Reducing Disparities in Care Using Quality Improvement: Findings from a Multi-Hospital Collaborative

RESEARCH OBJECTIVE: Disparities in healthcare represent a failure in the equity domain of quality. The Institute of Medicine's report Unequal Treatment recommended improved race and ethnicity patient data collection and the use of evidence-based guidelines in efforts to reduce disparities. Yet there is little practical experience with this approach. Our objective was to reduce disparities in cardiovascular care for African Americans and Latinos using these methods in a set of hospitals with large minority patient populations. STUDY DESIGN: Through a national competitive process, 10 institutions with large African American and/or Latino populations were selected to work for 29 months to improve cardiac care. The hospitals were trained in the collection of patient race, ethnicity and language data to allow reporting of stratified clinical quality measures. They were also trained on quality improvement techniques to foster use of evidence-based clinical practice. Performance was measured by monthly collection of 23 quality measures by race, ethnicity and language. These measures include the current Hospital Quality Alliance metrics for heart failure and acute myocardial infarction, as well as new “measures of ideal care” assessing whether patients received all evidence-based care for these conditions. POPULATION STUDIED: The ten hospitals are diverse institutions, serving rural, suburban, and urban markets. Six are not-for-profit, one for-profit, and three publicly owned hospitals. Three are community hospitals, three are teaching hospitals, and four are academic medical centers. PRINCIPAL FINDINGS: The ten hospitals have been able to report quality data by race, ethnicity and language. There were some significant differences in care between racial and ethnic groups within individual hospitals; some of these disparities narrowed during the initial phase of the collaborative. At other hospitals there were no disparities and care improved for all patients. The median hospital percentage of patients receiving all recommended heart failure care rose from 44% to 78% in the first year of the collaborative. CONCLUSIONS: The uniform collection of race, ethnicity and language data in support of disparities reduction is feasible. Use of evidence based quality improvement practices may be able to narrow racial and ethnic differences at individual hospitals, and can improve quality dramatically at institutions serving large numbers of minority patients. IMPLICATIONS FOR POLICY, DELIVERY or PRACTICE: National efforts to improve care could explicitly include disparities reduction as a goal. Incentives and mandates to improve patient race, ethnicity, and language data collection may be worthwhile given that hospitals can collect this data uniformly.