Citation

Veldhuizen S, Rush B, Urbanoski K. J. Stud. Alcohol Drugs 2014; 75(6): 1039-1046.

Affiliation

Social and Epidemiological Research Department, Centre for Addiction and Mental Health, Toronto, Ontario, Canada, Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada.

Copyright

(Copyright © 2014, Alcohol Research Documentation, Inc., Rutgers, The State University of New Jersey)

DOI

unavailable

PMID

25343663

Abstract

OBJECTIVE: Mental disorders are common among people presenting for the treatment of substance-related problems, and guidelines recommend systematic screening for these conditions. However, even short screens impose a considerable burden on clients and staff. In this study, we evaluated the performance of a single screening question (SSQ) relative to longer instruments.

METHOD: We analyzed a sample of 544 adults recruited from outpatient substance use disorder treatment centers in Ontario, Canada. Clients were asked a simple SSQ, followed by the Kessler Screening Scale for Psychological Distress (K6) and the Global Appraisal of Individual Needs Short Screener (GAIN-SS). Caseness was ascertained using the Structured Clinical Interview for DSMIV (SCID). We measured and compared performance using receiver operating characteristic (ROC) curve analysis and explored client characteristics associated with screen performance using ROC regression. Last, we used logistic regression to test whether SSQs can be usefully combined with other measures.

RESULTS: The prevalence of past-month disorder was 71%. The performance of the SSQ (AUC = 0.77, 95% CI [0.72, 0.83]) was similar to, and not significantly different from, those of the K6 (AUC = 0.78) and the GAIN (AUC = 0.79). The K6 and the GAIN performed better than the SSQ among people with a psychotic disorder. The addition of the SSQ slightly improved the performance of the other measures.

CONCLUSIONS: SSQs can offer screening performance comparable to that of longer instruments. Reasons for caution include the small number of possible thresholds, lower accuracy than other measures in identifying psychotic disorder, and possible differential functioning in different populations. Performance of all three screens was also moderate; when prevalence is high and resources are available, offering full assessments may be preferable to screening. Nevertheless, SSQs offer an intriguing area for further evaluation. (J. Stud. Alcohol Drugs, 75, 1039-1046, 2014).

Language: en