Research into Recovery and Wellbeing

A website about the work of the recovery research team at the Institute of Mental Health

The University of Nottingham

Using the adult Camberwell Assessment of Need (CAN) for research

I am not sure whether the adult CAN is suitable for my research. Can I have a sample copy?

You can download sample pages from each version of the CAN from the main adult CAN page, or borrow the CAN book from a library to help you decide whether to use CAN in your research. If you do decide to use CAN, you need first to buy (not borrow) the book.

CAN-R or CANSAS? What's the difference and which one should we use in our research study?

The CANSAS is a one-page assessment, and identifies only the need rating for each of the 22 CAN domains. CAN-R is a 22-page assessment, and records more information about each of the 22 domains. The most useful data to emerge from research using the adult CAN have been the level of Met and Unmet need, and the extra information assessed by the CAN-R has proved more difficult to interpret. Therefore, the CANSAS will be the most appropriate version for most research uses.

How can my researchers get trained in using the adult CAN?

The adult CAN book includes a training programme. We can also offer bespoke training to meet the needs of your research team. Email

Can other domains be added?

In developing the adult CAN, the challenge was to create an assessment that was comprehensive yet still as short as possible. Some domains that are important, such as spiritual problems and sleep, were therefore, after careful consideration, not included. Other domains can be added for research purposes, but these should be added on to the existing 22 domains. The psychometric properties of any extra domains will, of course, need to be established. Please note, any publication presenting CAN data should clearly separate data about the 22 CAN domains from data about any extra domains.

How should the adult CAN be used as an outcome measure?

Three summary variables are of most interest – total number of needs (ie Met + Unmet needs), total number of Met needs, and total number of Unmet needs (maximum 22 for each). Patient-rated Unmet need is the most useful variable: see

Staff and patient perspectives on unmet need and therapeutic alliance in community services.
Junghan U, Leese M, Priebe S, Slade M.
British Journal of Psychiatry. 2007, 191, 543-547.
Free article

Patient-rated mental health needs and quality of life improvement.
Slade M, Leese M, Cahill S, Thornicroft G, Kuipers E.
British Journal of Psychiatry. 2005,187, 256-261.
Free article

Does meeting needs improve quality of life?
Slade M, Leese M, Ruggeri M, Kuipers E, Tansella M, Thornicroft G.
Psychotherapy and Psychosomatics. 2004, 73, 183-189.

I have read about the 22 adult CAN domains being grouped into five clusters – Basic, Health, Social, Functioning and Services. How was this done, and should research with the adult CAN use these clusters?

The five clusters were based on expert judgement about domains that naturally cluster together, rather than any empirical analysis. They have not proved particularly useful in making sense of adult CAN data, so are not recommended.

Is staff and service user agreement on need ratings a measure of reliability?

No. Staff and service users assess needs differently, and this difference is not due to measurement error.

Has a system been developed for scanning adult CAN data?



Is there an electronic version of the adult CAN for research use?


Is it possible to include the adult CAN in a publication I am compiling?

The adult CAN is copyright-protected, and may not be published in any form without written permission. If the publication is being sold for profit, then it is unlikely that permission will be given. If the publication is being distributed on a not-for-profit basis, or as a consultation document within a service, then permission may be possible, providing the publication makes clear that the adult CAN book must be bought before the assessment is used. To request permission, e-mail