PREDICTING PREECLAMPSIA DURING PREGNANCY: THE VALUE OF BLOOD PRESSURE MEASUREMENT TRENDS TO IDENTIFY WOMEN AT HIGHEST RISK

Abstract

Objective: Hypertensive disorders of pregnancy, including preeclampsia are serious complications increasing the risk of morbidity and mortality for both mother and baby. We hypothesise that blood pressure (BP) trends during pregnancy are a useful, accessible and practical predictor of preeclampsia onset. In this study we aimed to determine: (i) the association between mean arterial pressure (MAP), Systolic BP (SBP) and Diastolic BP (DBP) at time of pregnancy booking with subsequent preeclampsia development, and (ii) whether repeated measures of BP thought pregnancy can improve prediction of preeclampsia. Design and method: Data were from 11,008 women with a singleton pregnancy who received antenatal care and gave birth across three metropolitan Sydney hospitals 2017-2020. BP measurements were represented as MAP, SBP and DBP. Three sets of logistic regression models were used to yield odds ratios (ORs) with 95% confidence intervals (CIs) for the association of preeclampsia with: (1) booking BP, (2) summary BP measurements (mean, maximum, standard deviation (SD)) and (3) growth curve parameters (mean BP, slope and standard error of slope) derived from mixed effects models, representing each women's BP trajectory. Akaike's information criterion was used to determine the ‘best fitting’ model. Results: Booking BP was taken between 5 and 15 weeks of pregnancy (median 14 weeks). There were 80,624 BP measurements taken throughout pregnancy, the number and timing varied between women: median (min, max): 8(3,17) measurements. There were 191 cases of preeclampsia. Booking MAP before 16 weeks’ gestation (OR (95%CI) (per SD higher) (1.78 (1.53, 2.06) yielded a better model fit than SBP (1.67 (1.44, 1.94) or DBP (1.66 1.43, 1.96)) for preeclampsia. Of the summary measures, maximum MAP yielded the best fitting model (OR (95%CI) (per SD higher): 4.37 (3.76, 5.11). Growth curve parameters were significant predictors of preeclampsia, and substantially improved model fit in comparison to booking BP alone. Conclusions: Booking visit MAP was more strongly associated with preeclampsia than SBP or DBP. Using repeated BP measures, in the form of summary MAP measures and BP trajectories, improved the prediction of preeclampsia.