Screen of drug use: Diagnostic accuracy for opioid use disorder

Abstract

AimsOpioid related deaths have more than tripled in recent years. Identifying and referring individuals with opioid use disorder (OUD) to treatment is one of the promising approaches to reduce opioid related deaths. However, using urine toxicology to identify opioid misuse is not reliable. This study validates the Screen of Drug Use (SoDU) to screen for OUD in the primary care setting, and establish its concurrent diagnostic validity among diverse subgroups of patients, including age, gender, race/ethnicity, marital status, educational level, and PTSD status. MethodsWe used data from 1283 primary care patients recruited in the VA in CA. This sample matched patient characteristics general VA population with mean age = 62, and 95% men. A total of 10.4% met DSM-5 criteria for any drug use disorder and 2.7% met criteria for OUD (with or without other drug use disorders). An opioid use abuse or dependence diagnosis based on the Mini International Diagnostic Interview was used as the criterion for having a DSM-5 opioid use disorder. ResultsThe SoDU was 100% sensitive (95% confidence interval [CI], 89.9%–100%), and 86.3% specific (95% CI, 84.3%–88.1%). When tested in subgroups of patients, the SoDU maintained 100% sensitivity in all subgroups. Specificity ranged from 74.5% to 94.2% for diverse subgroups of patients. ConclusionsThe SoDU is an appropriate instrument to screen for opioid use disorder in primary care. It is brief, easy to use, and has good concurrent diagnostic validity for diverse groups of patients.