While the central tenets behind the CSRI do not vary, regardless of where an economic evaluation is undertaken, it is important to make sure that the CSRI is appropriate for use in that country. There are two central challenges to amending the CSRI for other countries. First, service systems are very different in each country; and different services may be available, be provided by different agencies, or draw from different funding streams. The second challenge is that service titles common in one country may be unknown in another. Perhaps the same service is called something different, or a similar service name is used, but applied to a type of support that has a very different focus. To capture service use accurately, the integrated service lists must reflect the context in which the evaluation is set.
Below are some examples of CSRIs that have been adapted for use in other countries.
EPILSON and the CSSRI-EU
The challenges in making sure the CSRI is fit-for-purpose are even greater in cross-national research where comparisons across countries will be undertaken. Developed for the EPSILON study, the Client Socio-Demographic and Service Receipt Inventory – European Version (CSSRI-EU) was a careful adaptation of the CSRI for which researchers used translation and cross-country validation before administration to 404 people across five countries: Denmark, England, Italy, the Netherlands and Spain. Balancing local relevance with international generalisability mean that a number of separate pieces of work were undertaken to ensure validity across countries:
- Select an existing instrument for adaptation (the CSRI)
- Translate into other languages
- Conduct focus groups to refine the content and translation
- Revise the original instrument for content, terminology or language
- Pilot the instrument
- Re-examine the instrument’s performance and revise appropriately.
The EPSILON CSRI may be downloaded here.
This CSSRI-EU has a similar format to the CSRI, although the first section on background information includes questions on a wider set of socio-demographic details. Again, a three-month retrospective was employed as a period sufficiently long enough to pick up the wide range of services that patients might use, without stretching their recall powers past the level of reasonable accuracy. Response rates were very good, with a 100% rate of overall completion and low rates of missing data for individual items. These EU versions of the CSRI have been used as the basis for other international versions of the CSRI.
More information on the development of the international CSSRI-EU can be found here.
An English abstract for a paper (that was written in German) regarding the feasibility of the German adaptation of the CSSRI-EU (Roick, C et al) may be found here.
ICare and the CSRI
For our current ICare project, we have developed some guidance to help colleagues from other countries develop a CSRI for use locally, using Germany to illustrate the issues. This document can be found here.
QUATRO and the CSRI
The CSSRI was also used as part of the QUATRO study. The cost-effectiveness of this intervention (involving 409 participants) was investigated, and the paper can be found here.
This study collected data from four countries (England, Germany, Holland and Italy and Spain). The CSRIs administered in each country can be found here in:
In addition to the change of language, there are variations between some of these CSRIs. The Dutch CSRI has removed some services from the list of community-based day services as these would not be applicable. The Dutch form does also not include the supplemental hospital section. The German CSRI has included other services which are appropriate given their service provision context.
Developing the CSRI for Brazil
The CSRI has also recently been translated into Portuguese and adapted for use in the Brazilian healthcare system. This version and the accompanying guidance can be found here. An abstract from a conference presentation can also be found here and a report of this study will be available soon.