RESUMO
In order to evaluate the Cornell Medical Index-Health Questionnaire (CMI) as a screening device for depressive disorders, we ran tests on 306 patients (148 males and 158 females, 18-83 years of age with a mean of 57.7 years) who visited our clinic, complaining of tinnitus, dizziness, sore tongue and/or throat discomfort that are often associated with affective disorders.<BR>The tests all the patients underwent simultaneously were the following three: CMI, self-rating depression scale (SDS) and self-rating questionnaire for depression (SRQ-D).<BR>The number of “yes” responses (CMI-DEP score) to 24 depression-related questions on the CMI correlated significantly with the SDS score (Spearman's rank correlation coefficient (ρ) =0.570, p<0.0001, SDS score=1.6× “CMI-DEP score” +31.8) and with the SRQ-D score (Spearman's ρ=0.659, p<0.0001, SRQ-D score=0.9× “CMIDEP score” +5.1) as well.<BR>Referring to the classifiations for SDS scores (20-39: little depressive, 40-49: slightly depressive, 50-80: moderately depressive) and for SRQ-D scores (0-10: normal, 11-15: borderline, 16-36: possibly masked depression) reported by others, we classified the individuals into the three groups according to the CMI-DEP scores: normal (0-5), probably depressive (6-11) and depressive (12-24).<BR>Supposing that a depressive state was correctly diagnosed with the SDS method and the SRQ-D method as well, the sensitivity of the CMI-DEP method was relatively low (46.3% against SDS and 59.2% against SRQ-D), but the specificity of the method was high enough (84.3% against SDS and 85.1% against SRQ-D).<BR>These results suggest that the CMI-DEP classification can provide an accurate estimate of depressive disorders.