Development of a biophysical screening model for gestational hypertensive diseases

Abstract Background To investigate the possibility of using maternal biophysical parameters only in screening for the different types of gestational hypertensive diseases. Methods A total of 969 pregnant women were randomly screened in first and second trimester, of which 8 developed Early-onset Pre...

Full description

Bibliographic Details
Main Authors: Sharona Vonck, Anneleen S. Staelens, Dorien Lanssens, Kathleen Tomsin, Jolien Oben, Liesbeth Bruckers, Wilfried Gyselaers
Format: Article
Language:English
Published: BMC 2019-05-01
Series:Journal of Biomedical Science
Subjects:
Online Access:http://link.springer.com/article/10.1186/s12929-019-0530-0
Description
Summary:Abstract Background To investigate the possibility of using maternal biophysical parameters only in screening for the different types of gestational hypertensive diseases. Methods A total of 969 pregnant women were randomly screened in first and second trimester, of which 8 developed Early-onset Preeclampsia, 29 Late-onset Preeclampsia, 35 Gestational Hypertension and 897 women had a normal outcome. An observational maternal hemodynamics assessment was done via standardized electrocardiogram-Doppler ultrasonography, Impedance Cardiography and bio-impedance, acquiring functional information on heart, arteries, veins and body fluid. Preliminary prediction models were developed to test the screening potential for early preeclampsia, late preeclampsia and gestational hypertension using a Partial Least Square Discriminant Analysis. Results A combined model using maternal characteristics with cardiovascular parameters in first and second trimester offers high screening performance with Area Under the Curve of 99,9% for Early-onset Preeclampsia, 95,3% for Late-onset Preeclampsia and 94% for Gestational Hypertension. Conclusions Using biophysical parameters as fundament for a new prediction model, without the need of biochemical parameters, seems feasible. However, validation in a large prospective study will reveal its true potential.
ISSN:1423-0127