Abstract
This study aimed to develop and validate a prediction model for preterm birth in patients with intrahepatic cholestasis of pregnancy. A retrospective analysis was conducted using demographic and clinical data from 370 patients with intrahepatic cholestasis of pregnancy. Multivariate logistic regression was used to screen predictive factors and construct the prediction model. The model's discriminative ability, calibration, and clinical utility were evaluated. Six relevant independent predictors were identified: employment status, placental grading, pharmacotherapy status, lymphocyte count, neutrophil count, and serum total bile acid level. The constructed model demonstrated excellent discriminative ability and calibration in the training set, with an area under the curve of 0.85 (95% CI: 0.78-0.91) and a P value of 0.147 in the Hosmer-Lemeshow test. Decision curve analysis in the training set showed that the model provided clinical net benefit across a risk threshold range of 4% to 45%. This study established the first preterm birth prediction tool for intrahepatic cholestasis of pregnancy integrating multidimensional indicators. It provides an objective basis for early clinical identification of high-risk patients and has significant clinical value.