Abstract 4: The Association Between Admission Rate and 30-Day Risk Standardized Readmission Rate for Heart Failure and Acute Myocardial Infarction

Abstract

Background: There is wide variation in regional admission and readmission rates and the association between the two is unknown. We sought to determine this association and whether it differs between heart failure (HF) admissions, more discretionary, and acute myocardial infarction (AMI) admissions, less discretionary. Methods: We used 2006-2008 Medicare ICD-9-CM claims data to determine AMI and HF admissions. Hospital referral region (HRR) admission rates were determined using Medicare denominator files. We estimated hospital level 30-day risk standardized readmission rates using hierarchical logistic models adjusting for age, sex and comorbidities and then aggregated them to HRRs level RSRRs. The correlation (R 2 ) obtained by weighted linear regression characterized the relationship between admission rates and RSRRs for AMI and HF. Results: Among 306 HRRs, the median AMI admission was 9.2 per 1000 enrollees (range 2.8 to 26.5) and the median HF admission rate was 19.6 (range 7.6 to 41.4). Among the 306 HRRs, the median AMI RSRR was 0.196 (range 0.163 to 0.233) and the median HF RSRR was 0.242 (range 0.201 to 0.289). Regional AMI admission rate was weakly correlated with AMI RSRR (R 2 0.05, 95% CI 0.02 - 0.11) and HF RSRR (R 2 0.07, 95% CI 0.02 - 0.13). Regional HF admission rate was modestly correlated with both AMI RSRR (R 2 0.28, 95% CI 0.20 - 0.37) and HF RSRR (R 2 0.32, 95% CI 0.24 - 0.41). Conclusion: The modest association between HF admissions, a more discretionary admission condition, and both AMI and HF readmission rates suggest a systemic propensity to admit and readmit patients. In contrast, the same was not true of AMI, a less discretionary admission condition.