Abstract
OBJECTIVE: Failure to rescue, defined as death after a postoperative complication, is a key metric for evaluating hospital performance and surgical quality. We hypothesized that both patient characteristics and hospital performance are associated with failure to rescue risk after esophagectomy. METHODS: We analyzed esophagectomy cases from the Society of Thoracic Surgeons General Thoracic Surgery Database (2012-2023), excluding emergency surgeries and colonic interpositions. The primary outcome was failure to rescue, defined as death within 30 days or during the index hospitalization after a complication. Hospitals were grouped into quintiles based on their risk-adjusted standardized mortality ratio using hierarchical logistic regression. We compared failure to rescue rates between hospitals in the top (very low standardized mortality ratio) and bottom (very high standardized mortality ratio) quintiles using generalized estimating equations to account for clustering. Failure to rescue rates were further stratified across patient risk categories using established Society of Thoracic Surgeons risk metrics. RESULTS: A total of 28,626 patients from 318 hospitals were included. Top-tier hospitals performed more procedures than bottom-tier hospitals (10,431 vs 3527, P < .001) and had significantly lower unadjusted mortality (1.2% vs 6.6%, P < .001). Adjusted odds of failure to rescue were 4.4 times higher in bottom-tier hospitals (95% CI, 3.6-5.3), with similar trends after major complications (odds ratio, 3.83) and esophagectomy-specific complications (odds ratio, 7.27). Across all patient risk strata, bottom-tier hospitals had higher failure to rescue rates; high-risk patients had an odds ratio of 5.94 (95% CI, 3.6-9.6). CONCLUSIONS: Hospital performance is strongly associated with failure to rescue after esophagectomy, even after adjusting for patient risk. System-level interventions are needed to reduce variation in rescue capacity.