Bowel injury in gynecologic laparoscopy: a systematic review.

OBJECTIVE: To evaluate the incidence of bowel injury in gynecologic laparoscopy and determine the presentation, mortality, cause, and location of injury within the gastrointestinal tract. DATA SOURCES: The PubMed, EMBASE, ClinicalTrials.gov, and Cochrane Library databases were searched. Additional studies were obtained from references of retrieved papers. METHODS OF STUDY SELECTION: Included retrospective studies and randomized controlled trials reported the incidence of bowel injury in gynecologic laparoscopy. Studies were excluded if they were not in English or duplicated data. TABULATION, INTEGRATION, AND RESULTS: Two reviewers extracted data in duplicate from each study regarding incidence, cause, and location of bowel injury. Ninety studies published between 1972 and 2014 met eligibility criteria, representing 474,063 gynecologic laparoscopies. Six hundred four bowel injuries were reported for an incidence of 1 in 769 (0.13%, 95% confidence interval [CI] 0.12-0.14%). The rate of bowel injury varied by procedure, ranging from 1 in 3,333 (0.03%, 95% CI 0.01-0.03%) for sterilization to 1 in 256 (0.39%, 95% CI 0.34-0.45%) for hysterectomy. The small intestine was the most frequently damaged region of the gastrointestinal tract, representing 166 of 354 (47%) injuries. The majority of bowel injuries occurred during abdominal access and insufflation obtained using a Veress needle or trocar placement (201/366, 55% of injuries). Although most bowel injuries were recognized intraoperatively, diagnosis was delayed by more than 1 day in 154 of 375 cases (41%, 95% CI 36-46%). Bowel injuries were managed primarily by laparotomy (80%). Mortality occurred after bowel injury in 5 of 604, or 1 of 125 (0.8%, 95% CI 0.36-1.9%) cases. All deaths occurred as a result of delayed recognition of bowel injury (n=154), making the mortality rate for unrecognized bowel injury 5 in 154 or 1 in 31 (3.2%, 95% CI 1-7%). There were no deaths associated with intraoperatively diagnosed bowel injury. CONCLUSION: The overall incidence of bowel injury in gynecologic laparoscopy is 1 in 769 but increases with surgical complexity. Delayed diagnosis is associated with a mortality rate of 1 in 31.