Objective: The objective of this study was to identify clinical leak in diverted colorectal anastomoses.
Design: Cohort analysis.
Setting: The study was conducted in a subspecialty practice at a tertiary care facility.
Patients: Consecutive subjects undergoing colorectal anastomosis and proximal fecal diversion between July 16, 2007 and June, 31 2012.
Interventions: No intervention was applied.
Main outcome measures: Clinical anastomotic leak.
Results: Two hundred forty-five patients underwent a colorectal anastomosis with proximal fecal diversion. A total of 34 (14%) clinical leaks were identified at a median of 43 days. Clinical leaks were identified in 13 (5%) patients within 30 days of surgery (early leaks) and in 21 (9%) patients after 30 days of surgery (late leaks). Age, sex, use of neoadjuvant chemoradiotherapy, and method of anastomotic construction were similar in patients with clinical leaks as compared to those with no evidence of leak. However, clinical leaks were more likely to develop in patients with a diagnosis of inflammatory bowel disease or other diagnoses, i.e., radiation enteritis, ischemia, and injury/enterotomy. Patients with clinical leak were not more likely to have air leaks on intraoperative air leak testing.
Conclusions: In diverted anastomoses, most leaks become clinically apparent beyond 30 days. The standard practice of censoring outcomes that occur beyond postoperative day 30 will fail to identify a substantial fraction of leaks in diverted colorectal anastomoses.