[DSM III personality diagnosis in a group of chronic offenders].

A group of chronic offenders (n = 44) was investigated and their personality disorders coded according to two different methods, namely the ICD 9 (International Classification of Diseases 9) and the American DSM III (Diagnostic and Statistical Manual III). The total number of personality disorders diagnosed by DSM III was less than by ICD 9. This cannot be attributed only to the fact that according to DSM III cyclothymic disorders do not have a direct correlate in axis II-personality disorders, but are, instead, assigned to the group of affective disorders. The majority of persons in whom the diagnosis of cyclothymic personality disorders was made on ICD 9 were rated as having one of the "extroverted personality disorders" in DSM III. If the personality disorders are combined to form three groups - (a) eccentric (close to psychoses), (b) extroverted and (c) introverted - then the overall correlation between ICD 9 and DSM III is considerable. The DSM III system promises a higher reliability due to the existence of criteria for inclusion and exclusion. The diagnosis "borderline personality disorder" was compared with the DIB Gunderson interview (kappa 0.89). In the international classification DSM III borderline personality disorder was equivalent to a diagnosis of cyclothymic, irritable, hysteric or sociopathic personality disorder.