Despite advances in maternal medicine, pre-eclampsia continues to be a major contributor to maternal, fetal and neonatal mortality and morbidity. Pre-eclampsia is not a single disorder but a syndrome. The early onset disease is more severe, and is considered to have different pathophysiology than the late onset disease. It is unlikely that one single model will accurately predict both early and late onset disease. A brief by the HTA calls for a systematic review on the predictive accuracy of markers, separately and in combination for predicting pre-eclampsia, especially early onset disease. We have identified 59 systematic reviews on clinical characteristics, biomarkers and ultrasound in the prediction of pre-eclampsia, including our HTA report (HTA No. 12060), and 69 prediction models for pre-eclampsia.

However, clinical applicability of the aggregate meta-analyses is limited due to the observed heterogeneity in population, test characteristics including timing and cut off, and outcomes. Furthermore, they often do not account for multiple predictors, and are mainly focussed on any pre-eclampsia than early onset. An Individual Patient Data (IPD) meta-analysis will allow us to assess the differential accuracy of the tests in various subgroups according to the risk status. It will provide us with sufficient sample size to develop and validate a multivariable prediction model, for the clinically important outcome of early onset pre-eclampsia. Furthermore, by taking into account the clustering within studies, the developed model will avoid the model performance deterioration encountered in aggregate meta-analysis, when the individual’s baseline risk is different from the average estimated during model development.

Prior to use of prediction models in clinical practice, there is a need to successfully validate the model in multiple datasets external to the model development phase. This often takes many years to accomplish. None of the prediction models have robustly undertaken external validation to assess model performance in relevant population. Our International Prediction of Pre-eclampsia IPD Collaborative Network (IPPIC) of global researchers has provided access to the IPD from existing studies and large databases, thereby allowing us to both develop and validate the risk prediction models simultaneously, within an IPD meta-analysis framework. Our collaborative approach encourages consensus towards a single, well developed, and validated prognostic model, rather than a number of competing and non-validated models for the same clinical question.