The utility of routine postoperative upper GI series following laparoscopic gastric bypass

Abstract

Routine upper gastrointestinal (UGI) studies following laparoscopic Roux-en-Y gastric bypass (LRYGBP) have the potential advantage of early identification of anastomotic complications. The aim of our study was to evaluate the efficacy of routine postoperative UGI and its relationship to clinical outcomes. Over a three-year period, 516 patients underwent LRYGBP followed by routine postoperative UGI studies. Data were collected on the results of the UGI, clinical parameters, and patient outcomes. Study groups were composed of patients with a normal UGI (Group I, n = 455), abnormal UGI not requiring further intervention (Group II, n = 36), and abnormal UGI requiring further intervention (Group III, n =25). Statistical significance was set at alpha= 0.05 level for all analyses. The three study groups were not statistically different in mean age (42 years) or body mass index (BMI) (45) and were predominantly female (90%). Most patients had an uneventful postoperative course. Anastomotic complications (gastrojejunostomy and jejunojejunostomy) were uncommon (1.3%). The sensitivity of the UGI for anastomotic leak in this study was low (33%). However, all patients with alimentary limb obstruction (n = 3) had UGI evidence of this complication. Of the 516 UGI reports, there were only 25 (4.8%, Group III) that were abnormal and required some form of intervention ranging from serial imaging (84%) to reoperation (16%). Of the various clinical parameters examined, the patients in Group III demonstrated a significantly higher prevalence of fever (p < 0.001), tachycardia (p < 0.01), vomiting (p < 0.001), and postoperative day 1 leukocytosis (p < 0.005). Our data suggest that routine UGI after LRYGBP has limited utility as it may result in unnecessary intervention based on false-positive results or a delay in treatment based on false-negative results. We advocate selective UGI imaging following LRYGBP based on the patient's clinical factors, particularly fever and tachycardia.