3G-05 - Switching to Preferred Antihypertensives during Early Pregnancy in Individuals with Pre-Existing Hypertension and Risk of Pregnancy Loss: Emulating a Target Trial Using Real-World Data

Background: Hypertension among women with childbearing potential is on the rise. Maintaining proper blood pressure during pregnancy is crucial to prevent maternal and neonatal complications. However, limited evidence on the risk-benefit of switching antihypertensive medications in early pregnancy complicates clinical decision-making.

Objectives: To evaluate the risk of pregnancy loss associated with switching to preferred antihypertensives in the first trimester.

Methods: This retrospective cohort study used healthcare claims data from Merative MarketScan® Commercial Databases (2011-2020). We identified pregnant women aged 18 to 55 years with an essential hypertension diagnosis from ≥1 inpatient or ≥2 outpatient encounters in the 6 months before conception. Pregnancy episodes resulting in live delivery or pregnancy loss (i.e., spontaneous abortions or stillbirths) with continuous insurance enrollment from 6 months before pregnancy until 6 months after the end of pregnancy were identified using validated algorithms. We included pregnant women whose most recent antihypertensive prescription fill before pregnancy was neither for renin-angiotensin system-acting agents (contraindicated in pregnancy) nor for one of the three preferred agents (labetalol, nifedipine, or methyldopa). We emulated a target trial with two treatment strategies: switching to preferred antihypertensives within the first trimester (T1; grace period) versus not switching. We used the clone-censor-weight method with Cox proportional hazards regression to estimate the probability of an individual remaining uncensored at the end of T1, accounting for demographics, comorbidities, healthcare utilization measured at baseline, and time-dependent prenatal care initiation. The pregnancy loss risk was compared between strategies, and 95% CIs were derived from nonparametric bootstrap 500 replicates.

Results: Of 6,062 pregnancies with second-line therapy (mean [SD] maternal age: 34.7 [5.8] years), 2,444 (39.4%) had a fill for preferred agents by T1. At the end of T1, 1,927 and 2,543 pregnancies remained in the switching and non-switching groups, respectively. Switchers were younger and less likely to have baseline comorbidities. After adjusting for censoring weights, all covariates were balanced at the end of T1. The estimated risk of pregnancy loss was 26.1% (95% CI, 24.9%-27.4%) in switchers and 23.3% (95% CI, 22.1%-24.5%) in non-switchers. The estimated risk ratio was 1.12 (95% CI, 1.08-1.17). The sensitivity analysis restricted to pregnancies remaining by the end of T1 showed a risk ratio of 1.12 (95% CI, 0.88-1.43).

Conclusions: Further research is warranted to assess the choice of antihypertensives and optimal timing for switching, considering maternal and fetal outcomes.