Embedding Advance Care Planning in Hospital at Home for Older People With Frailty: A Quality Improvement Report

BackgroundAdvance care planning (ACP) is important for older people with frailty but is inconsistently delivered in Hospital at Home. Baseline staff survey and interview work identified low confidence, documentation uncertainty, time pressure, limited mentorship and organisational barriers.ObjectiveTo evaluate whether a theory-informed multicomponent strategy improved emergency care-planning documentation and nurse confidence, and whether the 30-day emergency department reattendance rate changed.MethodsA quality improvement project was implemented from June 2025 in a UK Hospital at Home service. Interventions included education, simulation, champions, peer mentorship, documentation redesign and co-design. Monthly Recommended Summary Plan for Emergency Care and Treatment (ReSPECT)/Do Not Attempt Cardiopulmonary Resuscitation (DNACPR) documentation and 30-day emergency department reattendance were analysed from January 2024 to June 2026 using run charts and baseline-centred statistical process control p-charts. Nurse confidence was assessed immediately before and after ACP training and analysed using a paired t-test.ResultsAcross 10 950 admissions, new ReSPECT/DNACPR documentation increased from 2.55% (160/6266) before intervention to 5.57% (261/4684) during intervention. Baseline-centred statistical process control showed an eight-point upward shift in documentation rate from November 20252 to June 2026. Nurse confidence increased from 4.71 to 7.43/10 (mean paired increase 2.71, 95% CI 1.86-3.57; P < .001). Thirty-day emergency department reattendance decreased from 22.25% to 21.26%, with an eight-point downward shift over the same period.ConclusionMulticomponent implementation was associated with improved emergency care-planning documentation and nurse confidence. A sustained downward shift in 30-day emergency department reattendance was also observed, although causal attribution is not possible in this uncontrolled service-level evaluation.