A Comparison of the Validity of the Cornell Scale and the MADRS in Detecting Depression among Memory Clinic Patients

Abstract

Background: The aim of the study was to compare the validity of the Cornell Scale for Depression in Dementia (CSDD) and the Montgomery-Aasberg Depression Rating Scale (MADRS) among memory clinic patients. Methods: The scales were independently completed for 125 patients. The diagnosis of depression was made by psychiatrists blinded to the depression scores. Results: The mean score of the Mini-Mental State Examination was 25.5 (SD: 4.6), of the CSDD 6.8 (SD: 4.9), and of the MADRS 8.5 (SD: 6.8). In receiver operation characteristics (ROC) analyses, the AUC for the CSDD was 0.73 (95% CI: 0.63–0.82) using the International Classification of Diseases-10 (ICD-10) criteria for depression, and 0.68 (95% CI: 0.57–0.79) using the DSM-IV criteria. The AUC was 0.88 (95% CI: 0.81–0.95) for the MADRS using the ICD-10 criteria, and 0.84 (95% CI: 0.76–0.92) using the DSM-IV criteria. Conclusion: Both scales are suitable as screening tools. According to the ROC analyses, the MADRS seems better at distinguishing depressed from nondepressed patients.

A Comparison of the Cornell Scale for Depression in Dementia and the Montgomery-Aasberg Depression Rating Scale in a Memory Clinic Population

“…The mean score of the MADRS was 9.7 (SD: 6.7), and 318 (61.2%) had a sum score of 7 and above. According to our previous validity study among the same group of patients, these cutoff points had the best sensitivity and specificity, and were therefore the most suitable for the present study [12]. …”

“…In a previous validation study using data from daily clinical practice, we filled in the two scales independently of each other, using two different sources of information (caregivers and patients). The correlation between the results of the two scales in that study was Spearman's rho = 0.43 [12]. A similar study, in a memory clinic in Brazil using the same methodology, also found a low correlation between the two scales of 0.46 [13].…”

“…However, the prevalence rates of depression in dementia differ greatly across studies depending on differences in patient populations and which diagnostic criteria are being used [8,9,10,11]. We have previously found that about 40% of the patients referred to a memory clinic in Norway fulfill the diagnostic criteria of a depressive disorder according to the 10th edition of the International Classification of Diseases criteria, and a similar figure was also reported in a memory clinic population in Brazil [12,13]. …”

“…Each item can be scored between 0 and 2 (absent, mild or intermittent and severe) with a total score of 38 [18]. The scale is designed for use in institutions, but in studies it is often used on outpatients [12,13]. The MADRS was completed by the physician who examined the patients, based on a semi-structured interview with the patient.…”

“…The MADRS is a scale applied by a clinician (physician or psychologist), and it is generally filled in after a semi-structured interview with the patients. It has been developed to follow changes in depressive symptoms over time, but is also used as a screening instrument to detect depression [12,21]. It contains some questions that overlap with those used to identify symptoms of dementia.…”