The CSRI is a practical instrument for recording factual information on resource use as reported by study participants and/or their carers. As with any research instrument, care should be taken at the design stage to ensure the content is relevant to the evaluation and that the questions are phrased in a manner that is easy to understand.

Generally, the CSRI does not require a high-intensity training course to use, as the questions and answers are, in the main, self-explanatory. However, where we have provided training, this has tended to cover two topics: an explanation of why the information is collected and what is then done with it, and some more detailed information around recording and coding given the service context and research questions, and to ensure consistency within the evaluation. This has been particularly useful where there are several interviewers, or where data are collected across different sites.

We have also developed manuals for some versions of the CSRI. These tend to follow a similar format: describing why the data are collected, and providing support for recording and coding consistency. Manuals are available to download from this website for the Homestart study of families experiencing stress, the iCST study of an individualised sessions of cognitive stimulation, a study investigating day hospital care as opposed to traditional inpatient care, and the twelve years on study looking at service use and costs for those leaving psychiatric hospitals.


If you require assistance with questions not covered below, please contact Jennifer Beecham (LSE: or KENT: or Martin Knapp (

Frequently asked questions:

1. Is the CSRI copyrighted? Do I need permission to use it?

The CSRI is not copyrighted although we ask those using a version of the CSRI to reference it appropriately

2. How do I reference the CSRI?

Beecham, J. and Knapp, M. (2001) Costing psychiatric interventions, in G. Thornicroft (ed.) Measuring Mental Health Needs, Gaskell, 2nd edition, 200-224.

This core reference is for a chapter in the second (updated) edition of a book published in 2001, with the original book having been published in 1992. You may also cite the weblink for this chapter.

3. Can I amend it to more closely meet the needs of my study?

The CSRI can be adapted to meet the needs of any new project. We suggest you first look at those available on these webpages (either featured or more examples) to see if there is one that will fit closely with your research questions, service context and user group. Careful scrutiny will help ensure the CSRI asks all the questions relevant to the research. It is particularly important that the services listed are the ones your participants are likely to use.

4. My study is funded by NIHR; can I only use the questions relating to health careservices?

Economic theory advocates taking a comprehensive and societal approach in economic evaluation and this is the basis on which the CSRI has been developed. However, there are circumstances where a tighter perspective is justified and the service list can be amended.

5. Can I use the CSRI in other countries?

Service systems vary considerably across countries, and are often known by different titles or are located in different agencies. It is advisable to amend the service list in the light of these differences. The international versions page has versions developed for some European countries, and articles discussing how international CSRIs were developed and validated for a study of schizophrenia. We have also included a document developed for the ICare project which discusses translating the CSRI for use in other countries, using Germany as an illustration.

6. Are there any questions on the CSRI which people find difficult to answer?

The CSRI is an adaptable instrument, and through repeated administration and iteration it has been discovered that some ways of asking questions yield better results than others. We would recommend taking great care with wording with regards to questions related to individual budgets/ direct payments. Respondents often have difficulty disentangling this from the receipt of benefits, and in order to obtain an accurate figure care should be taken to ensure the respondent is clear about what is being asked. For example, the GREAT CSRI uses a definition card within the interview.

7. Are there any recommendations about how to go about collecting information on unpaid care hours provided?

Asking carers for the number of hours of unpaid care they provide in a typical day/week can also be challenging. Sometimes carers will indicate that they spend each moment of every day caring. This may be true, but researchers may want to separate active (or more ‘hands-on’) care from supervision. Recently, the CSRI has been collecting this information by asking co-resident carers what percentage of hours in a typical day that the care recipient can be left alone for, which can be used to calculate the number of hours of care needed in a typical day. This can be seen for example in the CSRI used as part of the GREAT study. Carers not living with the person they care for are simply asked for their hours of care over a given period. These individuals typically find it easier to estimate care hours as the act of arriving or leaving the home of the cared for person more easily separates caring hours from non-caring hours.

8. Is it too intrusive to ask individuals for their salary?

Some respondents feel that questions about their income from employment are sensitive, and may not want to provide this information. Additionally, some individuals may only know their own income and may not be able to accurately estimate household income. An alternative approach is to ask for individuals what job they do and their typical working hours per week. The national average wage by job category (and age/gender if desired) can then be applied to produce an estimation of the income from employment.