Table 6 Refinement of the programme theory for future implementation

(1) If the danceSing care resources are delivered to the care homes, the ACs would consistently deliver the programme to the residents

- The adherence was best when the programme was given two to three times/week out of the four sessions recommended. Therefore, future studies could implement two to three sessions/week

- For danceSing care staff and researchers, fewer than ten participating care homes would facilitate and speed-up problem solving of any unforeseen barriers regarding the resources, such as tech issues or staffing problems

(2) If the programme is delivered in the care homes, the residents would want to participate regularly

- Participation could be improved if the danceSing care resources are part of the regular care home schedule, with an allocated time slot, playing music at the start of the sessions and briefing and encouraging residents before the start

(3) If the ACs provide the programme consistently, the participants will experience improved psychosocial health markers. Changes in pre-and post-survey data and qualitative interviews would evidence this

- Questionnaires evidenced improvements in several multidimensional health outcomes of participants. Indicators of frailty should be complemented with physical function tests. More objective measures such as actigraphs or endocrine measurements could provide additional information about specific health parameters

(4) If the ACs were given enough organisational support (e.g., resources and time), they would be engaged in this programme. This would establish shared learning and co-production between programme developers, care homes, and researchers

- Wider contexts (e.g., the care home system in Scotland or the COVID restrictions) and organisational contexts (e.g., time and staffing) are out of our control. They are therefore not included in this programme refinement. However, before future implementation of the danceSing care resources takes place, tech issues (e.g., internet access or big screens) and staff issues (e.g., a dedicated AC with back-up) could be factored in

(5) If the group sessions were adequately and consistently used, residents would feel more engaged in group activities, creating a community. This communal feeling would increase self-confidence and quality of life and reduce loneliness. This could potentially inspire future usage of the programme's resources

- Some care home residents who participated had cognitive impairment, influencing the adherence and understanding of measurement tools. Considering cognitive status of care home residents, and exploring how to tailor the intervention to those lacking capacity to consent due to cognitive impairment could make findings more generalisable in a future intervention