Diagnostic concordance for DSM-IV sleep disorders: a report from the APA/NIMH DSM-IV field trial

OBJECTIVE: The frequency and ranking of DSM-IV sleep disorder diagnoses of clinical patients with complaints of insomnia, as well as rates of diagnostic agreement and disagreement between two types of interviewers, were investigated. METHOD: Interviewers at five clinical sites assessed 216 patients referred for insomnia complaints. One sleep specialist and one general clinician interviewed each patient in an unstructured clinical interview, assigned DSM-IV diagnoses, and indicated their reactions to the diagnostic system. RESULTS: Insomnia due to another mental disorder was the most frequent DSM-IV diagnosis across sites, followed by primary insomnia. Interviewers at the five sites differed significantly in the rankings they assigned to different diagnoses. In addition, sleep specialists at most sites diagnosed psychiatric forms of insomnia more frequently than nonspecialists. Kappa values for agreement between the two types of clinicians on multiple DSM-IV sleep diagnoses ranged from 0.26 to 0.80 across sites, indicating moderate agreement overall. Kappa values for individual diagnoses varied across sites and specific diagnoses and ranged from poor to excellent. Interviewers' ratings of their confidence in diagnoses and the fit and ease of use of the DSM-IV categories also showed significant variability related to site and type of interviewer. CONCLUSIONS: The distribution of diagnoses highlights the importance of psychiatric and behavioral factors in the assessment of insomnia. Site- related variability indicates a need for greater standardization in the application of sleep disorder diagnostic criteria. Diagnostic concordance for these diagnoses, while only moderately good, likely reflects actual clinical practice and would be improved through the use of standardized (or structured) interviews and increased training.