Abstract
BACKGROUND: Interpreting data about intervention effectiveness requires an understanding of which intervention components were delivered and whether they were delivered as planned (fidelity of delivery). These studies aimed to develop a reliable measure for assessing fidelity of delivery of the Community Occupational Therapy in Dementia-UK intervention (COTiD-UK) (Study 1) and measure fidelity of delivery of COTiD-UK across sessions, sites and occupational therapists (Study 2). METHODS: The studies used a longitudinal observational design nested within a multi-site randomised controlled trial. Where practicable, all intervention sessions were audio-recorded. Fidelity checklists and coding guidelines were developed, piloted and refined until good agreement was achieved between two coders. Ten percent of sessions were purposively sampled from 12 sites and 31 occupational therapists. Transcripts were coded using checklists developed in Study 1; 10% of sets of intervention session transcripts were double coded to ensure that agreement was maintained. Percentages of components that were delivered were calculated for each session, site and occupational therapist. RESULTS: A reliable measure of fidelity of delivery for COTiD-UK was developed after several rounds of piloting and amendments. COTiD-UK was delivered with moderate fidelity across all six sessions (range: 52.4-75.5%). The mean range of fidelity varied across sites (26.7-91.2%) and occupational therapists (26.7-94.1%). CONCLUSIONS: A reliable, systematic method for measuring fidelity of delivery of COTiD-UK was developed and applied, and can be adapted for use in similar interventions. As COTiD-UK was delivered with moderate fidelity, there is a reasonable degree of confidence that intervention effects were attributable to COTiD-UK.