Emergency department care transitions for persons living with cognitive impairment and care partners: development of a novel artificial intelligence and care coach intervention
James Galske, Heidi Gil, Chitra Dorai, Kate Keefe, Edith Stern, Erica DeFrancesco, Cameron Gettel- Psychiatry and Mental health
- Cellular and Molecular Neuroscience
- Geriatrics and Gerontology
- Neurology (clinical)
- Developmental Neuroscience
- Health Policy
- Epidemiology
Abstract
Background
Emergency department (ED) presentation by persons living with cognitive impairment (PLWCI) often represents a sentinel event in the care continuum, yet little data exists to date regarding effective interventions to improve ED‐to‐community care transitions for PLWCI and their care partners. Therefore, we sought to develop and refine the innovative pairing of artificial intelligence (AI)‐enabled digital advisors and an occupational therapist‐led care coach intervention to improve ED‐to‐community care transitions for PLWCI and their care partners.
Methods
We used a mixed methods, multi‐phased approach to develop the intervention, with PLWCI and care partners sampled from the LiveWell Dementia Specialists network. In the first phase, we developed and iteratively revised a codebook and conducted recorded, semi‐structured focus groups using a standardized guide. Two researchers coded the professionally transcribed data using a combined deductive and inductive approach and analyzed transcripts to identify dominant themes and representative quotations. In the second phase, we performed a 2‐day design thinking workshop using a user‐centered design approach, involving PLWCI, care partners, health care professionals, community‐based partner organizations, technologists, and AI experts. Integration of findings from the two phases provided takeaways for refinement and optimization of the intervention for eventual pilot testing.
Results
In the first phase, we performed two focus groups, involving a total of 16 participants, and generated empathy maps regarding the ED experience, the subsequent ED‐to‐community care transition, and the role of technology. PLWCI and care partners expressed: 1) concerns that their unique needs related to cognition were not addressed in the ED setting, 2) frustration with the ED discharge process and their perceived ability to follow‐up in the outpatient setting and obtain desired resources, and 3) openness to the incorporation of AI to facilitate successful ED‐to‐community care transitions. In the second phase, we identified perceived barriers and facilitators for intervention adoption during open forum discussions, depicted the current ‘As Is’ and the desired ‘To Be’ states regarding ED‐to‐community care transitions, and provided a demonstration of the AI‐enabled digital advisors and care coach intervention.
Conclusion
PLWCI and care partners were integrally engaged in the co‐creation of a two‐component novel intervention to improve ED‐to‐community care transitions.