Test-retest reliability of the Camberwell Assessment of Need (CAN)

The Camberwell Assessment of Need (the CAN) is a widely used instrument in several countries to assess needs for the severely mentally ill. The instrument consists of ratings of needs made by patient and staff. This paper reports test-retest reliability for staff ratings in the Swedish version. Fifty-six patients, most of them with a diagnosis of schizophrenia, were interviewed by one person from the social services and by one from the psychiatric clinic. Afterwards the ratings were discussed and negotiated in order to arrive at one unanimous rating. The same procedure was replicated after 2-4 weeks. The intraclass correlation coefficient for the sum of needs was 0.93, and for the sum of unmet needs and met needs 0.80 for both. There were no means differences between the first and second rating. An analysis was also made of the test-retest scores of individual needs and subscales. The results in this study suggest that the summary scores of the CAN generally are reliable over time when using the method applied here. General conclusions about test-retest reliability for individual needs and subscales are more difficult to make and demand larger samples.