Discriminant possibilities of the Hamilton depression scale: ROC analysis

Abstract

The purpose of this study is to compare discrimination power of original and reconstructed version of Hamilton’s depression scale in separation of depressive vs. anxious patients and to suggest some possibilities which offer ROC analysis. The subjects of the study were 119 patients of Psychiatric clinic in Novi Sad. 67 of them were diagnosed with some of the forms of affective disorders and 52 with an anxious-phobic diagnosis. Results of ROC analysis suggest that both instruments can be used in distinguishing depressive from anxious patients, but reconstructed version shows greater sensitivity and specificity with optimal cut-off score. It also has more significant AUC, which refers to probability of prediction on the basis of the whole spectrum of the results. These data is commented in relation with current debates, between unitaristic and pluralistic oriented authors, about the nature of the anxious-depression relationship.

Keywords

• anxiety-depression relationship
• Hamilton's depression scale
• ROC - analysis
• sensitivity
• specificity