Diagnosing comorbidity in psychiatric hospital: challenging the validity of administrative registers

BACKGROUND This study will explore the validity of psychiatric diagnoses in administrative registers with special emphasis on comorbid anxiety and substance use disorders. METHODS All new patients admitted to psychiatric hospital in northern Norway during one year were asked to participate. Of 477 patients found eligible, 272 gave their informed consent. 250 patients (52%) with hospital diagnoses comprised the study sample. Expert diagnoses were given on the basis of a structured diagnostic interview (M.I.N.I.PLUS) together with retrospective checking of the records. The hospital diagnoses were blind to the expert. The agreement between the expert's and the clinicians' diagnoses was estimated using Cohen's kappa statistics. RESULTS The expert gave a mean of 3.4 diagnoses per patient, the clinicians gave 1.4. The agreement ranged from poor to good (schizophrenia). For anxiety disorders (F40-41) the agreement is poor (kappa = 0.12). While the expert gave an anxiety disorder diagnosis to 122 patients, the clinicians only gave it to 17. The agreement is fair concerning substance use disorders (F10-19) (kappa = 0.27). Only two out of 76 patients with concurrent anxiety and substance use disorders were identified by the clinicians. CONCLUSIONS The validity of administrative registers in psychiatry seems dubious for research purposes and even for administrative and clinical purposes. The diagnostic process in the clinic should be more structured and treatment guidelines should include comorbidity.