Hospital admissions for patients close to the end of life are considered ‘inappropriate’ in contemporary English health policy. Hospitals are supposedly unable to offer a ‘good’ death for patients, and dying there is thought to contradict patient choice, since patients are assumed to want to die at home. However, almost half of all deaths in England in 2015 occurred in hospital, and of these, nearly a third died within three days of admission. This thesis seeks to explore why these admissions are considered to be a problem and how they occur.

Through a systematic review of UK literature I found that it cannot be stated that most patients want to die at home, because of the extent of missing data (preferences not asked, expressed, reported or absent). This finding challenges the justification that admissions are inappropriate because they contravene patient choice. Similarly inconclusive evidence about the undesirability, cost, and lack of need for patients to be in hospital were also found in a review of policy. Together with analysis of historical trends in hospital and hospice provision, it is apparent that attitudes towards end-of-life admissions reflect existing tensions about the role of hospital as an acute provider, and as a place of death.

An analysis of interviews conducted with healthcare staff and next-of-kin involved in the admission of patients (case-patients) who died shortly after being admitted to Meadowbridge, a large English hospital explored these tensions further. I found that whilst hospital was not recognised as a place where ‘good’ deaths typically occurred, it was acknowledged as an emergency place of care. In this context, patients without obvious need for hospital care were nevertheless admitted to the hospital and the environment was subsequently recognised to offer distinct benefits. The need for emergency care reflected the difficulties of providing end-of-life care in the community. For dying to occur appropriately, home had to be adapted and care organised by healthcare staff. Both tasks were complicated by the unpredictability of dying, and family carers helped to absorb much of the uncertainty and support patients to die at home. Ambulance staff became involved when patients had care needs that exceeded care quickly and easily available in the community. When called to the case-patients, ambulance staff instituted familiar practices in transferring them to hospital. Hospital was recognised as a default place of care because ambulance staff struggled to facilitate alternative care and lacked sufficient professional authority to keep patients at home.

The admissions of the case-patients represent the best attempts of staff to navigate the tangled practices of end-of-life care. These practices are the result of the actions of the staff, which in turn both constrained and enabled their action in providing care to patients. The term ‘inappropriate’ to describe admissions does not encompass these attempts, and moreover, devalued the significant care provided by healthcare staff in the community and hospital.