Estimating Causal Effects of Third‐Stage Management on Postpartum Haemorrhage in a Midwifery Context: An Evidence Synthesis Approach for Constructing Directed Acyclic Graphs

BACKGROUND: Estimating the causal effect of third-stage management approaches on preventing postpartum haemorrhage (PPH) in the context of physiologic birth using observational data requires conditioning on specific variables, with selection relying on assumptions about their roles in the exposure-outcome pathway that are rarely made explicit. OBJECTIVES: To apply the evidence synthesis for constructing DAGs approach, incorporating findings from a systematic review, to develop a causal directed acyclic graph (DAG) that clarifies these assumptions and identifies the minimum set of variables needed to reduce bias in estimating the causal effects of physiologic third-stage care versus oxytocin prophylaxis on PPH. DATA SOURCES: MEDLINE, Embase, CINAHL, Web of Science, and Cochrane Central Register of Controlled Trials (to December 15, 2023), ClinicalTrials.gov (to July 8, 2024), and reference lists of eligible studies. STUDY SELECTION AND DATA EXTRACTION: The systematic review included randomised and non-randomised studies involving individuals with physiologic birth or minimal obstetric interventions. Two authors independently screened studies. DAG development was based on the subset of non-randomised studies. For each, one reviewer extracted outcome, exposure, control variables and mediators. SYNTHESIS: Eligible studies were analysed in three stages: (i) mapping each study's saturated implied graph; (ii) translating each posited connection using causal criteria to create study-specific DAGs; (iii) synthesising individual DAGs into an integrated DAG. The assumptions underlying this process were specific to the midwifery context in Ontario, Canada and translation was guided by midwifery expertise and existing literature. RESULTS: Four non-randomised studies were included. Expert consultation identified 20 factors influencing third-stage management. The integrated DAG comprised 339 directed edges connecting 35 covariates, yielding four minimal sufficient adjustment sets. CONCLUSIONS: The integrated DAG and minimal sufficient adjustment sets are valuable tools for informing future study design and analysis, helping to minimise bias in estimating the causal effect of physiologic third-stage care versus oxytocin prophylaxis on PPH in the context of physiologic birth, while also exposing the assumptions about causal relationships between variables to scrutiny.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.