Screen of drug use: Diagnostic accuracy for stimulant use disorder

Stimulant misuse, overdose, and related deaths have increased dramatically. Identifying and referring individuals with stimulant use disorder to treatment may reduce misuse and overdose. This study validated the 2-item Screen of Drug Use (SoDU; Tiet et al., 2015) to screen for stimulant use disorder (and for cocaine and amphetamine use disorders) in a VA primary care setting, and to establish its concurrent diagnostic validity among diverse subgroups of patients, including age, gender, race/ethnicity, marital status, educational level, and PTSD status.

Archival data from 1283 VA primary care patients recruited in California were examined. This predominantly male sample matched general VA primary care patient population characteristics. A total of 79 individuals (6.2%) met criteria for a stimulant use disorder. The criterion for having a stimulant use disorder was a DSM-IV cocaine use disorder and/or amphetamine use disorder diagnosis based on the Mini International Diagnostic Interview.

For stimulant use disorder, the 2-item SoDU was 93.67% sensitive (95% confidence interval [CI], 85.84%-97.91%), and 89.12% specific (95% CI, 87.22%-90.82%). When tested in diverse subgroups of patients, the sensitivity ranged from 66.67% to 100% and specificity ranged from 76.81% to 94.17%. When a follow-up question was added, the sensitivity was unchanged and the specificity was 99%, with lower false positive rate.

The SoDU, especially with a follow-up question, is an appropriate instrument for routine screening of stimulant use disorder in VA primary care settings. It has good concurrent diagnostic validity for diverse groups of patients.