The validity of the depression rating scales in discriminating between citalopram and placebo in depression recurrence in the maintenance therapy of elderly unipolar patients with major depression.

The World Federation of Societies of Biological Psychiatry guidelines for treatment of unipolar major depression has recommended three depression rating scales for evaluating outcome: The Hamilton Depression Rating Scale (HAM-D), the Montgomery-Asberg Depression Rating Scale (MADRS), and the Bech-Rafaelsen Melancholia Scale (MES). In this study we evaluated the ability of these scales to differentiate between citalopram and placebo in the recurrence prevention of unipolar depression. The study is a psychometric reexamination of a trial on the efficacy of citalopram versus placebo in the maintenance therapy of elderly patients with unipolar depression. Internal validity (the Cronbach coefficient alpha, the Loevinger coefficient of homogeneity, and factor analysis) of the three scales has been examined to evaluate their unidimensionality. In the outcome analysis for depression recurrence, the conventional cutoff scores of the three scales are used. In total, 60 patients received citalopram and 61 patients received placebo in the maintenance phase of 48 weeks. The results showed that the internal validity was higher for MES and MADRS than for HAM-D. Using the MADRS, 67.2 % of the patients on placebo and 31.6 % of the patients on citalopram developed a depression recurrence (ratio 2.12); using HAM-D17, 42.6 % on placebo and 13.3 % on citalopram developed a depression recurrence (ratio 3.20); and using the MES, 34.4 % on placebo and 11.7 % on citalopram developed a depression recurrence (ratio 2.94). The conventional cutoff scores of HAM-D17 and MES for depression recurrence indicated a ratio between citalopram and placebo of around 3, while the conventional cutoff scores of MADRS for depression recurrence indicated a ratio of only around 2. In future trials on the recurrence prevention of unipolar depression, a cutoff score of 25 rather than 22 on the MADRS is recommended.

RCT Entities:   Population Interventions Outcomes