Hospital variation in 'days alive and out of hospital' after colorectal cancer surgery: a national retrospective cohort study.

Abstract

Postoperative mortality varies between hospitals, largely driven by differences in failure to rescue rather than complication rates. Although failure to rescue focuses on mortality, it does not capture broader patient-centred outcomes. Days alive and out of hospital (DAOH) reflects survival, recovery, and hospital resource use. We hypothesised that hospital-level variation in DAOH within 90 days of surgery (DAOH₉₀) would be greater among patients with postoperative complications, reflecting 'capacity to rescue'. A retrospective cohort study was conducted using linked, population-based national registry data for patients undergoing colorectal cancer resection in Aotearoa New Zealand (2005-2020). Risk-adjusted DAOH₉₀ values were calculated using direct standardisation, stratified by the occurrence of complications. Hospital-level variation was analysed across percentiles of DAOH₉₀, and correlations with mortality and failure to rescue were assessed. Among 23 948 patients, 44.5% experienced postoperative complications. DAOH₉₀ was lower for patients with complications (median, 72 vs 82 days; P<0.001). Hospital variation in DAOH₉₀ was most pronounced at lower percentiles (25th percentile range, 67-75 days; 10th percentile range, 39-62 days), driven by variation amongst patients with complications. Minimal variation was observed among patients without complications. DAOH₉₀ correlated strongly with failure to rescue (r=-0.82) and mortality (r=-0.87), supporting its potential as a quality indicator. Hospital variation in DAOH₉₀ among patients with complications reflects differences in capacity to rescue. DAOH₉₀ extends the concept of failure to rescue to include nonfatal outcomes, providing a patient-centred quality measure that complements traditional indicators of surgical performance, and could aid in national benchmarking and quality improvement.