Preoperative Prediction of Difficult Laparoscopic Cholecystectomy: Evolution, Validation, and Emerging Machine-Learning Approaches—A Scoping Review

Background and Objectives: Preoperative prediction of difficult laparoscopic cholecystectomy (LC) has expanded from point-based scores to regression, nomogram, machine-learning, and quantitative imaging approaches. We examined whether this diversification was accompanied by stronger validation and clinical translation. Materials and Methods: Following a registered Open Science Framework protocol, we conducted a scoping review using Joanna Briggs Institute methodology and PRISMA-ScR. PubMed/MEDLINE, Scopus, and Web of Science were searched on 8 August 2026 for English-language publications from 1 January 2010 through 31 July 2026. Eligible studies evaluated structured multivariable tools using exclusively preoperative inputs. Screening and data charting were performed independently and in duplicate. Results: Searches identified 893 records; removal of 55 records outside the publication window and 372 duplicates left 466 for screening. Sixty reports were assessed as full text, and 59 were included; 47 (79.7%) were published from 2020 onward. Point-based scores predominated (45/59, 76.3%), followed by nomograms (5/59, 8.5%), regression-based probability models (4/59, 6.8%), machine-learning models (4/59, 6.8%), and one quantitative imaging tool (1/59, 1.7%). Four model-development publications (6.8%) included a geographic external validation component, including one with mixed temporal and geographic validation. Calibration was reported in nine publications (15.3%), decision-curve analysis in seven (11.9%), and SHAP-based explainability in two (3.4%). Two publications evaluated scores during clinical use; no controlled clinical-impact study was identified. Conclusions: Evidence of transportability and clinical impact remains limited. Future studies should prioritize standardized prediction targets, complete model specification, calibration, independent validation, and controlled evaluation of model-guided care.