A population-based retrospective cohort study of end-of-life emergency department visits by people with dementia: multilevel modelling of individual- and service-level factors using linked data

BACKGROUND: emergency department (ED) visits have inherent risks for people with dementia yet increase towards the end-of-life. Although some individual-level determinants of ED visits have been identified, little is known about service-level determinants. OBJECTIVE: to examine individual- and service-level factors associated with ED visits by people with dementia in the last year of life. METHODS: retrospective cohort study using hospital administrative and mortality data at the individual-level, linked to health and social care service data at the area-level across England. The primary outcome was number of ED visits in the last year of life. Subjects were decedents with dementia recorded on the death certificate, with at least one hospital contact in the last 3 years of life. RESULTS: of 74,486 decedents (60.5% women; mean age 87.1 years (standard deviation: 7.1)), 82.6% had at least one ED visit in their last year of life. Factors associated with more ED visits included: South Asian ethnicity (incidence rate ratio (IRR) 1.07, 95% confidence interval (CI) 1.02–1.13), chronic respiratory disease as the underlying cause of death (IRR 1.17, 95% CI 1.14–1.20) and urban residence (IRR 1.06, 95% CI 1.04–1.08). Higher socioeconomic position (IRR 0.92, 95% CI 0.90–0.94) and areas with higher numbers of nursing home beds (IRR 0.85, 95% CI 0.78–0.93)—but not residential home beds—were associated with fewer ED visits at the end-of-life. CONCLUSIONS: the value of nursing home care in supporting people dying with dementia to stay in their preferred place of care must be recognised, and investment in nursing home bed capacity prioritised.