The Social Behaviour Schedule (SBS) was examined as a measure of disability in a series of 66 chronic psychiatric patients, who were also assessed using the Social Role Performance Schedule (SRP) and the ninth edition of the Present State Examination (PSE). All three schedules have been used in previous studies in their present form. SBS and SRP measures were correlated with each other and neither showed any relationship with subjectively described current PSE symptoms. SBS deficit was associated with a clinical diagnosis of schizophrenia and with problems in supervised occupational activity. Long-term hospital residents showed greater SBS deficit than chronic patients in active rehabilitation programmes, although impaired SRP was the major correlate with the type of supervision provided. SRP measures can be compared with population norms, unlike measures derived from the SBS, which cover more severe ranges of disability.

A two-phase study of psychiatric prevalence has been carried out among the inhabitants (aged 18–64) of a Dutch health area (Nijmegen). In phase 1, a random sample of 3232 persons answered the GHQ–30 (response rate:75%). In phase 2,486 of the respondents were interviewed withthe full PSE within two weeks. The relationship of PSE-‘caseness’ (ID ≥ 5) and GHQ score was expressed in a logistic regression model, the parameters of which showed strong agreement with the Canberra results, for all socio-demographic variables examined, except for urbanization. means of the logistic model the point-prevalence of PSE cases was calculated at 7·3% (range5·5%-9·2%).

Prevalence did not differ significantly in men and women. Higher case rates were found in age category of 55–59 years, among divorced and widowed persons, the lower educational and occupational levels, the unemployed, chronically ill and unable to work, and with city people. This study adds further evidence to a growing body of epidemiological data suggesting similar rates and patterns of psychiatric disorder in populations in industrialized countries.

The Organic Brain Syndrome (OBS) and the Depression (D) scales derived from the Comprehensive Assessment and Referral Evaluation (CARE) were translated into Italian and used in a survey of geriatric institutions in Milan. During the survey validity and reliability tests of the scales were conducted.

Inter-rater reliability (total score weighted kappa) was highly satisfactory for both scales (0·96 for OBS and 0·83 for D scale). Reliability was assessed three times during the survey and showed good stability for both scales, with a slight but significant trend towards reduction over time for the D scale. Reliability of the D scale was significantly lower when the subjects interviewed scored highly on the OBS scale (severe cognitive impairment).

Criterion validity was highly satisfactory both for the OBS scale (cut-off point 4/5: sensitivity 77%, specificity 96%, positive predictive value 91%) and the D scale (cut-off point 10/11: sensitivity 95%, specificity 92%, positive predictive value 84%).

Results are discussed with special reference to longitudinal assessment of reliability, the choice of the cut-off point, and the context-dependent properties of questionnaires.

Two-stage screening for psychiatric disorder was carried out with a sample of 99 male and 95 female outpatients with insulin-dependent diabetes mellitus, using the General Health Questionnaire and the Clinical Interview Schedule. The estimated prevalence of psychiatric disorder in this sample was 18%: only 28% of the psychiatric morbidity so identified was detected by physicians. In contrast, physicians rated 10% of the sample as psychiatric ‘cases’ using a simple 6-point scale.