[Analysis of the reliability of diagnostic criteria and classifications in psychiatry]

The future development of mental health funding in France will be based primarily on administrative data and a diagnostic classification: ICD-10.

The objective of this study was to investigate the inter-rater reliability of diagnoses of schizophrenia, schizoaffective disorder, bipolar disorder and unipolar depression based on the ICD-10 classification. A cross-sectional study was conducted on all of the patients admitted to a public psychiatric hospital on the same day in January 2010. Cohen's kappa coefficient was used to estimate inter-rater reliability and diagnostic congruence between the treating psychiatrist and an independent psychiatrist using a standardized interview. In total, 112 patients participated in the study. 56.3% of participants were male and the mean age was 43.4 years (SD = 15.0). The proportion of discordant diagnoses ranged from 14.3% (schizophrenia and bipolar disorder) to 42.9% (schizophrenia and schizoaffective disorder). The coefficient of congruence was 0.48 for schizoaffective disorder, 0.56 for unipolar depression, 0.80 for schizophrenia and 0.90 for bipolar disorder. The reliability of schizoaffective disorder and unipolar depression diagnoses may seem limited. Further research on larger samples is needed to confirm the results of this study and to determine the impact of poor diagnostic reliability on future mental health funding in France.