Implementing the compassion intervention, a model for integrated care for people with advanced dementia towards the end of life in nursing homes: a naturalistic feasibility study

Background Many people with dementia die in nursing homes, but quality of care may be suboptimal. We developed the theory-driven 'Compassion Intervention' to enhance end-of-life care in advanced dementia. Objectives To (1) understand how the Intervention operated in nursing homes in different health economies; (2) collect preliminary outcome data and costs of an interdisciplinary care leader (ICL) to facilitate the Intervention; (3) check the Intervention caused no harm. Design A naturalistic feasibility study of Intervention implementation for 6 months. Settings Two nursing homes in northern London, UK. Participants Thirty residents with advanced dementia were assessed of whom nine were recruited for data collection; four of these residents' family members were interviewed. Twenty-eight nursing home and external healthcare professionals participated in interviews at 7 (n= 19), 11 (n= 19) and 15 months (n= 10). Intervention An ICL led two core Intervention components: (1) integrated, interdisciplinary assessment and care; (2) education and support for paid and family carers. Data collected Process and outcome data were collected. Symptoms were recorded monthly for recruited residents. Semistructured interviews were conducted at 7, 11 and 15 months with nursing home staff and external healthcare professionals and at 7 months with family carers. ICL hours were costed using Department of Health and Health Education England tariffs. Results Contextual differences were identified between sites: nursing home 2 had lower involvement with external healthcare services. Core components were implemented at both sites but multidisciplinary meetings were only established in nursing home 1. The Intervention prompted improvements in advance care planning, pain management and person-centred care; we observed no harm. Six-month ICL costs were 18 pound 255. Conclusions Implementation was feasible to differing degrees across sites, dependent on context. Our data inform future testing to identify the Intervention's effectiveness in improving end-of-life care in advanced dementia.