Background: Anastomotic leaks after LRYGB are uncommon and an independent risk factor for mortality.
Methods: Retrospective analysis of prospectively collected database from 1/1/01-11/1/07. All patients underwent linear-stapled GJ and JJ anastomosis.
Results: 1792 patients underwent LRYGB. 23 (1.3%) developed leaks. Demographics included age 44±9, female(75%), and BMI 51±8. There were 16 GJ leaks (0.9%). The incidence was evenly distributed over time. Air leak on intra-op testing was a risk factor for postoperative leak, occurring in 3/16 (23%). Routine postoperative UGI was negative in 7/16 (44%) patients with GJ leak. 11/16 (69%) were symptomatic. 9/16 (56%) were managed operatively on average post-op day 2±2. 7/16 (44%) were managed with drainage alone. There were no deaths. Average hospital stay was 8±7 days.There were 5 JJ leaks; all during the first half of our series. None were diagnosed by UGI. All patients were symptomatic and were managed operatively on post-op day 3±3. There were 2 deaths (40%), and hospital stay was 16±9 days.Sensitivity of UGI is higher for GJ than JJ leaks (56% vs 0%, p:0.04). Both are similarly symptomatic. Operative management was employed similarly. JJ leaks have a longer LOS (16±9 vs 8±7 days, p:0.02) and a higher mortality (40% vs 0%, p:0.04).
Conclusions: Leak is an uncommon complication of LRYGB (1.3%). The GJ leak presents minimal morbidity / mortality with early detection / treatment in the modern era of LRYGB. Routine UGI is questioned due to its poor sensitivity for leak detection. JJ leaks cause greater morbidity and mortality than GJ leaks. Quality improvement efforts should not overlook the importance of detection and prevention of JJ leaks.