BMJ Open. 2026 Sep 22;16(9):e121623. doi: 10.1136/bmjopen-2026-121623.
ABSTRACT
INTRODUCTION: Hypertensive disorders of pregnancy (HDP) are associated with maternal and neonatal morbidity. Although delivery is the definitive treatment, antihypertensive agents are often prescribed to control blood pressure to facilitate safe prolongation of pregnancy. International guidelines vary in their treatment recommendations and preferences for methyldopa, labetalol and nifedipine. Within the field of HDP, randomised trials are often underpowered and observational studies are limited in causal inference due to confounding. Therefore, we propose a target trial emulation with hospital-based electronic health record (EHR) data to compare methyldopa, labetalol and nifedipine in pregnancies complicated by HDP, focusing on maternal and perinatal outcomes.
METHODS AND ANALYSIS: We will emulate a target trial by using multicentre EHR data of four university and two non-university hospitals in the Netherlands covering the period from 2006 to 2025, although data availability varied across hospitals. With algorithm-based data extraction, we will collect EHR data from patients with an obstetric diagnosis-treatment code and a prescription of methyldopa, labetalol or nifedipine during pregnancy. Time zero between patients will be aligned to minimise the risk of immortal time bias. Therefore, patients will be stratified into gestational age groups at treatment initiation. Primary outcomes include severe hypertension and severe small for gestational age (SGA) below the third percentile. Secondary maternal outcomes include the development of pre-eclampsia, haemolysis, elevated liver enzymes and low platelets syndrome, acute kidney injury, maternal intensive care unit (ICU) admission, mortality and mode of delivery. Secondary perinatal outcomes include birth weight, SGA below the 10th percentile, fetal growth restriction (diagnosed based on ultrasound data), gestational age, preterm birth <37 and <34 weeks, Apgar score, neonatal ICU admission, neonatal hypoglycaemia and perinatal mortality. Weighted multivariable mixed-effects logistic or linear regression models will be used, depending on the outcome.
ETHICS AND DISSEMINATION: The Medical Research Ethics Committee of the University Medical Centre Utrecht (NedMec, 23-092/DB) confirmed that formal ethical approval was not required. Study findings will be disseminated through publication in peer-reviewed scientific journals.
PMID:42772874 | DOI:10.1136/bmjopen-2026-121623