Qualitative research to inform hypothesis testing for fidelity-based sub-group analysis in clinical trials: lessons...

Ensuring fidelity to complex interventions is a challenge when conducting pragmatic randomised controlled trials. We explore fidelity through a qualitative process evaluation, which was conducted alongside a pragmatic, multicentre, two-arm cohort randomised controlled trial: the REFORM (Reducing Falls with Orthoses and a Multifaceted podiatry intervention) trial. The paper aims, through a qualitative process evaluation, to explore some of the factors that may have affected the delivery of the REFORM intervention and highlight how project-specific fidelity can be assessed using a truly mixed-methods approach when informed by qualitative insights.

Semi-structured qualitative interviews carried out as part of a process evaluation. Interviews were analysed thematically.

Seven NHS trusts in the UK and a University podiatry school in Ireland. Interviews were undertaken face-to-face or over the telephone.

Twenty-one REFORM trial participants and 14 podiatrists who delivered the REFORM intervention.

Factors affecting fidelity included: how similar the intervention was to routine practice; the challenges of delivering a multifaceted intervention to a heterogeneous older population; and practical issues with delivery such as time and training. Trial participants' views of the intervention, whether falls prevention is a personal priority, their experience of being part of a trial and individual factors such as medical conditions may also have affected intervention fidelity.

Our process evaluation highlighted factors that were perceived to have affected the fidelity of the REFORM intervention and in doing so demonstrates the importance of considering fidelity when designing and evaluating pragmatic trials. We propose a number of recommendations of how important project-specific insights from qualitative work can be incorporated into the design of fidelity measurement of future trials, which build on existing conceptual fidelity frameworks. In particular, we encourage adopting a mixed-methods approach whereby qualitative insights can be used to suggest ways to enhance quantitative data collection facilitating integration through hypothesis generation, hypothesis testing and seeking explanation for trial findings. This will provide a framework of enabling measures of fidelity to be incorporated into the understanding of trial results which has been relatively neglected by existing literature.

ISRCTN Registry: ISRCTN68240461. Registered on 01/07/2011.