[Concepts of psychiatric diagnoses: a further investigation of the reliability of psychiatric diagnoses (author's transl)].

A total of 128 physicians and psychologists, working in six psychiatric hospitals, were asked to allocate 10 psychiatric diagnoses on the basis of typical clinical criteria (symptoms, course of illness, etiological and anamnestic data), with which they were supplied. The correct diagnosis was made in 45.3%, while in 82% the broad diagnostic group was correctly ascribed. The respondents were divided into two groups, according to the amount of diagnostic data they had been given (though in fact the information necessary for differential diagnosis was given equally to both groups). The group which made diagnosis on the grounds of more comprehensive data (ten characteristic diagnostic items) did significantly worse them the group which had few data (six characteristic items). The stereotypes for mania, endogenous depression, obsessional-compulsive neurosis, and antisocial personality disorder were correctly defined by 75% of more of the respondents. The profiles of the remaining six diagnostic groups were associated with the corresponding diagnoses by 50% or fewer of the respondents.