November 4, 1992

The Self-reporting of Cocaine Use-Reply

Author Affiliations

Emory University Atlanta, Ga

JAMA. 1992;268(17):2375-2376. doi:10.1001/jama.1992.03490170045016

In Reply.  —Drs Gleason and Barnum's references indicate that gas chromatography/mass spectrometry identifies cocaine use in patients who have negative EMIT cocaine metabolite assays, the assay used by us as our criterion standard. Using these data and the figures from our article, they have calculated the prevalence of cocaine use in our population to be an astonishingly high 51%. We agree that our prevalence rate was only a minimum estimate of recent cocaine use and appreciate his point that actual prevalence may be even higher.Dr Holland suggests a creative design to ensure informed consent and to address the question of how foreknowledge of urine testing would affect questionnaire responses. However, urine screening information based on those subjects who consent to urine testing is not easily extrapolated to those subjects who refuse consent for urine testing. Moreover, as Holland points out, analysis based on missing data can only approximate true