The Effect of Hospital Safety Net Status on the Association Between Bundled Payment Participation and Changes in Medical Episode Outcomes

Abstract

BACKGROUND: Under Medicare’s Bundled Payments for Care Improvement (BPCI) program, hospitals have maintained quality and achieved savings for medical conditions. However, safety net hospitals may perform differently owing to financial constraints and organizational challenges. OBJECTIVE: To evaluate whether hospital safety net status affected the association between bundled payment participation and medical episode outcomes. DESIGN, SETTING, AND PARTICIPANTS: This observational difference-in-differences analysis was conducted in safety net and non–safety net hospitals participating in BPCI for medical episodes (BPCI hospitals) using data from 2011-2016 Medicare fee-for-service beneficiaries hospitalized for acute myocardial infarction, pneumonia, congestive heart failure, and chronic obstructive pulmonary disease. EXPOSURE(S): Hospital BPCI participation and safety net status. MAIN OUTCOME(S) AND MEASURE(S): The primary outcome was postdischarge spending. Secondary outcomes included quality and post–acute care utilization measures. RESULTS: Our sample consisted of 803 safety net and 2263 non–safety net hospitals. Safety net hospitals were larger and located in areas with more low-income individuals than non–safety net hospitals. Among BPCI hospitals, safety net status was not associated with differential postdischarge spending (adjusted difference-in-differences [aDID], $40; 95% CI, –$254 to $335; P =.79) or quality (mortality, readmissions). However, BPCI safety net hospitals had differentially greater discharge to institutional post–acute care (aDID, 1.06 percentage points; 95% CI, 0.37-1.76; P =.003) and lower discharge home with home health (aDID, –1.15 percentage points; 95% CI, –1.73 to –0.58; P