Abstract 18137: Socio-economic and Racial Differences in Admissions to America's Lowest Performing AMI Hospitals: Implications for Pay for Performance
Background: Low performing hospitals have unique characteristics and tend to serve vulnerable patients. Little is known about hospital selection for patients living in proximity to low performing hospitals.
Methods: We used 2004 -06 Hospital Compare data to calculate composite AMI performance scores for hospitals. We categorized hospitals into low performing (lowest quintile score for all 3 years), high performing (top quintile score for all 3 years), and intermediate (all others). We used 2004-05 Medicare data to identify AMI patients living within 5 miles of a low performing hospital (n = 22365) and hospitals within 30 miles of patient home zip. We ascertained bypass when a patient was admitted to a hospital located at least 1 mile farther than the nearest low performing hospital. We used multinomial logit models to estimate the odds of selecting a low performing hospital over all available hospitals while accounting for distance to hospitals, provision of revascularization and patient socio-demographic characteristics. We used logistic regression to adjust odds of bypass for patient characteristics.
Results: One third of patients (n=7362) were admitted to low performing hospitals; of these, 58% did not have alternative hospitals within 10 miles. Patients admitted to low performing hospitals were older (81 vs 79), lived closer (2 vs 13 miles), more likely to live in zip codes with high social disadvantage (43% vs 38%) and had higher predicted mortality (21% vs 17%). In adjusted analyses, patients were less likely to select low performing hospitals over alternative hospitals (OR 0.33, 95%CI 0.31-0.35, p<.01). However, blacks and patients from socially disadvantaged areas were relatively more likely to select these hospitals (OR 1.25, 95% CI 1.16-1.34 and OR 1.67, 95% CI 1.54-1.78, p <.01). Blacks and patients from disadvantaged areas were also less likely to bypass the nearest low performing hospital (OR 0.83, 95% CI 0.75-0.93 and OR 0.83, 95% CI 0.78-0.89, p < .01).
Conclusion: Blacks and socially disadvantaged AMI patients are more likely to seek care at low performing hospitals and may not have alternative hospitals nearby. Punitive measures under proposed P4P programs could adversely impact care for vulnerable patients at these hospitals and widen disparities.
- © 2010 by American Heart Association, Inc.