Assembling a library of typical surgery video clips to construct a system for assessing the surgical difficulty of laparoscopic cholecystectomy

Abstract

Background To explore best practices for acute cholecystitis, it is necessary to construct a system to assess the difficulty of laparoscopic cholecystectomy (LC) based on intraoperative findings. In this study, multiple evaluators assessed videos of LC to assemble a library of typical video clips for 25 intraoperative findings. Methods We have previously identified 25 items that contribute to surgical difficulty in LC. For each item, roughly 30‐second video clips were submitted from videos of LC performed at member institutions. We then selected one typical video from the collected clips based on simple tabulation of the instances of agreement. Inter‐rater agreement was assessed with Fleiss's κ and Gwet's agreement coefficient (AC). Results Except in the case of two assessment items (“edematous change” and “easy bleeding”), κ or AC significantly exceeded 0.5 and the typical videos were judged to be applicable. For the two remaining items, the evaluation was repeated after clarifying the definitions of positive and negative findings. Eventually, they were recognized as typical. The completed video clip library contains 31 clips and is divided into five categories (http://www.jshbps.jp/modules/project/index.php?content_id=13). Conclusions This clip library may be highly useful in clinical settings as a more objective standard for assessing surgical difficulty in LC.

Application of a novel surgical difficulty grading system during laparoscopic cholecystectomy

“…1 “edematous changes,” was categorized as having “Fair” agreement. Regarding the previous study, “edematous changes” also revealed “Fair” agreement in the first round ( κ = 0.223, 90% CI: 0.096‐0.350; AC 1 = 0.314, 90% CI: 0.150‐0.479) 9 . Therefore, we adapted the definition of “edematous change” and added the description to the video clip.…”

“…Regarding the previous study, "edematous changes" also revealed "Fair" agreement in the first round (κ = 0.223, 90% CI: 0.096-0.350; AC 1 = 0.314, 90% CI: 0.150-0.479). 9 Therefore, we adapted the definition of "edematous change" and added the description to the video clip. To further improve agreement, we added absence of "edematous change," in which abundant visceral adipose tissue mimics edema at the time of dissection, to the video clip library.…”

“…Then, the items were scored by using the Delphi process in a multinational collaborative study involving Japan, Korea, and Taiwan (Table 1). 8 Subsequently, evaluation analysis was performed multiple times, and a library of a total of 31 typical video clips of all 25 surgical difficulty items was created 9 . In addition, this typical video clip library can be viewed from the website of the Japanese Society of Hepato‐Biliary‐Pancreatic Surgery (http://www.jshbps.jp/modules/project/index.php?content_id=13).…”

“…First, we assumed that the surgical difficulty grades would be divided into three categories: Grade A (maximum difficulty scores from 0 to 2); Grade B (maximum difficulty scores of 3 and 4); and Grade C (maximum difficulty scores of 5 and 6). In this case, according to our previous study, 9 the 95% confidence intervals (CI) were estimated as κ (no weighting) = 0.60 (95%CI: 0.43‐0.78) and κ (first‐order weighting) = 0.64 (95%CI: 0.47‐0.81). According to this calculation, in the case of 50 evaluators assessing difficulty with the three grades, a statistical power of 80% could be achieved by assessing 11 or more cases, assuming a true κ coefficient of 0.75, a cutoff of 0.5, and a one‐tailed α of 0.05.…”

“…8 In our previous research, we created a typical video clip library regarding all 25 items so that surgical difficulty can be evaluated more objectively. 9 The members submitted typical video clips considered to be consistent with the intraoperative findings and selected one from among the collected clips.…”