Objective: To determine if serial blood pressure (BP) measurements during pregnancy (before 30 weeks gestation) improve the prediction of preeclampsia, as compared with a single early BP in a routine clinical care study. Design and method: Data were from a multi-ethnic cohort of women with a singleton pregnancy who received antenatal care and delivered at one of three metropolitan hospitals in Sydney, Australia (2017-2020). BP was measured at booking visit (<16 weeks gestation) and subsequent scheduled antenatal visits. The primary BP parameter of interest was mean arterial pressure (MAP) calculated as: DBP+1/3(SBP–DBP). Logistic regression models were fit with preeclampsia as the outcome and the incremental predictive power of the following predictors was assessed using C-statistics and Likelihood Ratio Tests: (1) booking MAP (starting BP) (2) mean MAP (average MAP from booking to 30 weeks) (3) MAP slope (change in MAP over time) and (4) standard error of the slope (fluctuation of MAP around the slope). Results: There were 10052 women (186 preeclampsia events) that had 2 or more BP measurements (median 4 measurements) including booking visit (median 14 weeks) and between <30 weeks gestation. Higher booking MAP was associated with a higher odds of preeclampsia (OR (95%CI) per standard deviation higher 2.27 (1.99, 2.58), as was a higher mean MAP 2.90 (2.54, 3.31). Including the random slope parameter (C-statistic 0.7743, p=0.039), but not further adding the standard error of the slope (C-statistic 0.7739, p=0.108), improved the prediction of preeclampsia in addition to booking and mean MAP. Results held in the presence of confounders. Conclusions: Booking MAP was an important predictor of preeclampsia. Using all BP measurements before 30 weeks’ gestation enhances the prediction of preeclampsia by capturing the cumulative burden of higher BP levels (through measures such as the mean and slope). Clinically, the additive value of serial measures allows for identification of women who would benefit from closer surveillance in late pregnancy when preeclampsia is most likely to develop. Results will now be validated against an international research cohort study.
Harris et al. (Fri,) studied this question.