Nomogram predicting surgical risk of laparoscopic left-sided hepatectomy for hepatolithiasis.

Abstract

To explore the surgical risk factors of laparoscopic left-sided hepatectomy for hepatolithiasis and establish and validate a nomogram to estimate the corresponding surgical risks. Patients with hepatolithiasis who underwent laparoscopic left-sided hepatectomy were retrospectively enrolled. Demographic data, clinicopathological parameters, and surgical factors were collected. Three hundred fifty-three patients were enrolled and randomly divided into training set (n=267) and validation set (n=86) by 3:1. Conversion to laparotomy was used as a surrogate index to evaluate the surgical risk. Univariate analysis was used to screen potential surgical risk factors, and multivariate analysis using logistic regression model was used to screen independent surgical risk factors. Nomogram predicting the surgical risks was established based on the independent risk factors. Discrimination, calibration, decision curve, and clinical impact analyses were used to evaluate the performance of the nomogram on the statistical and clinical aspects both in the training and validation sets. Five independent surgical risk factors were identified in the training set, including recurrent abdominal pain, bile duct stricture, ASA classification ≥2, extent of liver resection, and biliary tract T tube drainage. No collinearity was found among these five factors, and a nomogram was established. Performance analyses of the nomogram showed good discrimination (AUC=0.850 and 0.817) and calibration (Hosmer-Lemeshow test, p=0.530 and 0.930) capabilities both in the training and validation sets. Decision curve and clinical impact analyses also showed that the prediction performance was clinically valuable. A nomogram was established and validated to be effective in evaluating and predicting the surgical risk of patients undergoing laparoscopic left-sided hepatectomies for hepatolithiasis.