Novel Application of Case-Control Analysis in Confounded Exposure-Response
*Yaning Wang, U.S. Food and Drug Administration 


Exposure-response analysis is routinely conducted in drug development and regulatory review. Since clinical trials are rarely randomized to drug exposure, a common criticism for exposure-response analysis is the potential confounding relationship between the drug exposure and some influential covariates that may become unbalanced due to the regrouping of patients based on the drug exposure. A case study will be presented to demonstrate how to apply case-control analysis to control the distribution of influential covariates when regrouping patients based on drug exposure results in unbalanced distribution of those factors. Case-control analysis also helps to clarify the contribution of drug exposure and patient covariates to the responses and avoid biased estimation of response size due to drug exposure alone, which is critical for sample size determination to study a new regimen.