Abstract 126: Massachusetts Health Reform Did Not Reduce 30-day Readmissions for Acute Myocardial Infarction among Massachusetts Residents Overall or among Racial/Ethnic Minorities

Abstract

Introduction: After Massachusetts (MA) health reform, the percent of uninsured residents fell, particularly among minorities. Studies have not examined the effect of this policy change on 30-day readmission rates, and disparities in such rates, following acute myocardial infarction (AMI). Hypothesis : 30-day readmission rates will decline in patients age 18-64 (affected by health reform) vs. patients age ≥ 65 (most unaffected) Methods: We used 2004-2009 MA discharge data. We compared all-cause 30-day readmission rates after a hospitalization for AMI overall, and among racial subgroups, for patients age 18-64 vs. age ≥ 65 in the period prior to and following reform. We used chi-square tests to compare unadjusted 30-day readmissions for AMI between groups. Treating patients age 18-64 as the intervention cohort, and those age ≥ 65 as the control cohort, we used logistic regression to conduct difference-in-difference analysis that estimates odds of readmission in the post-reform period vs. the pre-reform period in MA adjusted for secular changes. The model was also adjusted for age, gender, and medical comorbidity. To assess differences in white vs. minority disparities over time between the pre and post reform periods among patients age 18-64 vs. age ≥ 65, we used logistic regression to conduct difference-in-difference-in-differences analysis. Results: Among patients age 18-64, pre-reform and post reform readmission rates were 11.5% and 10.5%. Among patients age ≥ 65, they were 24.0% and 22.2%. The post-reform decrease among patients age 18-64 was not different than the decrease among patients age ≥ 65 (difference-in-difference adjusted OR (AOR) =1.0; 95% CI=0.9-1.2). Among patients age 18-64, blacks had higher readmission rates than whites pre and post-reform. Post-reform changes in readmission rates among non-elderly and elderly adults were not significantly different among whites, blacks, and Hispanics. There was no significant change in the presence of disparities between whites and blacks or between whites and Hispanics among non-elderly and elderly adults pre- and post-reform. Conclusions: A major coverage expansion in MA was not associated with a reduction in 30-day readmissions for AMI overall or a reduction in racial and ethnic disparities.