Routine upper gastrointestinal imaging is superior to clinical signs for detecting gastrojejunal leak after laparoscopic roux-en-y gastric bypass

Background: There are myriad symptoms and signs of gastrojejunal leak; prompt recognition is essential. Many surgeons use clinical predictors to guide selective use of upper gastrointestinal imaging (UGI). The appropriate practice remains undefined. Study Design: A review of patients who underwent primary laparoscopic Roux-en-Y gastric bypass between January 2002 and December 2008 was conducted. All underwent routine UGI studies on postoperative day 1. Actual gastrojejunal leak within 7 days of surgery (actual leak [AL], radiologic leaks), operative reports, patient charts, and postoperative vital signs were retrospectively reviewed. Results: There were 2,099 operations. Eight ALs (0.43%) occurred without associated mortality. UGI was positive in 7 AL patients and falsely positive in 6 patients. The AL patients underwent laparoscopy on postoperative days 1 and 3 (n = 5 and n = 1, respectively), laparotomy on postoperative day 3 (n = 1), and peritoneal drainage (n = 1). False-positive UGIs prompted laparoscopy (n = 3) and close observation (n = 3). Pulse was 100 to 120 beats per minute in 2 patients and fever (>38.5°C) was present in 0 AL patients. AL patients had osteogenesis imperfecta (n = 1), macronodular cirrhosis (n = 1), positive bubble test (n = 3), and concomitant splenectomy (n = 1). No jejunojejunostomy leaks were identified. Conclusions: Routine UGI after laparoscopic Roux-en-Y gastric bypass has greater sensitivity than clinical signs for detecting gastrojejunal leak. Delay in the diagnosis of leakage can impact mortality, and this suggests that indications for routine UGI might still exist. Tachycardia is not a reliable early marker of leak. There might be risk factors for leak in addition to vital signs, including patient medical history or intraoperative events, which should prompt routine UGI on postoperative day 1.