Abstract
BACKGROUND: Anastomotic leakage following anterior resection for rectal cancer may result in death. The aim of this study was to yield an updated, population-based estimate of postoperative mortality and evaluate possible interacting factors. METHODS: This was a retrospective national cohort study of patients who underwent anterior resection between 2007 and 2016. Data were retrieved from a prospectively developed database. Anastomotic leakage constituted exposure, whereas outcome was defined as death within 90 days of surgery. Logistic regression analyses, using directed acyclic graphs to evaluate possible confounders, were performed, including interaction analyses. RESULTS: Of 6948 patients, 693 (10·0 per cent) experienced anastomotic leakage and 294 (4·2 per cent) underwent reintervention due to leakage. The mortality rate was 1·5 per cent in patients without leakage and 3·9 per cent in those with leakage. In multivariable analysis, leakage was associated with increased mortality only when a reintervention was performed (odds ratio (OR) 5·57, 95 per cent c.i. 3·29 to 9·44). Leaks not necessitating reintervention did not result in increased mortality (OR 0·70, 0·25 to 1·96). There was evidence of interaction between leakage and age on a multiplicative scale (P = 0·007), leading to a substantial mortality increase in elderly patients with leakage. CONCLUSION: Anastomotic leakage, in particular severe leakage, led to a significant increase in 90-day mortality, with a more pronounced risk of death in the elderly.