Diagnostic Accuracy of the SLU AMSAD Scale for Depression in Older Adults Without Dementia

Abstract Objectives To evaluate the diagnostic accuracy of a short depression screening tool, the Saint Louis University (SLU) AMSAD depression scale, relative to the Geriatric Depression Scale–15 (GDS-15) and Montgomery-Asberg Depression Rating Scale (MADRS), and in relation to the Diagnostic and Statistical Manual of Mental Disorders–Fifth Edition (DSM-5) criteria for major depressive disorder, in cognitively intact older adults. Design Cross-sectional. Setting Outpatient geriatric psychiatry clinic. Participants Fifty older adults (aged ≥65 years) without dementia. Measurements GDS-15, MADRS, SLU AMSAD, and DSM-5 criteria for major depressive disorder were administered. Results Total scores (continuous variables) for the GDS-15, MADRS, and SLU AMSAD correlated significantly with the DMS-5 criteria for major depressive disorder (MDD) [area under the curve (AUC) ≥ 0.93, sensitivity = 0.93, and specificity ≥ 0.80]. Optimal cutoffs were 9+ for GDS-15, 18+ for MADRS, and 7+ for SLU AMSAD. When score results were categorized according to their known cutoffs for mild, moderate, and severe depression, AUC values were again high (range = 0.82-0.89), with adequate levels of sensitivity (0.87-0.93) and specificity (0.71-0.86), distinguishing no or mild depression from moderate or severe depression in relation to the DSM-5 diagnostic criteria for MDD. Conclusions and Implications Strong diagnostic accuracy was shown for the 3 scales. The SLU AMSAD performed as well as the GDS-15 and slightly better than the MADRS. The superiority of the SLU AMSAD is supported by the fact that it encompasses only 5 simply worded, simply scaled items to be used in busy clinical settings.