Facilitating implementation of research evidence (FIRE) : A randomised controlled trial and process evaluation of two models of facilitation informed by the promoting action on research implementation in health services (PARIHS) framework

Seers, Kate and Rycroft-Malone, Joanne and Harvey, Gill and Cox, Karen and Chricton, N.J. and Edwards, Rhiannon and Eldh, A.C. and Estabrooks, C. and Hawkes, Claire and Jones, Carys and Kitson, Alison and McCormack, Brendan and McMullan, C. and Mockford, Carole and Niessen, Theo and Slater, Paul and Titchen, A. and van der Zijpp, Teatske and Wallin, Lars (2019) Facilitating implementation of research evidence (FIRE) : A randomised controlled trial and process evaluation of two models of facilitation informed by the promoting action on research implementation in health services (PARIHS) framework. Implementation Science, 14 (Supp 1): S46. p. 18. ISSN 1748-5908

Abstract

Background: The PARIHS framework proposes that successful implementation of research evidence results from the complex interplay between the evidence to be implemented, the context of implementation and the facilitation processes employed. Facilitation is defined as a role (the facilitator) and a process (facilitation strategies/methods). Empirical evidence comparing different facilitation approaches is limited; this paper reports a trial of two different types of facilitation represented in the PARIHS framework. Methods: A pragmatic cluster randomised controlled trial with embedded process evaluation was undertaken in 24 long-term nursing care settings in four European countries. In each country, sites were randomly allocated to standard dissemination of urinary incontinence guideline recommendations and one of two types of external-internal facilitation, labelled Type A and B. Type A facilitation was a less resource intensive approach, underpinned by improvement methodology; Type B was a more intensive, emancipatory model of facilitation, informed by critical social science. The primary outcome was percentage documented compliance with guideline recommendations. Process evaluation was framed by realist methodology and involved quantitative and qualitative data collection from multiple sources. Findings: Quantitative data were obtained from reviews of 2313 records. Qualitative data included over 332 hours of observations of care; 39 hours observation of facilitation activity; 471 staff interviews; 174 resident interviews; 120 next of kin/carer interviews; and 125 stakeholder interviews. There were no significant differences in the primary outcome between study arms and all study arms improved over time. Process data revealed three core mechanisms that influenced the trajectory of the facilitation intervention: alignment of the facilitation approach to the needs and expectations of the internal facilitator and colleagues; engagement of internal facilitators and staff in attitude and action; and learning over time. Data from external facilitators demonstrated that the facilitation interventions did not work as planned, issues were cumulative and maintenance of fidelity was problematic. Implications for D&I Research: Evaluating an intervention - in this case facilitation - that is fluid and dynamic within the methodology of a randomised controlled trial is complex and challenging. For future studies, we suggest a theoretical approach to fidelity, with a focus on mechanisms, as opposed to dose and intensity of the intervention.