Construction and Validation of Clinical Prediction Model for Gallstone Columnar Chart

Objective: This study aims to develop a nomogram prediction model to predict the exact probability of gallstones in the general population. Methods: We divided 2282 individuals undergoing physical examinations into a training set (n=1708) and a validation set (n=574). We collected and analyzed their clinical characteristics and liver and gallbladder ultrasound results. Factors for constructing the prediction model were screened through univariate analysis and multivariate logistic regression analysis, and an interactive nomogram prediction model was created based on these factors. The performance of the prediction model was evaluated using the receiver operating characteristic (ROC) curve, calibration curve, decision curve, and clinical impact curve. Results: Multivariate logistic regression analysis indicated that age, HDL, high TBA (TBA ≥ 2.3μmol/L), RDW, and high NE% (NE% ≥ 58.2%) were independent risk factors for gallstones (P<0.05). An interactive nomogram prediction model was constructed based on these risk factors. Statistical evaluation showed that the model had moderate predictive value [area under the curve (AUC)>0.7] and good calibration, as well as good clinical benefit and impact. Applying the model to the validation set showed that the prediction effect of the validation set was similar to that of the training set (AUC=0.712), indicating that the model’s prediction results were relatively stable and had certain clinical application value. Conclusion: This nomogram clinical prediction model has good practical application value and generalizability. It can help screen individuals at risk for gallstones, assist medical workers in formulating personalized examination plans, reduce medical costs, save medical resources, and promote the development of public health.