Mirjam Kaijser

CI > 3.5 were again marked as key steps. Items with a mean > 3.5 were marked ‘advisable’. These criteria are summarized in Table 1. In line with earlier Delphi key step identification, it was hypothesized that two rounds would be sufficient for consensus.8,9 Table 1: Selection Process Based on the Limits of the 95% Confidence Interval and Means First round Second round Key step Lower limit CI > 4 Lower limit CI > 3.5 Advisable n.a. CI < 3.5; mean > 3.5 Re-evalua2on in second round Lower and upper limit CI 3-4 n.a Excluded / non relevant Upper limit CI < 3 CI < 3.5; mean < 3.5 CI = confidence interval Statistical Analysis Analysis was performed by SAS statistical software version 9.2. Consensus, or internal consistency, between experts was defined as a Cronbach’s alpha of at least 80% for each procedure. The responses of each sub-step were evaluated as continuous outcomes. Next, the correlations between the answers of the individual respondents were calculated for both procedures, as well as the overall correlations between all respondents, the Cronbach coefficient alpha. This analysis was repeated after the second round. RESULTS The survey of lead surgeons and website search resulted in the collection of contact details of 68 surgeons performing metabolic bariatric procedures, in 20 Dutch centres. A total of 38 surgeons participated (response rate 56%), representing 18 of the 20 clinics (90%). Roux-en-Y Gastric Bypass The LRYGB procedure was divided into nine surgical steps: operative set up, starting laparoscopy, creating the pouch, creating the biliopancreatic limb, performing gastrojejunostomy, creating the alimentary limb, performing jejunojejunostomy, check of the bypass, and finishing the procedure. Next, the surgical steps were divided into 73 sub-steps. A complete list of these is represented in the first column of Table 2.14–18 4 65 The crucial steps in gastric bypass and sleeve gastrectomy procedures

RkJQdWJsaXNoZXIy MTk4NDMw