As I walked beside Clapham Common towards Trinity Hospice a famous BMJ phrase was ringing in my ears: “Hospice care is deluxe dying for the few looked after by dowager duchesses.” That may not be the actual quote, but the point was that hospices, funded mostly by charity, could provide an excellent death while most NHS patients missed out. But why should the NHS be willing to fund bad expensive deaths in hospitals, but not excellent deaths cared for by (not necessarily in) hospices?

Trinity Hospice is the oldest in London, founded in 1891 as the Hostel of God. Originally patients were cared for by the Sisters of St James’ Servants of the Poor, but later the Sisters of St Margaret’s of took over. The nuns left long ago. Part of the striking building was built by Sir Charles Barry, the architect who designed the Palace of Westminster, as a home for his family. The extensive gardens are the best in Clapham, and supposedly Samuel Pepys, who died in Clapham, used to wander there. Anybody can go and enjoy the gardens and the café, but mostly people don’t want to enter the “house of death” until they have to—or maybe they just don’t know about the gardens and café.

I visited the wards when I was last at the hospice, and the 28 large and modern rooms look out onto the gardens through huge windows. If you have to die in an institution, as some of us will, then it’s hard to think of a better place. Your extensive family, if you have one, could gather around, watch over you, and see the sun rise and sink among the delphiniums.

But this time I’m here to understand how the hospice keeps going and think about its future. Dallas Pounds, a nurse by background, took over recently as chief executive and is changing the culture and finding a path for the hospice in uncertain times. She is confident that the clinical care is excellent; it’s the whole enterprise she has to think about.

In broad terms the hospice costs about £10m a year with a third coming from the NHS, a third from fund raising, and a third from its 24 shops. The hospice has a catchment area, agreed with the many other hospices in London, that covers all of Wandsworth, some of surrounding boroughs south of the river, and also some of Kensington, Chelsea, Hammersmith, Fulham, and Westminster north of the river. Much of central London is covered, and the Queen and the Prime Minister are both within the catchment area. Altogether the hospice covers three quarters of a million people, which I estimate must mean just over 10 000 deaths a year. The hospice doesn’t know exactly how many people who die are influenced by its care (and it’s not only about the dying), but it’s probably less than 10% of all the deaths.

The hospice deals with six clinical commissioning groups with Lambeth CCG as the lead. The CCGs are busy trying to sort themselves out, but Dallas expects that just like primary care trusts and their predecessors some will understand the value of end of life care more than others. The CCGs, I estimate, are probably spending around 2% of their budget on hospice care. As they organise themselves other bigger expenditures will be their priorities.

The income from the NHS has drifted down in real terms, says David Coggins, the finance director of the hospice. I ask how the NHS decides what it will pay. The answer is not clear, but it seems—surprise, surprise—that it tends to pay in cash what it paid the year before, meaning a fall in real terms. In short, nobody in the NHS thinks about it much. But will they, I ask, get round eventually to commissioning outcomes? They will, answers Dallas, but outcomes are hard to define in palliative care and the CCGs have a lot to think about.

Sally Bateson is in charge of fundraising, a vital part of the organisation. She and her team of 14 raised £3.6 million in 2013 and must raise at least that again in 2014. Sally is actually employed by Compton Funding, the UK’s leading fundraising consultancy. Hospice fundraising is a growing business for Compton. Around a third of the money comes from legacies, natural “feeding ground” for hospices I reflect. “We have a great case for support,” says Sally. The rest of the money comes from foundations, “high net worth individuals” (as the Americans call them), and events. There is a lucrative ball every two years, but there are also events like street parties. The events can take up a lot of the fundraisers’ time, but they are good for building relations with the community. The street I live in, a few hundred yards from the hospice held a street party.

Stefan Karas manages the 24 shops which raise around another £3 million each year. Immaculately and stylishly dressed, Stefan will probably like it if I describe him as “hard boiled.” He’s a professional, unsentimental retailer and expects his shops to be as good as any other shop on the high street. His competitors are not other charity shops but chains like T K Max. About 15% of his stock is designer clothes, and he gets the best price he possibly can. Prices are steadily rising in charity shops as they professionalise. Each shop has an employed manager who has a target and receives a bonus if he or she reaches the target. The shops are staffed by volunteers, but they are expected to be as professional as employed staff in other shops. The income from the shops has risen considerably, and Stefan is aiming for 30 shops. He’s very particular about sites: a few yards can make all the difference, he says. “Retail is detail,” he adds. As another example of detail he regrets clothes that are being discarded that could be steamed and sold for a good price: if I went through every bag, he says, I’d find those clothes.

Then I meet the four managers of the clinical services. Cathy Maylin runs the community services. She has 16 nurses and at the moment 600 patients on her books. She classifies her patients into three broad groups: those who have complex needs; the “simply dying”; and the “not ready to die, palliative care patients.” More than half of the patients are those with complex needs where specialist nurses are much needed. The “simply dying” usually need just one or two visits. It sounds to me as if the “not ready to die palliative patients” are a bit of a nuisance in that they are dying but haven’t accepted it—or their doctors haven’t. I probably misinterpreted that. Almost all of the patients cared for in the community die in their homes or in care homes. Those in care homes usually die easily, fading away, says Cathy.

Mel Johnson looks after what might be described as the heart of the operation, the inpatients. There are more than 50 nurses and four palliative care physicians. The physicians are all part time and female, and  two are shared with acute hospitals. Recruiting senior nurses is hard, but the hospice has a bank of nurses, which means that it rarely needs to employ agency nurses. The hospice runs at a bed occupancy of 82%, which is evidently standard for hospices. Dallas would like the occupancy to edge up to 85%.

Louise Malone manages the day or outpatient service. Until recently no doctors were involved, but nursing, physiotherapy, occupational therapy, and complementary therapy services are provided. When asked what they want, the patients list, in order, social support, physiotherapy, and complementary therapy. Louise has 120 patients at the moment, some of them inpatients, but she has an emerging vision of a “one stop shop” that meets all the needs of patients at the end of their lives. This is a part of the service where it’s hard to prove the value to commissioners, and GPs are likely to be particularly sceptical about paying for complementary therapy.

The service managed by Letizia Forrest is called patient and family support and is provided by a rare breed, “clinical social workers.” The clinical means mostly training in mental health, and they “follow NICE guidelines” in providing social and psychological support to almost all of the inpatients, some of those cared for in the community, and a few direct referrals. Letizia and a colleague also run Schwarz Rounds, monthly meetings that all staff can attend and where patients who have presented difficulties or particular issues are discussed in a facilitated meeting. The next meeting is on “hope.” Usually around 30 people attend, and everybody I asked thinks them very helpful and contributing to changing the culture.

There is also a bereavement service, but I didn’t manage to meet the manager of that service.

Just as the hospice depends heavily on fundraising and its shops so it depends on volunteers. Around 80 volunteers work in the building, 30 in the community, and 350 in the shops. Dallas likes to call them non-paid staff and make sure that they have specific roles and are well trained. She doesn’t want people drifting around the wards with no clear role. The CCGs tend, she says, to think of the volunteers as a free good—but they are not: managing and training them requires paid resource.

The figures are rough, but about £2 million is spent on community services and £8 million in the services in the building, not just the clinical but also the support services. Could it be the other way round, I ask. The answer is that it possibly could, but clearly it would need a long term strategy. It would also mean shrinking the building, and the hospice does have a proposal to convert the Barry House to flats. (I like the idea of moving there, but my wife says she doesn’t want to live in a hospice.)

Everybody I know who has had a relative or friend die in Trinity Hospice praises the services, and those relatives and friends are important for the fundraising. But the price per patient does seem to be more than the NHS is willing to pay. But is that right? Most people would much prefer to die under the care of a hospice (not necessarily in the hospice) rather than in intensive care, but the NHS doesn’t hesitate to pay for intensive care, even though it is more expensive than hospice care and probably a good bit of it futile. Can this be changed? If I was running a CCG it would be high on my agenda because those last months of life tend to be expensive and often poor: quality could be improved and money saved, the “win win” that commissioners seek.

Or could the hospice find a way to spread its services to more people, bringing down the cost per patient but not compromising on the quality? The answer is clearly not to have ten times as many beds in the building. The answer lies in the community.

Competing interest: RS was asked to visit the hospice after speaking at the Help the Hospices conference, and he will be contributing to the away day of the board. He will not be paid but likes the idea of being cared for by the hospice when he’s dying.

Richard Smith was the editor of the BMJ until 2004 and is director of the United Health Group’s chronic disease initiative.