On the validity of the Bech-Rafaelsen Melancholia Scale (BRMS)

Background: The findings published to date on convergent validity of the BRMS are mainly concerned with the correlation with other observer-rating scales for depression. In many studies on the evaluation of therapeutic interventions self-rating scales are used in connection with observer-rating scales. Therefore, findings on the relations among instruments of both groups are necessary in order to justify the combination of a particular observer-rating scale with a particular self-rating scale. In the ICD-10 three different degrees of severity of depressive episodes are distinguished. No data on the discriminant validity of the BRMS with respect to this new diagnostic classification are available till now. Methods: 45 depressed inpatients were assessed with two observer-rating scales (BRMS and DEPRES of the AMDP system) and two self-rating scales (BDI and DS). Results: The discriminant validity with reference to the three degrees of severity of depressive episodes as defined in ICD-10 is at r = 0.80 very high. The convergent validity of the BRMS is high at r = 0.70, related to the DEPRES. The correlations between the BRMS and the BDI as well as the DS were clearly lower, at 0.53 and 0.32 each. Conclusions: The only moderate convergent validity between self-rating and observer-rating scales is a strong argument for a multi-methodological approach in the context of therapy evaluation. (C) 1999 Elsevier Science B.V. All rights reserved.