Back

Benefit take-up in the last year of life: a population-based study using linked data for England and Wales

Davies, J. M.; Fairs, A.; Ayoubkhani, D.; Marshall, S.; Diggle, M.; Bradshaw, A.; French, M.; Stone, J.; Hussain, J.; Fimister, G.; Harding, R.; Sleeman, K.; Nafilyan, V.

2026-04-11 palliative medicine
10.64898/2026.04.10.26350614 medRxiv
Show abstract

ContextIn the UK, and in other countries, people living with a terminal illness are eligible for financial support to help with the costs of serious illness and to support their dignity and independence. This study investigates the take-up of benefits in the last year of life and identifies sociodemographic, clinical, and geographical factors associated with underclaiming. MethodsRetrospective cohort study using linked mortality, Census and benefits data for all people who died aged 16+ from chronic illnesses in England and Wales between 1 May 2018 and 30 April 2021. Outcome was receipt of non-means tested disability benefits in the last 12 months of life. We describe geographical variation in take up, and association with sociodemographic, clinical and geographical exposures using Poisson models. FindingsOur population included 1,049,493 eligible decedents, with an overall take-up rate of 65.9%. After adjusting for sociodemographic factors, variation in take-up by cause of death was wide: liver disease 44% (95% CI 43-45%), heart failure 52% (51-52%), cancer 62% (61-62%), dementia 75% (74-75%), and neurodegenerative diseases 90% (88-91%). Across Local Authorities, the age-and-sex-standardised take-up varied from 53% to 78%; rates were generally higher in more deprived areas, but not uniformly. ConclusionsIn England and Wales, 1 in 3 people who die from expected causes (120,000 each year) do not receive the benefits for which they are eligible. Our analysis uses novel data linkages and highlights clinical and sociodemographic groups and geographical areas that could be targeted with proactive take-up initiatives.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.