The importance of measuring cotinine levels to identify smokers in clinical trials.

First-time-in-humans studies of drugs (phase I) typically exclude unsuitable volunteers by testing for recreational drugs. However, volunteers are usually not screened for cotinine, a metabolite of nicotine, even though tobacco products may alter pharmacokinetic and pharmacodynamic parameters and withdrawal from tobacco may cause additional adverse events. The accuracy of personal histories as a means of excluding smokers was examined prospectively in three phase I units in the northeastern, midwestern, and southwestern United States. In studies intended for nonsmokers, 45 of 282 purported nonsmokers screened before enrollment tested positive for cotinine. This suggests that personal histories are unreliable in determining tobacco use in clinical trials designated for nonsmokers.