Background: Hypertensive disorders of pregnancy (HDP) are one of the leading causes of maternal morbidity and mortality worldwide. Management of these conditions can pose many clinical dilemmas, and can be particularly challenging during the immediate postpartum period. Models for predicting and managing postpartum hypertension are necessary to help address this clinical challenge.
Objective: To evaluate predictive models of blood pressure (BP) spikes postpartum (PP) and investigate clinical management strategies to optimize care.
Methods: A retrospective cohort study of PP women participating in remote BP monitoring. Postpartum BP spike was defined as BP ≥ 140/ 90 if on an antihypertensive and BP ≥ 150/100 if not on an antihypertensive. We identified 3 risk level patient clusters (low, medium, and high) in predicting patient risk of BP spike PP days 3-7. The variables used in defining these clusters were peak systolic blood pressure (SBP) before discharge, BMI, patient SBP per-trimester, heart rate, gestational age, maternal age, chronic hypertension and gestational hypertension. For each risk cluster, we focused on two treatments: 1) PP length of stay < or >/= 3 days and 2) discharge with or without BP medications. We evaluated the effectiveness of the treatments on different subgroups of patients by estimating the conditional average treatment effect (CATE) values in each cluster respectively using a causal forest. Moreover, for all patients, we considered discharge with medication policies depending on different discharge BP thresholds. We used a doubly robust policy evaluation method to compare the effectiveness of the policies.
Results: 413 patients were included, of which 267 (64.6%) had a PP BP spike. The treatments for patients at medium and high risk were considered beneficial, the 95% confidence intervals of constant marginal average treatment effect for antihypertensive use at discharge were (-3.482, 4.840) and (-5.539, 4.315); and for a longer stay were (-5.544, 3.866) and (-7.200, 4.302). For patients at low risk, the treatments were not critical in preventing a BP spike with 95% confidence intervals of constant marginal average treatment effect (1.074, 15.784) and (-2.913, 9.021) respectively. We considered the option to discharge with antihypertensive use at different BP thresholds a) ≥ 130 mmHg and/or ≥ 80, b) ≥ 140 mmHg and/or ≥ 90, c) ≥ 150 mmHg and/or ≥ 100, or d) ≥ 160 mmHg and/or ≥ 110, we found that policy (b) was the best option at P<0.05.
Conclusions: We identified 3 possible strategies to prevent outpatient BP spikes PP: 1) medium and high-risk patients should be considered for a longer PP stay or participate in daily home monitoring, 2) medium and high-risk patients should be prescribed antihypertensives at discharge, and 3) antihypertensive treatment should be prescribed if patients are discharged with BP ≥ 140/90.
Keywords: Hypertension; blood pressure spike; machine learning, random forest method; postpartum.
Copyright © 2024. Published by Elsevier Inc.