Development and Internal Validation of a Risk Prediction Model for Acute Cardiovascular Morbidity in Preeclampsia

Abstract

Background

Women with preeclampsia are at increased short-term risk of adverse cardiovascular outcomes during pregnancy and the early postpartum period. We aimed to develop and internally validate a risk assessment tool to predict acute cardiovascular morbidity in preeclampsia.

Methods

The study was conducted at an academic obstetrics hospital. Participants with preeclampsia at delivery from 2007 to 2017 were included. A model to predict acute cardiovascular morbidity at delivery and within 6 weeks after delivery was developed and evaluated. The primary composite outcome included pulmonary edema/acute heart failure, myocardial infarction, aneurysm, cardiac arrest/ventricular fibrillation, heart failure/arrest during surgery or procedure, cerebrovascular disorders, cardiogenic shock, conversion of cardiac rhythm, and difficult-to-control severe hypertension. We assessed model discrimination and calibration. We used bootstrapping for internal validation.

Results

A total of 4171 participants with preeclampsia were included. The final model comprised 8 variables. Predictors positively associated with acute cardiovascular morbidity (presented as odds ratio with 95% confidence interval) were: gestational age at delivery (20-36 weeks: 5.36 [3.67-7.82]; 37-38 weeks: 1.75 [1.16-2.64]), maternal age (≥ 40 years: 1.65 [1.00-2.72]; 35-39 years: 1.49 [1.07-2.09]), and previous caesarean delivery (1.47 [1.01-2.13]). The model had an area under the receiver operating characteristic curve of 0.72 (95% CI 0.69-0.74). Moreover, it was adequately calibrated and performed well on internal validation.

Conclusions

This risk prediction tool identified women with preeclampsia at highest risk of acute cardiovascular morbidity. If externally validated, this tool may facilitate early interventions aimed at preventing adverse cardiovascular outcomes in pregnancy and postpartum.