Prediction model of preeclampsia using machine learning based methods: a population based cohort study in China

Hypertensive disorders of pregnancy-chronic hypertension, gestational hypertension, and preeclampsia-are uniquely challenging as the pathology and its therapeutic management simultaneously affect mother and fetus, sometimes putting their well-being at odds with each other. Preeclampsia, in particular, is one of the most feared complications of pregnancy. Often presenting as new-onset hypertension and proteinuria during the third trimester, preeclampsia can progress rapidly to serious complications, including death of both mother and fetus. While the cause of preeclampsia is still debated, clinical and pathological studies suggest that the placenta is central to the pathogenesis of this syndrome. In this review, we will discuss the current evidence for the role of abnormal placentation and the role of placental factors such as the antiangiogenic factor, sFLT1 (soluble fms-like tyrosine kinase 1) in the pathogenesis of the maternal syndrome of preeclampsia. We will discuss angiogenic biomarker assays for disease-risk stratification and for the development of therapeutic strategies targeting the angiogenic pathway. Finally, we will review the substantial long-term cardiovascular and metabolic risks to mothers and children associated with gestational hypertensive disorders, in particular, preterm preeclampsia, and the need for an increased focus on interventional studies during the asymptomatic phase to delay the onset of cardiovascular disease in women. Show more

Reduction of maternal mortality is a target within the Millennium Development Goals. Data on the incidence of preeclampsia and eclampsia, one of the main causes of maternal deaths, are required at both national and regional levels to inform policies. We conducted a systematic review of the incidence of hypertensive disorders of pregnancy (HDP) with the objective of evaluating its magnitude globally and in different regions and settings. We selected studies using pre-specified criteria, recorded database characteristics and assessed methodological quality of the eligible studies reporting incidence of any HDP during the period 2002-2010. A logistic model was then developed to estimate the global and regional incidence of HDP using pre-specified predictor variables where empiric data were not available. We found 129 studies meeting the inclusion criteria, from which 74 reports with 78 datasets reporting HDP were analysed. This represents nearly 39 million women from 40 countries. When the model was applied, the overall estimates are 4.6% (95% uncertainty range 2.7-8.2), and 1.4% (95% uncertainty range 1.0-2.0) of all deliveries for preeclampsia and eclampsia respectively, with a wide variation across regions. The figures we obtained give a general idea of the magnitude of the problem and suggest that some regional variations might exist. The absence of data in many countries is of concern, however, and efforts should be made to implement data collection and reporting for substantial statistics. The implementation of large scale surveys conducted during a short period of time could provide more reliable and up-to-date estimations to inform policy. Copyright © 2013 Elsevier Ireland Ltd. All rights reserved. Show more