Causal Directed Acyclic Graphs to Mitigate Confounding Bias in Exposure-Response Analyses.

The use of exposure-response (E-R) analysis to support drug development and treatment individualization requires estimating the causal effect of drug exposure on response. This may be challenging when the E-R relationship is confounded. This perspective examines the problem of confounding in E-R analyses in oncology through the lens of causal inference, demonstrating how a causal directed acyclic graph (DAG) can be used as a powerful tool to understand these challenges and identify potential solutions.