Validation of full PIERS model for prediction of adverse outcomes among women with severe pre‐eclampsia

OBJECTIVE: To evaluate the accuracy of the Pre-eclampsia Integrated Estimate of RiSk (fullPIERS) model for the prediction of complications among patients with severe pre-eclampsia in northeastern Brazil. METHODS: A retrospective study was conducted of data for all patients with severe pre-eclampsia admitted to a center in Recife in 2014. The fullPIERS model was applied to calculate the predicted probability of complications. A receiver operating characteristic curve was constructed to determine the accuracy of the model, and the area under the curve was calculated. Multiple logistic regression analysis was performed. RESULTS: Among 325 included women, 55 (16.9%) had one or more complication before discharge. The area under the curve was 0.72 (P<0.001), determining a cutoff point for fullPIERS probability of 1.7%. Sensitivity was 60.0% and specificity was 65.1%; the positive likelihood ratio was 1.72 and the negative likelihood ratio was 0.61. In multivariate analysis, the only variable that remained significantly associated with complications was a fullPIERS probability of more than 1.7% (odds ratio 5.87, 95% confidence interval 3.16-10.89; P<0.001). CONCLUSION: The fullPIERS model is a useful tool for predicting complications in women with severe pre-eclampsia, with the cutoff point of 1.7%.