Competing Risk Model in Screening for Preeclampsia by Mean Arterial Pressure and Uterine Artery Pulsatility Index at 30-33 Weeks' Gestation

Abstract

Objective: To assess risk for preeclampsia (PE) based on maternal characteristics, mean arterial pressure (MAP) and uterine artery pulsatility index (Ut-PI) at 30-33 weeks' gestation. Methods: Screening study in singleton pregnancies including 2,140 that subsequently developed PE and 83,615 that were unaffected by PE, gestational hypertension or delivery of small-for-gestational-age neonates (normal group). We developed a survival time model for the time of delivery for PE by combination of maternal characteristics and history with MAP and Ut-PI multiple of the median (MoM) values (biophysical test). Data on third-trimester MAP and Ut-PI were available in 350 cases of PE and 13,878 of the normal group. The detection rate of PE requiring delivery within 4, 6 and 8 weeks of the visit was estimated. Results: In pregnancies with PE the log₁₀ MoM values of MAP and Ut-PI were inversely related to gestational age at delivery. Biophysical testing detected 90, 65 and 53% of PE with delivery within 4, 6 and 8 weeks of the visit, at a fixed false-positive rate of 5%. Interpretation: Testing by maternal characteristics, Ut-PI and MAP at 30-33 weeks could identify 90% of pregnancies developing PE and requiring delivery within the subsequent 4 weeks.

Competing Risks Model in Screening for Preeclampsia by Biophysical and Biochemical Markers at 30-33 Weeks' Gestation

“…The fitted regression models for firstly, gestational age in weeks at delivery with PE according to maternal characteristics and history, secondly, log 10 uterine artery PI, log 10 MAP, log 10 PlGF and log 10 sFlt-1 at 30-33 weeks in unaffected pregnancies, and thirdly, marker log 10 MoM values at 30-33 weeks of gestation at time of delivery for pregnancies with PE were reported previously [14,15]. …”

“…At risk cut-off of 1:100, the estimated FPR, DR and PPV for PE requiring delivery within the subsequent 4 weeks were 6, 91 and 8%, respectively, in screening by maternal factors, uterine artery PI and MAP (biophysical test) [14], 4, 93 and 11% in screening by maternal factors, serum PlGF and sFlt-1 (biochemical test) [15] and 2, 95 and 18% in screening by maternal factors and all biomarkers (combined test).…”

“…The following steps were carried out in our previous studies [14,15]: firstly, the values of uterine artery PI, MAP and serum PlGF and sFlt-1 were log 10 transformed to make their distribution gaussian, secondly, backward stepwise multiple regression analysis was used to determine which of the factors amongst the maternal characteristics and gestation were significant predictors of the log 10 artery PI, log 10 MAP, log 10 PlGF and log 10 sFlt-1, adjusting for the adverse pregnancy outcomes as specified (PE, GH and SGA), thirdly, the distribution of markers was expressed as multiple of median (MoM) in all cases, correcting for the significant predictors as defined in the multiple regression and fourthly, in the cases of PE regression analysis was used to determine the relationship between log 10 MoM values with gestational age at delivery.…”

“…Bayes' theorem was used to combine the prior information from maternal characteristics with uterine artery PI, MAP and serum PlGF and sFlt-1MoM values [13,14,15,22]. The distribution of gestational age at delivery with PE was defined by two components: firstly, the prior distribution based on maternal characteristics and secondly, the distribution of uterine artery PI, MAP and serum PlGF and sFlt-1 MoM values with gestational age in pregnancies affected by PE.…”

“…In high-risk pregnancies the distribution is shifted to the left and the smaller the mean gestational age the higher is the risk for PE. In a study involving the measurement of uterine artery pulsatility index (PI) and mean arterial pressure (MAP) at 30-33 weeks' gestation in 350 cases that subsequently developed PE and 13,878 unaffected pregnancies, the estimated detection rate (DR) of PE requiring delivery within 4 weeks was 91% at a fixed false positive rate (FPR) of 5% [14]. In another study in which maternal serum placental growth factor (PlGF) and soluble fms-like tyrosine kinase-1 (sFlt-1) were measured at 30-33 weeks in 118 cases of PE and 3,734 unaffected pregnancies, the estimated DR for PE requiring delivery within 4 weeks of the visit was 95% [15].…”