Association Between Proteinuria and Pregnancy Outcomes in Women With Preterm Pre‐Eclampsia Undergoing Expectant Management: Secondary Analysis of Two Randomised Trials

Abstract

Objective Evaluate whether proteinuria levels, in women undergoing expectant management for pre‐eclampsia, are associated with pregnancy latency and adverse outcomes. Design Secondary analysis of data from two randomised trials. Setting Tygerberg Hospital, Cape Town, South Africa. Population Women undergoing expectant management for pre‐eclampsia diagnosed between 26 and 32 weeks. Methods Latency was analysed using parametric time‐to‐event methods to estimate increases. Composite outcomes were assessed using Poisson regression. Main Outcome Measures Pregnancy latency Composite adverse maternal or perinatal outcomes. Results Two hundred and ninety‐one women were included. Women with < 3 g/24 h had a median prolongation of 15.3 days (interquartile range (IQR) 4.3 to 25.6), compared to 10.0 days (IQR 5.0 to 19.6) with proteinuria between 3 to 5 g/24 h; and 4.9 days (IQR 3.8 to 11.7) for those with ≥ 5 g/24 h. Increasing levels were also associated with increases in subsequent adverse composite maternal outcomes: 8.8% of women with < 3 g/24 h experienced maternal complications, compared to 16.1% with 3 to 5 g/24 h, and 27.9% with ≥ 5 g/24 h. There was no association with composite perinatal adverse outcomes. The sFlt‐1/PIGF ratio performed slightly better in predicting latency (area under the receiver operating characteristic curve (AUC) 0.78 vs. 0.67) but similarly in predicting adverse composite maternal outcomes (AUC 0.67 vs. 0.64). Both had poor performance for predicting composite adverse perinatal outcomes (AUC 0.61 vs. 0.56). Conclusion Levels of proteinuria strongly correlate with pregnancy latency and adverse maternal outcomes, but not perinatal outcomes, in women with pre‐eclampsia undergoing expectant management. They perform similarly to angiogenic markers. Proteinuria levels may confer useful prognostic information to assist clinical decision‐making.