Treatments for pregestational chronic conditions during pregnancy: emulating a target trial with a treatment decision design

TL;DR

This paper proposes a novel method to estimate causal effects of early pregnancy treatments by emulating target trials centered around clinical decision points, demonstrated through two large database protocols.

Contribution

It introduces a treatment decision design approach to address methodological challenges in causal inference during early pregnancy.

Findings

Protocols for two target trials using administrative data.

Emulation of a target trial centered on clinical landmarks.

Emphasizes importance of careful causal effect estimation.

Abstract

As a solution to methodologic challenges inherent to estimating causal effects of exposures in early pregnancy, we suggest emulating a target trial using a treatment decision design, wherein time zero is centered around clinical landmarks where treatment decisions may occur, such as the date of preconception counseling or prenatal care initiation. These ideas are illustrated via protocols for two target trials in large administrative databases, antidepressant use for pre-existing depressive disorder and antihypertensive medication use for mild-to-moderate chronic hypertension. Careful consideration of these issues is critical to the identification of the causal effects of early-pregnancy pharmacotherapies on pregnancy outcomes.