The CORE system for sub-typing melancholic depression: Adaptation and psychometric properties of the Spanish version

Abstract We report the development procedure for a Spanish version of the CORE measure for subtyping melancholic depression and its psychometric properties. The sample comprised two subsets of depressive patients: 117 in-patients clinically assessed after admission and before discharge; and 34 in- and out-patients who were video-recorded, with each recording being rated by independent raters and re-rated six months later. The internal consistency of the Spanish version of the CORE was high, with Cronbach's alpha coefficient of 0.93 for the total CORE score. High intra-class correlation coefficient for the total score (0.96) and the three subscales (around 0.8) indicate high inter-rater reliability. Baseline and six-month ratings of video-recordings showed high test-retest reliability. Regarding construct validity, the correlation between the CORE and the Newcastle endogeneity scale scores was 0.64. A ROC curve generated an optimal cut-off of 10 to predict the definition of endogeneity/melancholia based on the Newcastle score >=6 and the DSM-IV-TR definition of melancholia. The comparison between the CORE scores at admission and at discharge revealed large differences, indicating demonstrable sensitivity to change. The Spanish version of the CORE system is a reliable and valid instrument for sub-typing depression in terms of melancholic versus non-melancholic sub-types.