Pilot Validation of a Brief Screen Tool for Substance Use Detection in Emergency Care

Screening and brief intervention for substance use in health care systems is recommended to identify and intervene with patients who abuse alcohol and other substances. However, there is limited research on the utility of short, single-item questions to identify illicit substance users.

Pilot validation of two single-item screening questions to detect illicit substance use, one for marijuana and one for other illicit drugs. The goal was to identify sensitive, time-efficient screening questions that can be easily integrated into busy health care settings.

A cross-sectional design was used. At intake, along with questions for tobacco and alcohol, nurses administered two brief screen questions to adult patients seen in designated areas in a large urban medical center. After patients were triaged to rooms, health educators (blind to brief screen responses) administered the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) as the reference standard.

On the ASSIST, 14% and 9% of participants reported risky marijuana and illicit (nonmarijuana) drug use, respectively. Sensitivity values for the marijuana and street drug questions were 72% (95% confidence interval [CI] 67% to 78%) and 40% (95% CI 32% to 48%), respectively. Specificity values for the marijuana and street drug questions were 96% (95% CI 95% to 97%) and 99% (95% CI 98% to 99%), respectively. Values differed minimally as a function of patient characteristics.

It is important to use validated questions to identify substance misuse so that individuals are not missed in the screening process. It is the possible that administration protocols play a role in detection rates. Future research is needed to identify easy-to-administer drug use screening questions.