0
0
A recent study published on February 4 in PLOS Medicine has found that prediction models designed to assess the risk of severe complications in women with preeclampsia lose their accuracy after the first 48 hours of hospital admission. The research, conducted by Henk Groen from the University of Groningen, the Netherlands, and colleagues, highlights significant limitations in the current models used to monitor women with preeclampsia beyond the initial period of hospitalization.
Preeclampsia is a potentially life-threatening condition that can develop during pregnancy, and it affects approximately 5–20% of women diagnosed with it, leading to severe complications. The PIERS models (Pre-eclampsia Integrated Estimate of RiSk), which include the PIERS Machine Learning (PIERS-ML) model and the logistic-regression-based fullPIERS model, were developed to assess the likelihood of adverse maternal outcomes in the first 48 hours following hospital admission. While both models were originally designed for short-term assessments, they are often used beyond this period, leading to concerns about their ongoing effectiveness.
The study analyzed data from 8,843 women diagnosed with preeclampsia at a median gestational age of 36 weeks, collected between 2003 and 2016. The researchers examined both the PIERS-ML and fullPIERS assessments along with health outcomes. The findings revealed that while the PIERS-ML model performed well in identifying both the very high-risk and very low-risk groups over time, its ability to accurately predict risks for the larger high-risk and low-risk groups deteriorated significantly after 48 hours. Meanwhile, the fullPIERS model performed even worse.
Despite the findings, the study’s authors noted that clinicians may still use these models for ongoing risk assessments after the first 48 hours. However, they advised that predictions should be treated with increasing caution as the pregnancy progresses. The authors also emphasized the need for more accurate and reliable prediction models that perform well over time.
“Pregnancy hypertension outcome prediction models were designed and validated for initial assessment of risks for mothers; this study shows that such ‘static’ models, if used repeatedly over days, yield increasingly inaccurate predictions,” the authors stated.
The research underlines the need for better tools to monitor and predict the progression of preeclampsia in pregnant women, ensuring timely interventions and improving maternal health outcomes.
Disclaimer: The findings presented in this article are based on a study published in PLOS Medicine. The accuracy of preeclampsia prediction models may vary, and clinical decisions should always be made by healthcare professionals based on individual patient circumstances.
