Based on previous work of Glarner et al. and Kramp et al. it was aimed to create an independence-scaled procedure-based assessment on LRYGB and to assure that this assessment can differentiate between different expertise levels of trainees with the use of video recordings.20,21 A high level of independence may eventually lead to professional entrustment or certification for this procedure. MATERIALS AND METHODS First, a new PBA was designed based on a previously introduced PBA in laparoscopic cholecystectomy (LC).21 Second, LRYGB procedures performed by residents and surgeons on different levels were video recorded. After selecting specific fragments of these procedures, twelve independent expert metabolic bariatric surgeons gave feedback with the PBA, BOSATS and OSATS. Finally, the experts were invited to give their opinions on the use of the different assessment scales by questionnaire. Creation of the Procedure-Based Assessment. In previous research LRYGB was divided into multiple steps and substeps with a hierarchical task analysis. A Delphi consensus analysis among Dutch metabolic bariatric surgeons defined the key steps of the procedure: the steps that are necessary to safely accomplish the procedure. Based on this Delphi consensus the LRYGB was divided in 9 steps and 44 advised or crucial substeps.24 Following Kramp et al. in their research on a PBA in LC, these 9 steps were combined into 6 key steps and 30 substeps to improve usability of the PBA.21 The different items were combined with a 5-point operative independence scale introduced by Glarner et al..20 The complete PBA is shown in Appendix I. Sample Size Calculation As this study aims to validate a new assessment the sample size calculation was based on comparable research in the existing literature. In a study validating a PBA for LC, which also used one video per expertise level, a relevant difference between levels was found using 10 experts as assessors.21 As LRYGB was expected to be of higher complexity level than LC, it was assumed that the mutual differences were larger. No data were available to state this difference in a sample size number. In other research Zevin et al. calculated a minimum inclusion of 8 assessments in a comparable study on LRYGB.23 84 5
RkJQdWJsaXNoZXIy MTk4NDMw