Directed acyclic graphs (DAGs) in oncology research: applications and illustrated example.

Directed Acyclic Graphs (DAGs) are increasingly recognized as powerful tools in research and can be used for elucidating complex causal relationships inherent in cancer development, progression, and treatment outcomes. In oncology, where multifactorial risk factors are the norm, the ability to visualize and interrogate these relationships is critical. In observational studies, DAGs offer a rigorous framework for identifying confounders, mediators, and colliders, allowing researchers to estimate causal effects. This study presents the development and application of a DAG in a real-world observational study involving skin cancer. We outline practical steps for constructing DAGs and discuss how to use them to select variables that help control for confounding. Our approach provides a valuable guide for oncologists, epidemiologists, and other cancer researchers aiming to enhance the transparency and validity of causal claims across a wide range of oncologic contexts-from prevention and early detection to survivorship and health disparities.