Health and care service utilisation in the last year of life before non-sudden death in Wales, 2014-2023, by palliative care registration: a population-based retrospective cohort study.
Rhiannon K Owen, Rowena Bailey, Helen Daniels, Athena McBride, Ashley Akbari, Elinor Curnow, Alison Cooper, Natalie Joseph-Williams, Adrian Edwards, Maria Parry, Idris Baker
Abstract
Open AccessBackground: End-of-life health and care service provision are complex processes. We aimed to quantify the uptake of health and care services in the last year of life before death from non-sudden causes by palliative care registration. Methods: Population-scale linked administrative and health data in the last year of life for Welsh residents who died of non-sudden causes were modelled using multi-state models between 2014 and 2023. Cox regression were used to estimate hazards for transitions between care settings, including people's homes, care homes with and without nursing, emergency, elective and other hospital admissions, and death. The primary outcome was rate of transition reported as hazard ratios (HR) with corresponding 95% confidence intervals (95% CI) adjusted for age, sex, rurality, area-level deprivation, and palliative care registration. Secondary outcomes included expected length of stay (ELOS). Findings: Our analyses included 267,199 individuals, with 1,845,572 transitions. There were 74,045 (27.7%) individuals registered for palliative care, under-represented groups included men, most-deprived and living alone. Most time was spent at home, with 370,752 (90.3%) of 410,441 emergency admissions from home. There was a 23% (HR 1.23 [95% CI 1.22-1.25]) increased transition rate of emergency admissions from home for palliative care registered compared with unregistered individuals, with a decreased expected length of stay (ELOS 25.34 [95% CI 25.34-25.34] vs 26.87 [26.87-26.87]). Emergency admissions from care homes with and without nursing were 17% (HR 0.83 [95% CI 0.80-0.86]) and 18% (HR 0.82 [95% CI 0.79-0.85]) lower for palliative care registered compared with unregistered individuals, with an increased rate of discharge from emergency hospital settings (HR 2.00 [95% CI 1.92-2.09] and 1.62 [1.54-1.69]). Interpretation: Palliative care status was associated with health and care utilisation at the end-of-life. Efficient identification of individuals needing palliative care and additional support at home should be prioritised for system optimisation. Funding: Health and Care Research Wales Evidence Centre.