Development and Validation of a Prenatal Doppler Hemodynamic-Clinical Risk Model for Predicting Adverse Perinatal...

Background/Objectives: Umbilical artery thrombosis (UAT) is a rare but potentially catastrophic obstetric complication associated with a high risk of adverse perinatal outcomes. However, clinically applicable risk stratification tools for pregnancies complicated by UAT remain limited. Therefore, this study aimed to develop and validate a multivariable clinical prediction model integrating maternal clinical characteristics, fetal characteristics, and Doppler-derived hemodynamic parameters for fetuses with UAT. The model may facilitate early risk stratification and support individualized prenatal management and clinical decision-making. Methods: This retrospective study included 114 fetuses with UAT at the Women and Children's Hospital of Chongqing Medical University between January 2017 and January 2025. Candidate predictors were predefined based on previous literature, clinical knowledge, and their potential associations with adverse pregnancy outcomes in UAT, encompassing maternal characteristics, UAT-related factors, Doppler-derived hemodynamic parameters, and relevant clinical factors. Predictors were selected using LASSO with 10-fold cross-validation and entered into multivariable logistic regression to construct a nomogram. Model discrimination, calibration, and clinical utility were assessed using AUC, calibration plots, and decision curve analysis, respectively. The optimal cutoff was determined by the maximum Youden index. Bootstrap resampling was used for internal validation and generalizability assessment. Results: Of the 114 fetuses, 57 (50.0%) experienced adverse pregnancy outcomes. The final nomogram incorporated maternal age, umbilical artery pulsatility index (UA-PI), ductus venosus pulsatility index (DV-PIV), uterine artery pulsatility index (UtA-PI), gravidity, diabetes mellitus, fetal growth restriction (FGR), and UAT laterality. Higher UA-PI (OR, 7.15; 95% CI, 1.01-50.57), higher DV-PIV (OR, 13.18; 95% CI, 2.34-74.41), and FGR (OR, 7.53; 95% CI, 1.77-31.97) were associated with increased odds of adverse outcomes, while UAT laterality showed a borderline association (OR, 2.61; 95% CI, 1.00-6.81). The nomogram achieved an AUC of 0.791 (95% CI, 0.706-0.876), with 70.2% sensitivity and 78.9% specificity at the optimal cutoff. Bootstrap calibration showed a mean absolute error of 0.042. Decision curve analysis demonstrated a positive net benefit across threshold probabilities of 11-92%. Conclusions: This study establishes an internally validated model for predicting adverse perinatal outcomes in fetuses with UAT using clinical characteristics and gestational age-adjusted Doppler hemodynamic parameters available at the time of diagnosis. The model may provide a basis for risk stratification and individualized antenatal surveillance.