Process evaluation with cost analysis of the Move for Life cluster randomised feasibility trial for inactive adults aged 50 years and older

Enrique Garcia Bengoechea^1*Ciaran Doyle¹Amanda M. Clifford²Anna Hobbins³Liam Glynn⁴Andrew O'Regan⁴Chloe Forte⁵Andrew W. Murphy⁶Paddy Gillespie³Catherine B. Woods^1*

• ¹Department of Physical Education and Sport Sciences, Faculty of Education and Health Sciences, University of Limerick, Limerick, Ireland
• ²University of Limerick Faculty of Education and Health Sciences, Limerick, Ireland
• ³University of Galway J E Cairnes School of Business and Economics, Galway, Ireland
• ⁴University of Limerick School of Medicine, Limerick, Ireland
• ⁵University of Bristol Medical School, Bristol, United Kingdom
• ⁶University of Galway School of Medicine, Galway, Ireland

Background: An ageing population combined with a decline in physical activity with age have prompted calls for investment in physical activity programmes and services for older people. The Move for Life (MFL) intervention aims to augment existing community-based physical activity programs for middle-to-older aged adults in Ireland using behaviour change strategies and peer support. The intervention was found to be feasible and has the potential to impact desired outcomes in intended ways. This study reports on the process evaluation of the MFL cluster randomised feasibility trial. Methods: The process evaluation used a sequential explanatory mixed methods design. Guided by the Medical Research Council framework, the evaluation sought to uncover links between implementation variables, mechanisms of impact and context of the intervention. The evaluation also addressed the dimensions of sustainability and cost of the intervention, as recommended in the RE-AIM framework. Results: While generally the trial reached the target audience, recruitment of males and individuals from hard-to-reach groups remained challenging. The evaluation revealed potential mechanisms underlying the effects of the intervention, notably improved attitudes toward physical activity, more positive appraisal of physical activity programmes and instructors, enhanced interpersonal interaction, greater social support, and higher use of individual and group behaviour change techniques. Regarding sustainability, MFL showed potential to strengthen the relationship between individuals and their Local Sports Partnership. The MFL intervention was associated with a statistically insignificant increase in mean cost relative to usual provision, and a statistically significant increase in mean cost relative to the control arm. While the evaluation supports the original MFL intervention logic model, it also suggests three major areas for improvement. These are the need to reconsider the use of the participant handbook for the delivery of behavioural skills as well as the number of behavioural change strategies included, particularly for shorter duration programme formats, and better integrate the training of peer mentors and physical activity programme instructors. Conclusions: The findings highlight the essential role of process evaluation to inform intervention design to facilitate optimal implementation and illustrate the value of considering several perspectives in understanding, collecting and using cost estimates from interventions.