Health Service Utilisation of People Living with Psychosis: Validity of Self-report Compared with Administrative Data in a Randomised Controlled Trial

Background: Self-reported service use informs resource utilisation and cost estimates, though its validity for use within economic evaluations is uncertain. Objective: The aim of this study is to assess agreement in health resource-use measurement between self-reported and administrative data across different resource categories, over time and between different recall periods by subgroups among Australians living with psychosis. Methods: Data were obtained for 104 participants with psychotic disorders from a randomised controlled trial. Agreement between self-reported resource-use questionnaires and administrative data on community-based services and medication use was assessed through estimating differences of group mean number of visits and medications used and intraclass correlation coefficients (ICC) over multiple time periods. Results: ICC showed moderate agreement across most time periods for general practitioners, psychiatrists and mental health medications. No clear trends were discernible over time, between varying lengths of recall periods nor across participant subgroups. Conclusion: Despite poor agreement, when measuring visits to psychologists and other health professionals, small overall differences in group mean number of visits indicate that self-reported data may still be valid for use in economic evaluations in people living with psychosis.