Process Evaluation of a community-based physical activity intervention to prevent mobility-related disability in retired older adults: The REtirement in ACTion (REACT) Study

Abstract

The REACT Study assesses the effectiveness of a community-based PA intervention designed to reduce the progression of mobility-related disability in older people (≥65 years). The aim of this thesis was to conduct a mixed-methods process evaluation of the REACT intervention to understand how it was implemented, the causal mechanisms involved and the impact of contextual variables that led to intervention outcomes. Systematic review methods were used to determine that there is low utilisation of process evaluations (53.6%) in PA trials targeting older adults and a lack of focus on intervention fidelity and qualitative process evaluation of mechanisms of impact. Multi-observer coding of audio recordings (n=25) of intervention sessions were used to assess the delivery fidelity of REACT BCTs. A modified Dreyfus Scale (score 0 to 5) was used to assess delivery fidelity, with competent delivery fidelity defined as a score of ≥3. Mean intervention delivery fidelity approached competence (2.5, SD=0.45), with variation between BCTs and intervention groups. Framework analysis of semi-structured interviews was used to explore participant experiences of the REACT intervention longitudinally at six-and-12-month interviews. Perceived benefits (physical, social, emotional and behavioural) moderated REACT attendance and PA outcomes (PA and/or exercise independent of REACT). Perceptions of autonomy, competence/ self-efficacy and relatedness mediated both REACT attendance and PA outcomes. Enjoyment, a supportive facilitator and supportive group dynamics were key motivators of REACT attendance. Physical health, mental health and time were key barriers. Frail participants, scoring between 4-7 on the Short Physical Performance Battery (SPPB) experienced fewer perceived benefits and more barriers to REACT attendance and PA compared to pre-frail peers with SPPB scores between 8-12. There is a lack of process evaluation being conducted in PA trials targeting adults, which this thesis aimed to address. Improvements are needed to enhance delivery fidelity in future implementation of the REACT intervention. Exercise that is enjoyable and utilises social interaction to facilitate a shared identity around PA motivates participation. Process evaluations whilst vital for interpreting intervention outcomes raise pragmatic challenges that researchers need to navigate to successfully utilise this methodology.

Date of Award	14 Feb 2022
Original language	English
Awarding Institution	University of Bath
Supervisor	Fiona Gillison (Supervisor) & Emma Solomon-Moore (Supervisor)