Drug interactions should be evaluated in patients.

Drug development involves learning and confirming in healthy volunteers and ultimately in patients. This approach should apply to drug-drug interactions because requirements to extrapolate pharmacokinetic and pharmacodynamic findings from healthy volunteers to patients are not uniformly met, and the designs of interaction studies often have little relevance to actual clinical use. Guidance to health-care professionals on a safe and effective dose for individual patients should include information from drug-drug interaction studies in the intended patients.