The Department of Health's end of life care strategy published earlier this month pledged to allow more people with terminal illnesses to choose where they die. Clinicians at the Royal Marsden have made this possible through a pilot scheme

Around 500,000 people die in England each year. In London, there were 52,991 deaths in 2005.

However, patients' place of death does not often correspond with their preferences. Most people say they would prefer to die at home, yet the reality is that most deaths (58 per cent) occur in NHS hospitals. Only 18 per cent happen at home, 17 per cent in care homes, 4 per cent in hospices and 3 per cent elsewhere. At the Royal Marsden Hospital, of the 1,108 referrals to the palliative care team in 2005, 436 patients died in hospital (39 per cent).

An audit of urgent hospital admissions in 2004 showed 19.1 per cent were for pain and 54.7 per cent for other symptoms. The question was why were these symptoms not resolved by community professionals? The main reason was that there was a lack of careful planning to ensure safe transfer of patients from acute to community care, a lack of identified key workers and a lack of identified professionals responsible for individual aspects of care in the community.

The Royal Marsden Hospital2Home pilot project was set up to improve end of life care, decrease the death rate in hospital and increase the number of deaths at home.

Strategy for change

The pilot is funded by the Royal Marsden Charity and co-ordinates care for patients who are not receiving further active treatment. When a patient is referred to the programme, the programme's clinical nurse specialist assesses them and a managed care plan is drawn up. With the permission of the patient, the clinical nurse specialist then sets up a case conference in the patient's home. The patient's GP and all other community staff involved in the patient's care are invited. The conference is chaired by either the GP or the patient.

The agreed care plan is documented and distributed. At present, it is not possible to send this document electronically as independent/voluntary sector providers do not all have access to encrypted, secure network facilities.

The primary aims of the project are to allow more patients to die in their preferred place, give patients more choice by planning ahead, reduce the number of unnecessary acute hospital admissions and improve the quality of life of the terminally ill. Its secondary aims are to disseminate specialist palliative care skills and information that pertains to a particular patient to generalists, and to offer support to generalists through discussions of managed care plans.

Measuring improvement

So far, 88 case conferences have taken place and 60 patients have died. A far greater proportion of these than is usual have been able to die in their preferred location, at home or in a hospice rather than in hospital. This has been achieved through team working in which there are identified key workers, team members have defined responsibilities, and the patient is the central focus. Patients and their carers reported feeling "safe" and their satisfaction levels are high.

Place of death

Place of death

Numbers of patients

Patient's preferred place of care/death

How place of death for the pilot compares with national average



20 home + 1 hospice

Home (35% vs 18 %)▲



24 hospice + 7 home

Hospice (53% vs 4%)▲




Hospital (10% vs 58%)▼

Transit to hospice


1 hospice

Lessons for the future

The Hopsital2Home case conference model has been shown to decrease unnecessary hospital admissions and increase the number of patients being cared for and dying in their preferred place.

In addition, it has shown:

  • palliative care can be delivered by generalists if they are given sufficient information and training;

  • faxing the detailed document of the care plan is labour intensive, time consuming and not contemporaneous. The project highlights the need for end of life care to be part of the national IT programme. A palliative care entry in the summary care record in primary care could hold all essential patient information (such as do not resuscitate status, medications, plans for acute interventions);

  • an economic evaluation of the packages of care that includes carer burden is required.