Assurance of surgical quality within multicenter randomized controlled trials for bariatric and metabolic surgery: a systematic review.

Abstract

BACKGROUND\nSurgical quality assurance methods aim to ensure standardization and high quality of surgical techniques within multicenter randomized controlled trials (RCTs), thereby diminishing the heterogeneity of surgery and reducing biases due to surgical variation. This study aimed to establish the measures undertaken to ensure surgical quality within multicenter RCTs investigating bariatric and metabolic surgery, and their influence upon clinical outcomes.\n\n\nMETHODS\nAn electronic literature search was performed from the Embase, Medline, and Web of Science databases to identify multicenter RCTs investigating bariatric and metabolic surgery. Each RCT was evaluated against a checklist of surgical quality measures within 3 domain: (1) standardization of surgical techniques; (2) credentialing of surgical experience; and (3) monitoring of performance. Outcome measures were postoperative weight change and complications.\n\n\nRESULTS\nNineteen multicenter RCTs were included in the analysis. Three studies undertook pretrial education of surgical standard. Fourteen studies described complete standardization of surgical techniques. Four studies credentialed surgeons by case volume prior to enrollment. Two studies used intraoperative or video evaluation of surgical technique prior to enrollment. Only two studies monitored performance during the study. Although there were limited quality assurance methods undertaken, utilization of these techniques was associated with reduced overall complications. Standardization of surgery was associated with reduced re-operation rates but did not influence postoperative weight loss.\n\n\nCONCLUSION\nThe utilization of methods for surgical quality assurance are very limited within multicenter RCTs of bariatric and metabolic surgery. Future studies must implement surgical quality assurance methods to reduce variability of surgical performance and potential bias within RCTs.