Causal clarity: directed acyclic graphs in medical research

Purpose of review This review explains the rationale for using directed acyclic graphs (DAGs) to support causal inference in observational epidemiologic research. It outlines why DAGs have gained prominence as a method for identifying appropriate covariate adjustment sets and contrasts this with earlier, association-based approaches to confounder selection. Recent findings Recent literature demonstrates that DAGs are increasingly used in epidemiology but are often presented with unnecessary conceptual and technical complexity. This has reduced their accessibility and limited their uptake among clinicians involved in evidence-based medical research. Summary This paper provides a clear and clinician-oriented explanation of DAGs using a simple, real-world clinical example. It demonstrates how DAGs can be applied in practice to identify a minimal adjustment set when addressing a causal research question.