Outcomes of laparoscopic gastric bypass in a randomized clinical trial compared with a concurrent national database.

Abstract

RCTs are the standard for assessing medical interventions, but they may not be feasible and their external validity is sometimes questioned. This study aimed to compare results from an RCT on mesenteric defect closure during laparoscopic gastric bypass with those from a national database containing data on the same procedure, to shed light on the external validity of the RCT. Patients undergoing laparoscopic gastric bypass surgery within an RCT conducted between 1 May 2010 and 14 November 2011 were compared with those who underwent the same procedure in Sweden outside the RCT over the same time interval. Primary endpoints were severe complications within 30 days and surgery for small bowel obstruction within 4 years. Some 2507 patients in the RCT were compared with 8485 patients in the non-RCT group. There were no differences in severe complications within 30 days in the group without closure of the mesenteric defect (odds ratio (OR) for RCT versus non-RCT 0·94, 95 per cent c.i. 0·64 to 1·36; P = 0·728) or in the group with closure of the defect (OR 1·34, 0·96 to 1·86; P = 0·087). There were no differences between the RCT and non-RCT cohorts in reoperation rates for small bowel obstruction in the mesenteric defect non-closure (cumulative incidence 10·9 versus 9·4 per cent respectively; hazard ratio (HR) 1·20, 95 per cent c.i. 0·99 to 1·46; P = 0·065) and closure (cumulative incidence 5·7 versus 7·0 per cent; HR 0·82, 0·62 to 1·07; P = 0·137) groups. The relative risk for small bowel obstruction without mesenteric defect closure compared with closure was 1·91 in the RCT group and 1·39 in the non-RCT group. The efficacy of mesenteric defect closure was similar in the RCT and national registry, providing evidence for the external validity of the RCT.