This study shows a good inter-rater reliability in assessment of the CVS in LC videos between surgeons, considered as the gold standard, and surgical residents. This suggests that surgical residents are proficient in the competency assessment of CVS during an LC. At a vast majority (71% of cases), there was concordance between surgeons and surgical residents in assigning either a score of 0 to 4 or a score of 5 or 6, which is clinically relevant, as the literature defines a satisfactory CVS when the score is 5 or 6 [17]. Within this subgroup (29% of cases) where there was no agreement, surgical residents significantly more frequently assigned a score of 0–4 (63%), suggesting that they appear to score more cautiously than that of experienced surgeons. Fortunately, in the minority (37%) of cases, surgical residents assessed a satisfactory CVS, while surgeons disagreed. These could be the cases in practice where a BDI is likely to occur, highlighting the importance of minimizing such occurrences as much as possible.

Unedited LC videos were utilized in this study for CVS assessment. Existing literature discusses CVS assessment based on intraoperative photographs, of both the anterior and posterior views, offering a cost-effective and easily storable alternative. However, the literature indicates that video recording is still the superior method for accurate CVS assessment [19].

Employing SQA tools to assess the quality of surgical procedures and provide constructive feedback is crucial for enhancing postoperative results. Extensive research has consistently demonstrated the significant influence of surgical quality on clinical outcomes [2, 3, 20,21,22,23]. Notably, a national training program called Lapco, implemented in England for specialist colorectal surgeons, successfully employed competency-based supervised clinical training, resulting in reduced mortality and morbidity rates [24]. Applying CVS scoring of LC videos on a larger scale and thus creating valuable feedback opportunities may enable surgeons and especially surgical residents to make faster progress in performing the procedure. This, in turn, may mitigate BDI and positively impact patient outcomes following LC.

One limitation of this study is the absence of comprehensive explanation or training session on how to exactly use the assessment tool and when to assign specific scores. This tool does not pose a significant challenge in terms of usability, but literature suggests that pre-assessment training for the raters could contribute to more consistent results in video-based quality assessment [25]. Despite the tool’s simplicity, providing such training could potentially elevate the ICC scores by fostering consensus among surgeons and surgical residents when assigning scores to specific criteria from the same video.

Scoring videos using assessment tools is inherently time-consuming. As artificial intelligence (AI) gains prominence in the medical field, there is potential for its future integration to facilitate the expeditious determination of the moment of CVS within the entire, unedited LC videos [26]. This would allow reviewers to assess CVS faster. Next, an ideal scenario involves the development of a computer model capable not only of pinpointing the moment of CVS, but also of discerning whether the CVS is deemed “good” and therefore indicating whether the operator can start with ligation of the cystic artery and duct. The literature speaks of a “good” CVS if the score of the 6-point assessment tool is 5 or 6. Training such a model necessitates exposing it to numerous frames of LC videos depicting CVS instances with varying scores. This AI model could eventually be applied intraoperatively, providing a “green” or “red” light post-CVS. In this example, a “green” light would empower the surgical resident to complete the surgery independently, while a “red” light would prompt the surgeon's involvement for guidance. Before such applications become reality, this study establishes that surgical residents can effectively self-assess their surgeries retrospectively and video-based, an essential step in maintaining a steep learning curve.