In Reply Dr Lozano comments that drug interactions can be key confounding variables in pharmacogenetic studies and wonders if the results of our study might be changed if we had considered comedications. We agree that pharmacogenomic data can be influenced by nongenetic factors (eg, coadministered medicines and drug interactions) and that such confounding factors might lead to overestimation or underestimation of an association. In fact, uncharacterized extrinsic factors or unclear phenotypes of patients usually lead to negative findings in pharmacogenomic studies or replication analyses.
Chung W, Hung S. Genetic Factors Associated With Severe Cutaneous Adverse Reactions—Reply. JAMA. 2014;312(20):2166. doi:10.1001/jama.2014.14148