Relative Effects of the Hospital Readmissions Reduction Program on Hospitals That Serve Poorer Patients

IMPORTANCE: Hospitals that serve poorer populations have higher readmission rates. It is unknown whether these hospitals effectively lowered readmission rates in response to the Hospital Readmissions Reduction Program (HRRP). OBJECTIVE: To compare pre-post differences in readmission rates among hospitals with different proportion of dual-eligible patients both generally and among the most highly penalized (i.e. low performing) hospitals. DESIGN: Retrospective cohort study using piecewise linear model with estimated hospital-level RSRRs as the dependent variable and a change point at HRRP passage (2010). Economic burden was assessed by proportion of dual-eligibles served. SETTING: Acute care hospitals within the United States. PARTICIPANTS: Medicare fee-for-service beneficiaries aged 65 years or older discharged alive from January 1, 2003 to November 30, 2014 with a principal discharge diagnosis of acute myocardial infarction (AMI), congestive heart failure (CHF), and pneumonia. MAIN OUTCOME AND MEASURE: Decrease in hospital-level RSRRs in the post-law period, after controlling for the pre-law trend. RESULTS: For AMI, the pre-post difference between hospitals that service high and low proportion of dual-eligibles was not significant (−65 vs. −64 risk-standardized readmissions per 10000 discharges per year, p = 0.0678). For CHF, RSRRs declined more at high than low dual-eligible hospitals (−79 vs. −75 risk-standardized readmissions per 10000 discharges per year, p = 0.0006). For pneumonia, RSRRs declined less at high than low dual-eligible hospitals (−44 vs. −47 risk-standardized readmissions per 10000 discharges per year, p = 0.0003). Among the 742 highest penalized hospitals and all conditions, the pre-post decline in rate of change of RSRRs was less for high dual-eligible hospitals than low dual-eligible hospitals (−68 vs. −74 risk-standardized readmissions per 10000 discharges per year for AMI, −88 vs. −97 for CHF, and −47 vs. −56 for pneumonia, p < 0.0001 for all). CONCLUSIONS AND RELEVANCE: For all hospitals, differences in pre-post trends in RSRRs varied with disease conditions. However, for the highest-penalized hospitals, the pre-post decline in RSRRs was greater for low than high dual-eligible hospitals for all penalized conditions. These results suggest that high penalty, high dual-eligible hospitals may be less able to improve performance on readmission metrics.