The Adaption of the Client Sociodemographic and Service Receipt Inventory for Costing Mental Health Services in Brazil.

Abstract

Background: There is a scarcity of tested instruments for measuring mental health services and costs. The Client Sociodemographic Service Receipt Inventory (CSSRI) is the most used tool in economic evaluation in mental health in Europe; it was translated into five languages, and it was mainly used to evaluate deinstitutionalisation process in mental health system reform.

Aims of the study: To translate and adapt to the Brazilian healthcare system, and to test its inter-rater reliability, validity and its feasibility in a deinstitutionalized sample of psychiatric hospital living in residential facilities.

Method: The translation and adaptation of CSSRI to Brazilian context was done by a focus group with eight experts on public mental health services, covering all the available Brazilian healthcare services. Decisions on the extent of conceptual overlap between British and Brazilian version were discussed until reaching expert consensus. The inter-rater reliability and applicability of this version, called ``Inventário Sociodemográfico de Uso e Custos de Serviços - ISDUCS'', was tested in a sample of 30 subjects with moderate to severe mental disorders living in residential facilities. Because the lack of medical record or another source, ISDUCS's validity was assessed using Kappa coefficient agreement to compare between resident`s answers and their professional carers`answers.

Results: The same structure of the original instrument was kept, with an additional list of items for costing consumable services. The main modifications were on items related to education, occupational status and on detailed descriptions of public health services. The agreement between two mental health raters was good to excellent for the majority of items, with Kappa coefficient ranged from 0.6 to 1.0. Because 43% of the sample was unable to answer questions about regularly taken medications and consultations with health professionals, an exploratory analysis was done to identify potentially related variables. Greater severity of psychiatric symptoms and lower independent living skills were related to the inability to answer these questions. Agreement between residents and carers was good to excellent for socio and demographic variables, living situation and occupational status, income, visits to a psychologist, occupational therapists and social workers.

Conclusion: ISDUCS is the first tool for economic evaluation including mental health services translated and adapted to Brazilian context. Despite the widespread use of CRSSI among people with schizophrenia in Europe, this study found that greater severity of symptoms led to high rate of missing responses. Inter-rater reliability was excellent as a whole. Small sample size didn't allow generalisation of results of this preliminary testing.

Implications for health provision and use: ISDUCS may be suitable for people with mental illness but requires additional sources of information such as carers and medical records. ISDUCS could be used for monitoring health service use in general practice.

Implications for health policies: Despite some limitations, this instrument was used to measure mental health service costs in three Brazilian studies, generating data for supporting local mental health policies, for boosting empirical research in the country and for supporting modelling studies.

Implications for further research: It should be tested further in other health settings and samples.