Background and Aims: The severity of liver fibrosis is closely related to the incidence of complications after hepatectomy. Thus, a comprehensive and accurate preoperative assessment of the patient’s liver fibrosis is of great importance for surgical procedure selection and prognosis of patients. Thus, this study was aimed to explore whether certain indicators or a prediction model constructed from these indicators can accurately and comprehensively predict the severity of liver fibrosis in patients. 
Methods: The clinical data of 106 patients who underwent hepatectomy from September 2018 to December 2019 were collected for retrospective analysis, and those patients were divided into none/low-stage fibrosis group (50 patients) and high-stage fibrosis group (56 patients) based on the histological classification of liver fibrosis (Laennec staging system). Firstly, all the tested indexes of the two groups of patients were assessed by univariate analysis, and then those with significant differences were included in the multivariate regression analysis to screen out the independent prediction indicators and establish an integrated prediction model. Finally, the receiver operating characteristic (ROC) curve was established to evaluate the predictive efficacy of the independent predictive indicators and the integrated prediction model.
Results: Results of univariate analysis showed that there were significant differences between the two groups in terms of white blood cell (WBC), platelet (PLT), prothrombin time (PT), creatine (Cr), indocyanine green retention rate at 15 minutes (ICG15), width of portal vein and velocity of portal blood flow (all P<0.05). Results of multivariate regression analysis revealed that ICG15 and the width of portal vein were independent predictors for high-stage liver fibrosis (both P<0.05), and the established integrated predictive model based on the two variables was Logit (P)=–6.026+0.44×ICG15+0.299×width of portal vein. Moreover, the area under ROC curve (AUC) of the integrated predictive model was 0.88, with a sensitivity of 89.3% and a specificity of 74% at the cut-off value of 0.359. The predictive efficacy of the integrated prediction model was superior to either prediction indicator alone.
Conclusion: ICG15 and the width of portal vein are independent evaluation indicators for the severity of liver fibrosis, and the integrated prediction model combined with ICG15 and portal vein width offers more accurate preoperative assessment of the severity of liver fibrosis in patients, which has certain clinical reference value.