BackgroundAnastomotic leak following colorectal resection is associated with morbidity, mortality, and poor bowel function. Minimal data exists on the relationship between anastomotic technique, intraoperative leak test, and subsequent clinical leak, particularly on the utility of performing end-to-end versus non-end-to-end anastomoses to avoid postoperative leaks. This study’s aim was to analyze potential associations between anastomotic construction, intraoperative anastomotic assessments, and clinical leak. MethodsWe conducted a retrospective cohort study of patients at a tertiary care center with colorectal cancer, who underwent left-sided colorectal resections with colorectal or coloanal anastomoses, comparing anastomotic technique. Outcomes were rates of intraoperative air leak, incomplete anastomotic donuts, and postoperative clinical leak. Univariate and multivariate analyses were performed to evaluate potential association between anastomotic technique and intraoperative anastomotic assessments and subsequent leak. Results/OutcomesAmong 844 patients, 27 (3.2%) had intraoperative leak, 6 (0.7%) had incomplete donuts, and 27 (3.2%) experienced clinical leak. 500 (59.2%) patients had end-to-end anastomoses and 344 (40.7%) had non-end-to-end. There were no significant differences in demographics or comorbidities between groups (p>0.05) or rate of incomplete donuts (p=0.07). End-to-end anastomosis was associated with significantly more intraoperative air leaks than non-end-to-end on univariate (4.9% vs 1.2%, p=0.005) and multivariate analysis (OR 3.6; 95% CI 1.01-12.5, p= 0.049). There was no difference in postoperative clinical leak between groups on univariate (3.0% vs 3.5%, p=0.69) or multivariate analysis (OR: 0.97; 95% CI: 0.40-2.34; p=0.94). ConclusionEnd-to-end anastomosis is associated with higher rates of intraoperative air leak than non-end-to-end anastomosis, even after adjusting for potential confounders.
Read full abstract