The validity of the Montgomery–Aasberg depression rating scale as a screening tool for depression in later life

Abstract

The aims of the study were to examine the validity of the MADRS and to compare it with the validity of the Cornell Scale for Depression in Dementia (CSDD). We included 140 patients without dementia, with mean age 81.5 (sd 7.7) years. Trained psychiatric nurses interviewed all of them using the MADRS. In addition, for 70 patients caregivers were interviewed using the CSDD. A psychiatrist who had no access to the MADRS or the CSDD results made a diagnosis of depression according to the DSM-IV criteria for major depression, and the ICD-10 criteria was also applied for the 70 patients assessed with the CSDD. Twenty-two out of the 140 had depression according to the DSM IV criteria, whereas 25 out of 70 had depression according to the ICD-10 criteria. The area under the curve (auc) in a receiver operating characteristic analysis was 0.86 (95% CI 0.79-0.93) for the MADRS using the DSM-IV criteria. The best cut-off point was 16/17 with sensitivity of 0.80 and specificity of 0.82. The AUC for the CSDD was 0.83 (95% CI 0.71-0.95). The recommended cut-off score on the CSDD of 7/8 was valid but not the best in this study. The patients were diagnosed with a diagnosis of depression by only one psychiatrist, and the procedures in the two centres were not exactly the same. The MADRS has good discriminating power to detect depression in elderly persons and should be preferred to the CSDD for use with persons without dementia.