Development of a novel intraoperative difficult score for minimally invasive cholecystectomy

Abstract

Presenter: Brittany Calkins | Washington University, St. Louis Background: The rate of biliary injuries from minimally invasive cholecystectomy has remained high and static for over two decades. The Critical View of Safety has been recommended to reduce the risk of biliary injuries however in the presence of severe inflammation or hostile anatomy it may be impossible to obtain the Critical View of Safety. In these situation there are multiple “bail-out” methods described including aborting the procedure, converting to open, or performing a sub-total cholecystectomy. Currently there is a gap in the cholecystectomy literature for characterizing the intraoperative difficulty threshold for when a bail-out method should be implemented. The goal of this study was to test the ability of a novel intraoperative difficulty score to predict early in the operation whether a bail-out technique would be required. Methods: From January 2014 to February 2019 cholecystectomy videos were collected and de-identified. A five point scale was developed to rate the videos (see table 1). An online platform was developed to review and score the videos. Ten reviewers were recruited to test the scale including two faculty hepatobiliary surgeons (experts) and eight non-clinical participants (non-experts). The videos were accelerated and edited to include the first 2 – 3 minutes of the operation during which the gallbladder was visualized prior to analysis. Inter-coder reliability was evaluated using Krippendorff’s alpha and regression models were used to evaluate the difficulty scores ability to predict the need to implement a bail-out technique. All statistics were performed using SPSS (v25). Results: 62 cholecystectomy videos were collected for analysis. The median length of the final videos for review was 37.5 (IQR 29-43) seconds with a median time of 46.2 (IQR 38-53) seconds per reviewer to grade the videos. The bail-out rate in our sample was 42.9%, the majority of which were converted to subtotal cholecystectomies. The inter-coder reliability between experts and non-experts was 0.675 with an average difficulty score of 3.0 (SD=1.01). Regression models showed that the scale was able to significantly predict bail-out b=0.29, ß=0.56, p ≤.01. Conclusion: This novel difficulty scale tested herein was able to predict the need to implement a bail out technique during minimally invasive cholecystectomy. The ability to predict early in the operation the need for conversion provides surgeons heightened awareness of a difficult gallbladder.