispanic veterans to White veterans. FINDINGS/RESULTS: There is significant variation across VA sites in the racial-ethnic distribution of the patient populations they serve, and in other patient and facility characteristics. Among veterans with EPRP quality measures, the overall mean proportion of racial-ethnic minorities was 10% for outpatients and 19% for inpatients, but facility-level percent minority ranged from 1% to 60% for outpatients, and 1% to 64% for inpatients. Facility-level racial-ethnic minority percent was correlated with facility patient income category, with racial-ethnic minorities more likely than whites to have income <$15,000 (inpatient and outpatient odds ratios each 1.4; 95% confidence interval 1.4-1.5). The extent of within-facility disparities varied significantly across sites, and accounted for more variation than between-facility differences. Seventy-four percent of White veterans achieved targets for all clinical outcome measures in contrast to 70% of American Indian/Alaskan Natives, 69% of Hispanics, 69% of Native Hawaiian/Other Pacific Islanders, and 67% of African-Americans. Achievement of clinical outcomes was similar between Asians and Whites. Sixty-seven percent of White veterans received all processes of care for which they were eligible, in contrast to 71% of Asians, and 64% of all other groups. Disparities in healthcare quality were most apparent between African-Americans and Whites for global and individual clinical outcome measures (control of blood pressure, LDL-cholesterol, and diabetes), colorectal cancer screening (CRCS), receipt of immunizations, and tobacco use. Hispanic-White disparities were also present for global clinical outcome measures and CRCS. Facility location in the midwest or western geographic regions was associated with worse clinical outcome measures for African-Americans compared with a Northeast location. VA facility type was associated with Hispanic-White disparities, with use of VA contract rather than VA-owned sites associated with worse clinical outcome measures for Hispanics, but not for Whites. Facility patient mix was associated with quality of care for minorities, with worse blood pressure control for African-Americans in facilities serving greater proportions of uninsured. Recognition of physician performance, and having an on-site primary care training program, were both associated with lower Hispanic-White disparities in CRCS. IMPACT: Area contextual factors, facility characteristics, and primary care practice characteristics are all associated with achievement of quality indicators for racial-ethnic minority VA users, with several of these factors also contributing to VA facility-level healthcare disparities. Focused efforts to improve clinical outcome measures in African-American and Hispanic VA healthcare users should account for facility and local area context. Almost all of the disparities in quality were explained by within-facility disparities, suggesting that VA medical centers should measure and address racial-ethnic gaps in care for their patient populations. Future research should identify potential causal factors within healthcare sites, such as the clinician-patient relationship and encounter, including clinician-patient communication. The significant variation in the racial-ethnic composition of VA sites, coupled with the finding of greater within-facility variation, suggests that some efforts to improve quality for minorities should focus on high-minority-serving sites. External Links for this ProjectDimensions for VADimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.Learn more about Dimensions for VA. VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address. Search Dimensions for this project PUBLICATIONS:Conference Presentations
DRA:
Health Systems Science
DRE: Etiology Keywords: Disparities, Organizational issues, Rural MeSH Terms: none |