Early pregnancy screening for hypertensive disorders in women without a-priori high risk

Objectives To evaluate the performance of mean arterial pressure, uterine artery pulsatility index and soluble endoglin level alone or in combination in screening for hypertensive disorders in pregnant women without a-priori high risk. Methods This was a nested casecontrol study of women with singleton pregnancies without a-priori high risk who developed pregnancy-induced hypertensive complications. Women were enrolled into the study at 1114 weeks' gestation, when mean arterial pressure and uterine artery pulsatility index were recorded and a blood sample was taken for measurement of soluble endoglin. Women were followed up in the clinic to detect development of any hypertensive disorder. Each affected case was matched with two normotensive control women with uncomplicated pregnancies that resulted in phenotypically normal infants. Mean values for each variable were compared between cases and controls. Sensitivities, positive predictive values and negative predictive values at fixed specificity were derived from receiveroperating characteristics (ROC) curves. Results During the study period, 2120 patients were examined. Of these, 170 (8.02%) were excluded because they were lost to follow-up and in 52 (2.45%) there was fetal death or miscarriage before 24 weeks' gestation. Thus, 1898 cases formed the cohort population. Of these, 89 (4.69%) patients developed complications (study group), including 16 (0.84%) cases with early pre-eclampsia (PE), 60 (3.16%) with late PE and 13 (0.68%) with gestational hypertension (GH). There were 49 (2.58%) cases of spontaneous preterm delivery before 34 weeks. The rest of the cohort population (1760 (92.73%) patients, the base cohort) were not affected by PE or GH. The control group comprised 178 patients. The best model for the prediction of any of the types of hypertensive disorders was one that combined mean arterial pressure with soluble endoglin (area under the ROC curve (AUC), 0.83). The predictive value of the three combined markers was highest for screening for early and late PE (AUC, 0.86 and 0.83, respectively). When each marker was considered alone, the highest prediction of any type of hypertensive disorder was achieved by mean arterial pressure (AUC, 0.73). Sensitivity was lowest for detection of GH when screening both by individual and by combined markers. Conclusion First-trimester screening can be useful in predicting women at high risk of developing hypertensive disorders of pregnancy but more prospective longitudinal studies are needed. Copyright (c) 2012 ISUOG. Published by John Wiley & Sons, Ltd.