Science & Healthcare

Background: One of the main causes of maternal and perinatal morbidity and mortality around the world is hypertensive disorder during pregnancy. After diagnosis of severe preeclampsia it is necessary of accurate risk assessment and management for both the mother and the fetus at different times. The prognostic model is an alternative basis for clinical practice, for predicting patient's future outcomes and for making decisions to improve them.

The aim: analysis of literature data of models for the prediction of pre-eclampsia and its complications.

Materials and methods. During the article's preparation an analysis was conducted of 55 English and Russian publications from PubMed, Clinical Key, Web of Science Core Collection, eLibrary, Google Scholar databases for the last 10 years, from January 2009 to June. Inclusion criteria: publications that contained a forecasting tool (model) containing three or more variables and that provided a probability of outcome, or suggested diagnostic or therapeutic actions, used discrimination and/or calibration to describe the performance of the forecasting model, internal and external validation of the model. The following search queries were used for the search: "preeclampsia and prognosis", "preeclampsia and complications" "EPH AND probability learning", "Hypertension-Edema-Proteinuria Gestosis AND prediction", "prediction of preeclampsia", "predictors of gestosis", "complications of preeclampsia".

Results. The research publications contained models for early prediction of preeclampsia and adverse maternal and perinatal outcomes. This literature review helped to find problems with predicting adverse perinatal outcome in preeclampsia, all existing models have one important limitation for generalization-they are all designed for single pregnancy, there is a relative high cost of models, and there was no information about the impact of the model on clinical practice: the number of days of treatment, the number of unnecessary diagnostic and therapeutic measures, complications.

Conclusions. due to the presence of systematic failure and limitations in the studies found, as well as the lack of adequate external validation checks, the reliability and reliability of existing models for predicting preeclampsia and its complications are very questionable.