A personhood and citizenship training workshop for care home staff to potentially increase wellbeing of residents with dementia: intervention development and feasibility testing of a cluster randomised controlled trial

Abstract

BackgroundIn the UK, one third of people with dementia live in residential care homes, a sector where high staff turnover negatively affects continuity of care. To examine the effect of including personhood and citizenship principles in training, interventions need to be robustly tested, with outcomes relevant to residents with dementia.MethodsPhase one intervention development: The training intervention (PERSONABLE) comprised five reflective exercises facilitated by a mental health nurse/researcher. PERSONABLE was informed by four focus groups, and one field exercise, consisting of care home staff and family members. Phase two feasibility testing: Participants were (i) care home residents with dementia and (ii) care home staff working in any role. After baseline measurements, care homes were randomly allocated to (i) staff receiving PERSONABLE training or (ii) training as usual. Feasibility outcomes were the recruitment and attrition of care homes, residents and staff members (measured ten weeks between randomisation and follow-up), the acceptability of the training intervention PERSONABLE, and acceptability of outcome measures. The care home environment was evaluated, at baseline, using the Therapeutic Environment Screening Survey for Residential Care Homes. Measurements conducted at baseline and follow-up were resident wellbeing (Dementia Care Mapping™), staff knowledge of and confidence with personhood and citizenship (Personhood in Dementia Questionnaire and a perceived ability to care visual analogue scale). Inter-rater agreement for Dementia Care Mapping™ was undertaken at follow-up in one intervention and one training as a usual care home.ResultsPhase one: The developed reflective approach to the PERSONABLE exercises appeared to give staff a holistic understanding of residents living with dementia, seeing them as autonomous people rather than reductively as persons with a condition. Phase two: Six care homes, 40 residents and 118 staff were recruited. Four residents were lost to follow-up. Twenty-nine staff in the PERSONABLE arm of the study received the training intervention. In the PERSONABLE arm, 26 staff completed both baseline and follow-up measurements compared to 21 in the training as the usual arm. The most common reason for the loss to follow-up of staff was leaving employment. For the outcome measure Dementia Care Mapping™, the proportion of overall agreement between the two observers was 18.6%. High attrition of staff occurred in those homes undergoing leadership changes.ConclusionWith the right approach, it is possible to achieve good engagement during trial recruitment and intervention delivery of care home managers, staff and residents. Organisational changes are a less controllable aspect of trials but having a visible researcher presence during data collection helps to capitalise the engagement of those staff remaining in employment. Tailored, brief and flexible training interventions encourage staff participation. Simplification of study methods helps promote and retain sufficient staff in a definitive randomised controlled trial. This study found that some components of Dementia Care Mapping™ work effectively as an outcome measure. However, inter-rater reliability was poor, and the practical implementation of the measurement would need a great deal of further refinement to accurately capture the effect of a training intervention if delivered across a large number of clusters. The Dementia Care Mapping™ measurement fidelity issue would be further complicated if using multiple different unacquainted observers.Trial registrationRegistered with the ISRCTN under the title: Does a dementia workshop, delivered to residential care home staff, improve the wellbeing of residents with dementia? Trial identifier: ISRCTN13641553. Registered: 30/05/2017 http://www.isrctn.com/ISRCTN13641553.