DAG With Omitted Objects Displayed (DAGWOOD): a framework for revealing causal assumptions in DAGs

Directed acyclic graphs are widely used to represent causal assumptions in epidemiologic research, but the implicit background assumptions and omitted variables in a DAG often go undisclosed, creating opportunities for hidden causal assumptions to bias study design and analysis.

Target RWE researchers introduced the DAGWOOD framework—a structured approach for revealing and documenting the causal assumptions typically hidden in conventional DAGs, including omitted common causes, mediators, and selection nodes.

The DAGWOOD framework operationalizes transparent causal assumption disclosure in study design, providing Target RWE and pharma partners with a structured tool for FDA-facing pre-specification documents that require explicit causal assumption articulation.

Transparent causal assumption disclosure framework for regulatory-grade RWE study design, DAGWOOD as a structured tool for FDA pre-specification and study protocol documentation