Interventional Delivery for Hypertensive Disorders: A Causal Approach to Maternal and Infant Health Outcomes

• Activity Number: 83 - Contributed Poster Presentations: Health Policy Statistics Section
• Type: Contributed
• Date/Time: Monday, August 3, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Jerson Cochancela* and Roee Gutman
• Companies: Brown University and Brown University
• Keywords: causal inference; bayesian methods; hypertension; MITSS; multiple imputation; propensity score

Abstract

One of the most common medical problems arising during pregnancy is hypertension, complicating 9-10% of cases [1]. Hypertensive conditions are associated with increased risk of stroke or placenta abruption for the mother and low birth weight or decreased oxygen delivery for the infant. Some conditions clearly call for delivery, but many require week-by-week judgment to balance mother and infant risks. Under the potential outcomes framework [2], we examine the individual level causal effects of the decision to deliver at 34-38 weeks gestation in mothers with non-severe hypertensive disorders on neonatal intensive care unit admission, adverse respiratory outcomes of the infant, prolonged hospital stay, a neonatal morbidity index, and unplanned Cesarean delivery. We demonstrate the utility of Multiple Imputation with Two Splines in Subclasses [3,4] to generate credible estimates of the causal impact of intervention and provide investigators with empirical guidance across a crucial period of gestation.