In This Issue

Three articles detail the process and results of reliability tests for proposed DSM-5 diagnoses and cross-diagnosis symptom domains (figure). Clinicians of varied disciplines interviewed 2,243 patients at 11 academic medical centers. Clarke et al. (p. 43) describe the study design. Each patient was interviewed at different times by two interviewers with training comparable to what would be available in actual practice. Regier et al. (p. 59) report that of 15 adult diagnoses, 10 had very good or good test-retest reliability. Of eight diagnoses tested in children and adolescents, four had very good or good reliability. A tool new to DSM is a symptom questionnaire completed by the patient before the diagnostic interview. Narrow et al. (p. 71) found that the reliabilities for the symptom items were generally good to excellent for adults and parent-reported child symptoms. In an editorial, Freedman et al. (p. 1) highlight the good reliability of borderline personality disorder and relate the questionable reliability of major depressive disorder to its heterogeneity.