Detecting depressive disorders in drug abusers: a comparison of screening instruments

Previous investigators have reported a high prevalence of depressive symptoms in drug-dependent patients. Given the responsiveness of depressive disorders to both psychological and pharmacological treatments, it is desirable to find an economical, efficient screening instrument to detect depressive disorders in this population. In this study, 6 depression symptom screening scales (Beck Depression Inventory, Hamilton Depression Scale, Raskin Depression Scale, Degree of Illness Rating, Symptom Checklist 90 Overall, and Depression Subscale) based on either clinician interview or patient self report, were compared according to their utility in detecting cases of depression among 64 applicants for treatment at a substance abuse treatment unit of a community mental health center. The criteria for a case of depression were the Research Diagnostic Criteria (RDC) which are specified and operationalized. Cases identified using previously described cutoff scores on the screening scales were compared to rates based on the RDC and sensitivity and specificity were determined. The results showed that: (1) although the sensitivity of the symptom scales was applicable, ranging from 65--94%, the specificity was less impressive, ranging from 39--61%, and (2) the Beck Depression Inventory, a 13-item patient self report was the most sensitive and specific and is recommended for screening drug-dependent populations for depression.