Laparoscopic Surgery Decreases Anastomotic Leak Rate in Sigmoid Colectomy for Diverticulits
Liliana Bordeianou, MD, Melissa Levack, MD, Patricia Sylla, David Rattner, MD, David Berger, MD
Massachusetts General Hospital, Boston, MA
PURPOSE OF STUDY
Early studies comparing laparoscopic surgery for diverticulitis to open failed to show any advantages to the laparoscopic approach. The purpose of our study was to compare these approaches in a cohort of patients who underwent elective laparoscopic or open sigmoid colectomy by a group of surgeons who have performed ≥ 20 laparoscopic colonic resections prior to the study period.
This was a retrospective review of 249 patients who underwent elective open (N=127) or laparoscopic (N=122) sigmoid colectomy with primary anastomosis for diverticulitis between July 2001 and February 2008. Our primary endpoint was the combined rates of free and contained anastomotic leaks (diffuse peritonitis or localized peri-anastomotic abscess). We used intent to treat analysis to compare groups as far as other differences such as demographic factors, medical comorbidities, surgical indications and ultimate surgical outcomes (Chi square, t-test). A logistic regression model was then fitted to determine predictors of anastomotic leaks while controlling for these differences.
SUMMARY OF RESULTS
Laparoscopic and open patients were similar in age ( p=0.56), sex (p=0.62), history of diagnosed intraabdominal abscess (9.8% vs 11.8%, p= 0.68) and history of IR drainage prior to surgery (4.1% vs 4.7%, p=0.81), but open patients had a higher comorbidity index (1.6 vs. 1.2, p= 0.035). Splenic flexure mobilization was much more frequently performed in the laparoscopic patients (82.8% vs. 26.7%, p<0.0001); however the length of colon resection (19.9 vs. 19.1 cm, p=0.231) was similar. Conversion was required in 22 laparoscopic patients (18%). The postoperative rates of wound infections, early small bowel obstructions, ileus, renal and cardiac complications were similar. However, laparoscopic patients had lower rates of anastomotic leaks (2.5% vs. 8.7%, p=0.036). This finding held true on logistic regression analysis (OR 0.67, 95%CI 0.008-0.567, p= 0.01) even when controlling for age (p=0.9), Charlson Comorbidity Index (p=0.2), splenic flexure takedown (p=0.4), and length of resected bowel (p=0.9).
Anastomotic leaks are less common following laparoscopic sigmoid resection than open sigmoid colectomy. Although splenic flexure mobilization was more commonly performed in laparoscopic patients, other factors seem to account for the observed differences.