Faithful but flexible: Intervention fidelity in clinical trials of complex interventions in healthcare

Thesis

The focus of clinical trials is typically interventions’ efficacy, or whether they attain their desired outcomes. Comparatively less attention is focused on understanding how or why interventions succeed, or fail to attain, those outcomes. This may be particularly important in trials of complex interventions such as surgery or physiotherapy, which are multifaceted and often tailored to individual participants, providers, or settings, increasing the potential for variations in intervention delivery and effects.

The correspondence between the intervention that was planned and what was actually delivered in a trial is the intervention’s fidelity. Several benefits for high levels of intervention fidelity have been proposed. However, whether or how much fidelity influences clinical trials’ treatment effect estimates had not previously been determined. A lack of a uniform definition for fidelity and its key components may also hinder intervention delivery in clinical trials and the translation of evidence-based interventions to clinical practice.

The principle aim of this thesis was to investigate the effects of intervention fidelity on the results, interpretation, and appropriateness to change practice of clinical trials of complex interventions. Through systematic review, “Best-fit” framework synthesis, reliability study, meta-epidemiological study, and fidelity assessment in an on-going pragmatic randomised controlled trial (RCT), it addressed several important knowledge gaps for intervention fidelity in complex intervention clinical trials. It estimated the prevalence of fidelity monitoring and reporting in complex intervention RCTs. It contributed an empirically-based intervention fidelity framework and a reliable, internally consistent fidelity assessment checklist tailored to the unique needs of researchers who seek to assess clinical trials in rehabilitation. It also estimated empirically, for the first time, the magnitude and direction of bias in treatment effect estimates arising from poor intervention fidelity in rehabilitation RCTs. Fidelity assessment in the ACL-SNNAP trial also provided new insights into the feasibility, applicability, and impact of monitoring intervention fidelity and participant adherence in complex intervention pragmatic trials.

The findings of this thesis showed that intervention fidelity is important, not just on a theoretical, but also on an empirical level. This thesis provides new and important information for the conduct and interpretation of clinical trials in rehabilitation and other complex interventions in healthcare.

Authors: Paez, A

• DOI: 10.5287/ora-jbo0rmakg
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: compliance; randomised clinical trials; treatment effect estimates; Intervention fidelity; surgery; Clinical trials; Adherence; bias; treatment fidelity or treatment adherence; evidence-based medicine; complex interventions; treatment effects; meta-epidemiological; rehabilitation
• Subjects: Clinical trials--Reporting; Evidence-based medicine; Clinical trials; Clinical epidemiology; Data integrity; Physical therapy; Surgery; Clinical trials--Statistical methods; Patient compliance; healthcare; Medical rehabilitation