Predicting complications in pre-eclampsia: external validation of the fullPIERS model using the PETRA trial dataset

Abstract

ObjectiveThe internally validated fullPIERS model predicts adverse maternal outcomes in women with pre-eclampsia within 48h after eligibility. Our objective was to assess generalizability of this prediction model. Study designExternal validation study using prospectively collected data from two tertiary care obstetric centers. MethodsThe existing PETRA dataset, a cohort of women (n=216) with severe early-onset pre-eclampsia, eclampsia, HELLP syndrome or hypertension-associated fetal growth restriction was used. The fullPIERS model equation was applied to all women in the dataset using values collected within 48h after inclusion. The performance (ROC area and R-squared) of the model, risk stratification and calibration were assessed from 48h up to a week after inclusion. ResultsOf 216 women in the PETRA trial, 73 (34%) experienced an adverse maternal outcome(s) at any time after inclusion. Adverse maternal outcome was observed in 32 (15%) cases within 48h and 62 (29%) within 7 days after inclusion. The fullPIERS model predicted adverse maternal outcomes within 48h (AUC ROC 0.97, 95% CI: 0.87–0.99) and up to 7 days after inclusion (AUC ROC 0.80, 95% CI: 0.70–0.87). ConclusionsThe fullPIERS model performed well when applied to the PETRA dataset. These results confirm the usability of the fullPIERS prediction model as a ‘rule-in’ test for women admitted with severe pre-eclampsia, eclampsia, HELLP syndrome or hypertension-associated fetal growth restriction. Future research should focus on intervention studies that assess the clinical impact of strategies using the fullPIERS model.