In healthcare studies, participants often complete resource-use measures (RUMs). RUMs are often adapted on a study-by-study basis, with limited psychometric testing. This study undertook qualitative interviews with experienced health economists to test the usability, face validity and content validity of a newly developed RUM (ModRUM). ModRUM is a brief, generic RUM with a core module (6 questions) on healthcare use and modules to increase depth and breadth. A purposeful sampling strategy was used to recruit health economists with recent experience of implementing RUMs. Health economists were sent two versions of ModRUM (core module and depth extension). During telephone interviews, pre-defined questions on the relevance, clarity, conciseness, and omission of items were asked. Audio-recordings were transcribed and transcripts coded. Analysis drew on methods of constant comparison, where participants’ comments were continually compared to identify common themes. Findings were discussed with the research team, who decided whether ModRUM should be adapted or the suggestion should be explored with patients. Ten health economists from the United Kingdom were interviewed (mean duration: 44 minutes). Common themes emerged for the core module, with most questions considered adequate. Feedback on depth questions varied on whether more, less or alternative detail should be captured. For example, many thought it was inadequate to group community-based non-GP healthcare professionals in one question. ModRUM was revised to incorporate suggested changes agreed by the research team, including for example, an option to pre-specify numerous non-GP healthcare professionals. Our aim is to develop a well-validated, brief, generic RUM, which is needed to increase comparability of health economic research. In previous work, we used consensus methods to identify items and develop ModRUM. In this work, we used qualitative methods to enhance the usability and validity of ModRUM. Our findings in this study suggest that we are closer to achieving a valid measure.
Read full abstract