Factors associated with transition from community settings to hospital as place of death for adults aged 75 years or older: a population-based mortality follow-back survey

Objective: To identify factors associated with end of life (EoL) transition from usual place of care to hospital as place of death for people aged 75 years or older (75+). Design: Population-based mortality follow-back survey. Setting: Deaths over six months in 2012 in two unitary authorities in England, covering 800 square miles with over one million residents. Participants: A random sample of people aged 75+ who died in a care home or hospital and all those who died at home or in a hospice unit. Cases were identified from death registrations. The person who registered the death (a relative for 98.9%) completed the survey. Measurements: Our main outcome was EoL transition to hospital as place of death versus no EoL transition to hospital. We used multivariable modified Poisson regression to examine factors (illness, demographic and environmental) related to EoL transition to hospital. Results: 443/882 (50.2%) responded, describing the care received by people who died from mostly non-malignant conditions (76.3%) at mean age 87.4 years (SD= 6.4). 32.3% transitioned to hospital and died there (n=146). Transition was more likely in respiratory disease compared to cancer (Prevalence Ratio [PR] =2.07, 95%CI 1.42- 3.01) and for people with severe breathlessness (PR=1.96, 95%CI 1.12-3.43). Transition was less likely if EoL preferences had been discussed with a healthcare professional (PR=0.60, 95%CI 0.42-0.88) and when there was a key healthcare professional (PR=0.74, 95%CI 0.58-0.95). Conclusion: To reduce EoL transition to hospital for older people this study suggests a need to improve the symptom management of breathlessness in the community and better access to a key healthcare professional skilled in coordinating care, communication, facilitating complex discussions and in planning for future care.