PRERISK Study: A Randomized Controlled Trial Evaluating a sFlt-1/PlGF-Based Calculator for Preeclampsia Hospitalization.

Abstract

BACKGROUND A model based on the soluble Fms-like tyrosine kinase-1/placental growth factor ratio, gestational age, and the urinary protein-to-creatinine ratio (PRERISK calculator) has been developed to predict preeclampsia-related maternal-fetal complications. Here, we tested whether this model can reduce hospital admissions without increasing complication rates among women with suspected or confirmed preeclampsia. METHODS In this multicenter, open-label, randomized controlled trial conducted at 5 Dutch medical centers, women with suspected or confirmed preeclampsia were randomly assigned to the intervention group, where admission was guided by the PRERISK score using a 5% cutoff, or to the control group receiving routine care with a concealed PRERISK score. Two co-primary outcomes were the incidence of maternal-fetal preeclampsia-related complications (noninferiority) and the proportion of women with a hospitalization ratio (=admission days/inclusion days) ≤0.05 (superiority). RESULTS The intervention and control groups included 442 and 435 women, respectively. In the intention-to-treat analysis, complications occurred in 41.6% of the intervention group versus 39.5% of the control group (adjusted relative risk 1.06 [95% CI, 0.92-1.22]; P=0.43). The proportion of women achieving a hospitalization ratio ≤0.05 was 23.6% in the intervention group and 26.3% in the control group (adjusted relative risk, 0.89 [95% CI, 0.71-1.13]; P=0.34). The latter was comparable in the per-protocol analysis (adjusted relative risk, 0.87 [95% CI, 0.64-1.19]; P=0.38), while in this analysis, complications occurred in 47.8% of the intervention group (n=251) versus 41.7% of the control group (n=365; adjusted relative risk 1.19 [95% CI, 1.03-1.38]; P=0.02). CONCLUSIONS Routine screening with the PRERISK score and a 5% cutoff in patients with suspected or confirmed preeclampsia does not decrease hospitalization and is therefore not recommended. REGISTRATION URL: https://onderzoekmetmensen.nl/nl/trial/48687; Unique identifier: NL63386.078.17, NL-OMON48687.