Outcomes for Gastrointestinal Oncologic Surgery Episodes in the Bundled Payments for Care Improvement Advanced Program

Objective: The goal of this study was to understand the association of participation in a bundled payment program with financial and clinical outcomes for patients who underwent gastrointestinal operations for malignancy. Summary Background Data: Multidisciplinary cancer care is complex and expensive. Medicare alternative payment models aiming to reduce costs and improve outcomes have been designed for chemotherapy and radiation therapy, but oncologic surgery has yet to be directly targeted by such attempts at payment reform. Methods: Retrospective cohort study using the 100% Medicare file for claims initiated between January 1, 2017 - September 30, 2019. Surgical episodes at BPCI-A hospitals were compared to matched episodes at non-participant hospitals. The outcomes of interest were differential change in 90-day Medicare payments and clinical outcomes (readmissions, mortality, healthy days at home). Results: Patient characteristics at BPCI-A and non-participant hospitals were similar. BPCI-A participation was not associated with significantly decreased payments for patients who underwent operations for malignancy ($32,687 to $31,929, difference -$758) versus non-participant hospitals ($34,006 to $32,932; difference -$1,075) (difference-in-differences (DID) $316; 95% CI -$2976, $2709). Participants had similar changes in spending for malignant compared to benign conditions (triple difference -$624; 95% CI -$3411, $2164). There was no significant improvement in readmission (DID 3.82%; 95% CI -1.12, 8.76), mortality (-1.22%; -3.60, 1.17), or healthy days at home (0.10 d; -2.18, 2.39) associated with BPCI-A participation. Conclusions: BPCI-A participation was not associated with savings or improvement in quality measures for episodes involving gastrointestinal operations for malignancy. Long-term effects of participation and its impact on oncologic outcomes merit further study.