Table 1 Detail of targeted and random assessment of fidelity drift
We pragmatically aimed to sample at least 20% of interventions for drift. Drift was assessed by sampling a total 213 of different types of visits (1st intervention visit, review visit or phase review visit). We had set criteria to inform the sampling for assessment of drift: | |
|---|---|
Targeted assessment | • All interventionists who failed any certification assessment (ie 1st intervention, 1st Review, 1st Phase review visit), • Interventionists with high withdrawal rates (more than two participants withdrawn from interventionist contact) • Interventionists with in-sufficient number of audio recordings (less than 80% audio recorded out of those who provided consent) • Interventionists where CFHealthHub data indicated a lower than expected number of visits completed, and/or action and coping plans created. It is noted that as action/coping plans are central to the delivery of this behaviour the inclusion of these was reviewed regularly and targeted retraining delivered even before drift was assessed formally |
Random assessment | • If the targeted assessment was less than 10% of all interventionist visits, random assessment of interventionists / sites took place in order to bring the total sample of assessments up to 10% of all interventionist visits |