Sensitivity, specificity, and discordance with self-report of nail sample testing for alcohol and cannabis.

Abstract

Nails accumulate the alcohol metabolite, ethyl glucuronide (ETG), and the cannabis metabolite, carboxy- delta-9-THC over 3-6 months. Few studies have examined nail toxicology testing's sensitivity and specificity and the agreement between nail testing and self-reported alcohol and marijuana use. In an ongoing clinical trial, 1101 veterans completed initial telephone questionnaires and were then asked to mail nail clippings for substance use analysis. We examined sensitivity and specificity of ETG and carboxy- delta-9-THC in nails compared to self-report of alcohol use patterns (the AUDIT-C) and substance-related harms (alcohol and THC subscales of the ASSIST). We then examined factors associated with discordance between nails and self-report. Almost two-thirds (707/1101) of respondents mailed in nail clippings. Those with returned nails were disproportionately married, white race, older, and less depressed. At a threshold of 8pg/mg, sensitivity was only.50 to detect risky alcohol use and.49 to detect alcohol-related issues. Sensitivity for marijuana issues was only.61. Specificity was greater than.77 for all measures. Factors associated with positive nails/negative self-report (i.e. false positives) for risky alcohol use on the Audit-C included more pain and being unmarried; false positive nails for alcohol-related issues on the ASSIST were associated with being unmarried and non-Hispanic ethnicity. False positive nails for THC-related issues on the ASSIST were associated with being African American, Hispanic, and having had legal issues. At standard cut-offs, nail measures had low sensitivity and higher specificity. The groups who disproportionately submit positive nails/negative self-report could have substance use patterns not adequately captured by self-report, inaccurate self-report due to social pressures, or distinct drug metabolism.