Development and validation of a risk prediction model for placental abruption in patients with preeclampsia

Introduction

To develop a validated risk prediction model for placental abruption in preeclamptic patients with singleton pregnancies firstly.

Methods

Data from 1448 preeclamptic patients with singleton pregnancies who delivered between January 2013 and December 2022 were reviewed. Variables, including demographic characteristics, laboratory test results, comorbidities, and aspirin use were collected and analyzed. The preeclamptic patients were divided into a training set and a validation set according to the time of delivery. Logistic regression with a backward stepwise elimination method was used for variable screening and nomogram construction. The area under the receiver operating characteristic curve and calibration curve were used to evaluate its accuracy. Decision curve analysis and clinical impact curves were conducted to assess predictive performance.

Results

Finally, 1448 preeclamptic patients were included. We collected 50 variables for further analysis. Multivariate logistic regression analysis revealed that severity, subtype, premature rupture of membranes, urinary casts, diastolic blood pressure, aspartate aminotransferase, serum potassium, and fibrin degradation product levels were predictors of placental abruption. These factors were used to construct the nomogram model, which showed good concordance and accuracy. The area under the receiver operating characteristic curve values of the training set and the validation set were 0.767 (95 % CI = 0.728–0.806, P < 0.001) and 0.800 (95 % CI = 0.728–0.872, P < 0.001). Calibration curves revealed significant agreement between the nomogram model and actual observations. Receiver operating characteristic curve analysis and decision curve analysis indicated that the nomogram had good predictive performance.

Discussion

The prediction model can accurately estimate the risk of placental abruption in preeclamptic patients.