Background: Frailty is common in later life and can lead to adverse health outcomes. Services aimed at preventing decline in early stages of frailty may support older people to remain independent for longer. We developed and tested a new service, HomeHealth, in a randomised controlled trial (RCT). HomeHealth was a multi-domain behaviour change service based in the voluntary sector in England targeting mobility, socialising, nutrition, and psychological wellbeing. 

Objective: To describe the population reach, fidelity, acceptability, context, and mechanisms of impact of the HomeHealth service. 

Design and Methods: Mixed-methods process evaluation of a randomised trial. 

Setting and participants: HomeHealth trial participants (older people aged 65 years+ with mild frailty) and service providers.

Data sources and analysis: Population reach was evaluated through comparison to local census data. Fidelity of audio recorded appointments was assessed by two independent raters using a structured checklist. Using data from appointments attended, types of goals set and progress towards goals, we described appointment characteristics, goals and signposting, and evaluated three mechanisms of impact: 1) Effect of appointment attendance on independence, 2) Effect of goal progress upon independence, 3) Whether selecting a particular goal type led to improvements in the corresponding intermediate outcome. We thematically analysed qualitative interviews with 49 older people, seven HomeHealth workers and eight stakeholders to explore acceptability and context. 

Results: HomeHealth participants were similar with regards to deprivation, education, and housing status to the local older population, but with lower rates of minority ethnic groups. HomeHealth was delivered with good fidelity (81.7%) in voluntary sector organisations. Appointments were well attended (mean 5.33 out of the 6 intended), but attendance was not associated with better independence scores at 12 months (mean difference (MD) 1.29 (-8.20 to 10.78)). Participants varied in progress towards goals within appointments (mean progress 1.15/2.00), but greater goal progress was not associated with improved independence scores at 12 months (MD -0.40 (-2.38 to 1.58)). Mobility goals were most frequently selected (49%), but type of goal had no impact on independence and little impact on intermediate outcomes. Forty one percent were signposted or referred to other supportive services, with ongoing support where needed throughout this process. Qualitative data indicated that HomeHealth was acceptable, empowering for those who saw a need for change and fitted well within host voluntary sector organisations.   

Limitations: Census data was only available for all adults aged over 65 in local areas rather than a mildly frail population, who are likely to be older, female and less diverse, and therefore population reach calculations may be less accurate. Goal progress was assessed using a simple scale rather than a validated instrument. 

Conclusions: HomeHealth represents an acceptable and implementable intervention for older people with mild frailty, but may work via different mechanisms than those intended. 

Future work: Future work should explore how to best screen older people with mild frailty for readiness to change to maximise benefits from similar services and identify other possible mechanisms of effects. 

Study registration: ISRCTN54268283