OBJECTIVES: Quality metrics for episode-based bundled payment models have traditionally been tied to the 30-day period post event or procedure. The 30-day model has been criticized for being too short to accurately reflect patient outcomes and institutional performance. Acute myocardial infarction (AMI) is one episode of care included in the Centers for Medicare & Medicaid Services Bundled Payments for Care Improvement Advanced model, which ties payments and penalties to a 90-day period. This review aimed to understand if and why 90 days may be a critical timepoint for AMI ‘episodes of care’.

METHODS: A targeted literature review was conducted using MEDLINE to identify publications from real-world databases, healthcare claims analyses and clinical trials following AMI from 2011–2021 that included 90-day outcomes assessments.

RESULTS: Following AMI, cardiovascular (CV) events and hospital readmissions occurred frequently beyond the traditional 30-day period. High rates of readmissions were reported at 90 days (18% to 38.5%), with a substantial proportion occurring after the first 30 days (37% to 45%). Between 2.1% and 21.8% of 90-day readmissions were attributed to recurrent AMI. The 90-day timepoint may be particularly relevant for patients at a high risk of recurrent CV events; rehospitalization rates are reportedly substantially higher in high-risk versus low-risk patients (43% vs. 13%). Older age, female sex, low income and comorbidities e.g., diabetes mellitus were consistently associated with higher 90-day CV event and rehospitalization rates.

CONCLUSIONS: Significant clinical and economic burden following AMI occurs beyond 30 days. Although 90-day rehospitalization rates have decreased over time the change is not substantial indicating that novel interventions are needed to address risk. The risk period for recurrent CV events following AMI should be redefined and expanded to 90 days. Additionally, clinical trials for new AMI interventions should include 90-day endpoints to better understand treatment benefit.