External validation of the fullPIERS model in women with mild pre-eclampsia

Rethans, B. (Bente) (2015) External validation of the fullPIERS model in women with mild pre-eclampsia. thesis, Medicine.

Text RethansB.pdf Restricted to Registered users only Download (369kB)

Abstract

Background:Worldwide, pre-eclampsia is a leading cause of maternal and perinatal morbidity and mortality. Therefore the need for a validated model to predict adverse maternal outcomes is evident. Von Dadelszen et al. developed and internally validated the fullPIERS model, a model that has the ability to predict adverse maternal outcomes within 48 hours up to 7 days after inclusion using 6 predictor variables. Before adopting such a prediction model it is important to assess generalizability by externally validating the original model in different populations as we did in this study. Methods:The HYPITAT study was a multicenter, parallel, open-label randomized controlled trial undertaken in six academic and 32 non-academic hospitals in the Netherlands between October 2005 and March 2008. This study showed that induction of labour in women with a singleton pregnancy complicated by gestational hypertension or mild pre-eclampsia after 37 weeks gestation is associated with improved maternal and neonatal outcomes than expectant monitoring. For validation purposes of this study we used the mild pre-eclampsia subgroup from the HYPITAT dataset. (N=246). The fullPIERS prediction model equation was used to calculate individual probability scores. Missing prediction parameters were retrieved by screening charts of the participating patients. We assed performance of the model using AUC (ROC), risk stratification and calibration within 48 hours up to 7 days after admission. Results:Of the 246 women included from the HYPITAT study, 17(6.9%) women experienced an adverse maternal outcome at any time after inclusion. Adverse maternal outcome were in three cases observed within 48 hours and in 10 cases within 7 days after inclusion. The FullPIERS predicted adverse maternal outcomes within 48 hour and showed an area AUC ROC of 0.974 (95% CI 0.942-1.000). Up to 7 days after admission the AUC ROC showed an area of 0.730 (95% CI 0.542-0.917). FullPIERS successfully stratified the women into different risk categories. Conclusions:The fullPIERS model showed good performance in a subgroup of mild pre-eclampsia patients. It showed a positive trend in successfully stratifying the patients into different risk categories. The model can be used as a rule out test for women admitted with mild pre-eclampsia. However due to the population subgroup had a small numbers of adverse outcomes, no significant conclusion can be drawn. In future external validation needs to be done in a prospective appropriately powered study.

Item Type:	Thesis (Thesis)
Supervisor name:	Facultary Tutor: and Franssen Dr. M.T.M Gynecologist University Medical Centre Gr
Supervisor name:	Daily tutor and Akkermans, Drs J. Leiden University Medical Center and Local Tutor. and Ganzevoort Dr. W. Gynecologist-perinatologist and Academic Medical Center Amsterdam
Faculty:	Medical Sciences
Date Deposited:	25 Jun 2020 10:59
Last Modified:	25 Jun 2020 10:59
URI:	https://umcg.studenttheses.ub.rug.nl/id/eprint/1983

Actions (login required)

View Item