Where People With Cancer Die, and Where They Would Choose To

Asked where they would want to die, most people who give an answer, including most people with cancer, say home. But the share who say so varies widely, some people prefer a hospice or hospital, and some change their minds as illness progresses. Where people with cancer actually die varies even more between countries, and the research links dying at home most consistently with practical support: palliative care at home, community nurses and doctors who visit, and family who are able to be there.

Place of death is not, on its own, a measure of a good death. A hospice bed or a hospital ward can be exactly the right place for one person, and home the right place for another. What matters is whether people have a real choice, whether that choice is heard, and whether the help exists to carry it out. This study, written mainly for hospices and palliative care services, looks at what surveys have found about preferences, what death records show about where people with cancer die, and what research has linked to dying at home. It goes deeper on place of death than our companion study on end-of-life and palliative care for people with cancer, which also covers global access to palliative care and pain relief.

Most people say home, but not everyone, and preferences can move

The most thorough review of preferences pooled 210 studies covering just over 100,000 people in 33 countries. It found moderate evidence that most people prefer to die at home: that was the finding in 75% of studies. But the range was wide. Among the higher-quality studies, excluding outliers, the share of patients preferring home ran from 31% to 87%; for family caregivers it was 25% to 64%, and for the general public 49% to 70%. In the 10 studies that asked the same patients more than once, 20% of 1,395 people changed their preference, although the review notes this change was not tested statistically. Its authors urged policy-makers to secure hospice and palliative care elsewhere "for those who think differently or change their mind"^[1].

Who is asked, and how, shapes the answer. The table below sets out three very different kinds of survey.

Each method has limits. A healthy member of the public imagining advanced cancer is answering a different question from someone who is dying, and a relative reporting a wish after a death may not know it, or may remember it through grief. The UK review found that studies often dropped people whose preference was never recorded, and warned that "caution should be exercised if asserting that most patients prefer to die at home"^[5]. A small 2026 qualitative study that followed 14 people with life-threatening illness in the Netherlands, Portugal, Uganda and the United States over time, eight of them until death, found home was still the most preferred place, but that other places came to the fore because of the strain of receiving care, the anticipated trauma of a death at home, and urgent care needs^[6]. A preference recorded once, early on, is a starting point for conversation, not a fixed target.

Where people with cancer actually die varies several-fold between countries

Death certificates record a place of death for almost everyone, which makes them the most complete evidence we have. They show how far apart countries are. Across 12 Latin American countries, 491,929 deaths in 2016 to 2018 had cancer as the underlying cause. On average 31.1% happened at home, ranging from 14.9% in Brazil to 81% in Guatemala. Home deaths were more common among older people, people in rural areas and people with less formal education, and less common in countries with more hospital beds and doctors, wider universal health coverage and higher health spending^[7]. A high share of home deaths can therefore reflect choice, or a lack of other options, and death records alone cannot say which.

In Sweden, among all adults who died of cancer between 2013 and 2019, hospital was the most common place of death (48.7%), followed by nursing homes (25.6%) and the person's own home (23.5%). Over those years home deaths rose from 21.8% to 24.7% and hospital deaths fell from 49.2% to 47.1%. The study tested whether national palliative care policy had changed where people died; the authors reported that "the hypothesis was not confirmed", that trends varied between regions, and that the results "point to inequity in palliative cancer care"^[8].

Our companion study sets out a 14-country comparison from 2008 and more recent trends in the United States, England and Scotland, where cancer deaths have been moving out of hospital towards home and, in the United States, hospice facilities. We do not repeat those figures here.

Comparing countries needs care. A 2026 survey of 16 researchers who had analysed place-of-death data from 67 countries notes "substantial international variation in how it is recorded", and found ambiguous or inconsistent terms and a "lack of detailed categories for relevant settings (e.g. hospice, nursing home)"^[9]. Where those categories are missing, deaths in a hospice or care home cannot be told apart from deaths elsewhere. The figures above come from different registers and periods, so they show the size of the differences, not a league table.

Many people do die where they hoped, and support is linked to whether they do

The question that matters most to families is whether people die where they wanted to. That is harder to measure, because it needs a preference recorded before death and the place recorded after it. A systematic review of 18 studies found that the share of people who died in their preferred place (often called congruence) ranged from 30% to 91%. It was highest, 59% to 91%, in the eight studies of people receiving specialised home care; two studies of hospital inpatients found 43% to 46%, and the lowest figure, 30%, came from a study that asked where people would want to die "if everything were possible". Across several studies, physician support, hospice enrolment and family support were linked to people dying where they wished^[10]. The review was not specific to cancer.

Preferences are not always for home. In a prospective study of 2,638 people with cancer in 37 inpatient palliative care units in Japan, South Korea and Taiwan, published in 2025, 70% to 80% died in the place they had said they preferred; the availability of end-of-life care was the most common reason people gave for preferring, and dying in, a palliative care unit; even so, 13% to 22% of those admitted still preferred home^[11]. These were people already in specialist inpatient care, so the figures describe that group, not everyone with cancer in those countries.

The gap between preference and outcome is also a gap in data. England's national survey of bereaved people reported preferences and places of death, and asked whether the person "died in the right place", but its published bulletin does not cross-tabulate preference and place, so it cannot say what share of people who wanted to die at home did so^[4].

Dying at home is linked to home care, visiting clinicians and carers who can be there

A 2006 systematic review of 58 studies, covering over 1.5 million people in 13 countries, in studies where at least 80% of participants had cancer, found six factors strongly associated with dying at home: low functional status, the person's own preference, receiving home care, how intensive that home care was, living with relatives, and support from extended family^[12]. Most of these are about the support around a person, not the illness itself.

A London study gives a closer view. Relatives of 352 people who died of cancer in 2009 and 2010, 177 in hospital and 175 at home, completed a questionnaire (a 40% response). Four linked factors explained over 91% of home deaths: the person's preference, their relative's preference, home palliative care, and district or community nursing. People who had three or more home visits from a GP had about six times the odds of dying at home (adjusted odds ratio 6.27, compared with one visit or none), and where a relative was able to take more than 14 days off work the adjusted odds ratio was 9.94, compared with fewer than four days. These are wide estimates from one study. Relatives reported that people who died at home had more peace in their last week, with no significant difference in pain, and after adjustment the relatives' grief was somewhat less intense^[13]. That is an association reported by bereaved relatives, not proof that home is better for everyone. It does show how much a death at home rests on family members, and on whether their work and finances allow them to be present; our study on carers and families looks at that strain.

The strongest evidence on services comes from trials. A Cochrane review of 23 studies found that home palliative care more than doubled the odds of dying at home (odds ratio 2.21, from a meta-analysis of seven trials with 1,222 participants). It also found small but statistically significant reductions in symptoms, which its authors say applied in particular to people with cancer, no effect on caregiver grief, and inconclusive evidence on cost-effectiveness^[2].

One model studied is a community team that can be reached at any hour. In Ontario, 3,109 people who received care from community specialist palliative care teams in 2009 to 2011, about 80% of them with cancer, were compared with 3,109 similar people receiving usual care. The teams cared for people at home and aimed to be available "regardless of time or day". Fewer people cared for by the teams died in hospital: 16.2%, against 28.6%, a relative risk of 0.46^[14]. In England's survey of bereaved people, covering all causes of death, urgent care services had the lowest share of "excellent" ratings of any service, at 26%^[4].

Earlier palliative care is linked to fewer deaths in hospital

Making home possible takes time: equipment, medicines, nursing visits and a plan for what to do at night. Population studies suggest that when palliative care starts matters. In Taiwan, among 177,403 people aged 65 and over who received palliative care and died between 2010 and 2020, 61.5% died in hospital and 35.8% at home. Referral more than 180 days before death was associated with lower odds of dying in hospital (odds ratio 0.84), use of home palliative care with much higher odds of dying at home (odds ratio 3.93), and having cancer, compared with other conditions, with higher odds of dying at home (odds ratio 1.40)^[15].

In Finland, a national register study followed all 1,199 people who died of pancreatic cancer in 2019. Among the 270 people (22.5%) first in contact with specialist palliative care more than 30 days before death, 56% died in hospital, 22% on a specialist palliative care ward and 19% at home. Among those whose first contact came in the last 30 days, or who had none, 79% died in hospital, 5% on a specialist ward and 13% at home^[16].

These are associations. People referred early may be different from those referred late in ways the data cannot capture, such as how quickly the illness changed or how much family support they had. But they fit the trial evidence on home palliative care, and they point in the same direction as the wider evidence on late referral, which our planned study on who gets palliative care and when will cover.

What the evidence asks of us

The research does not say everyone should die at home. It says most people would like to, some would not, some change their minds, and whether a wish is met depends heavily on services and on families. That points to a few clear tasks for commissioners, funders and policymakers.

• Fund the support that makes home possible. Home palliative care, community and district nursing and GP home visits are among the factors most consistently linked to dying at home, and home palliative care more than doubled the odds in trials^[12][13][2].
• Make help reachable at any hour. Community teams designed to be available at any time were linked to fewer hospital deaths, and urgent care was the lowest-rated service in England's bereaved people survey^[4][14].
• Refer earlier. Earlier contact with palliative care was linked to fewer hospital deaths in Taiwan and Finland^[15][16].
• Support carers as part of the plan. Home deaths rest on relatives, including whether they can take time off work^[13]. People who live alone or without family nearby need services that do not assume a carer is there^[12].
• Keep hospice and other places open as real choices. Preferences vary and change, so beds in hospices and palliative care units remain essential^[1][11].
• Record preference as well as place, and record place clearly. Asking again as illness changes, counting people whose wishes were never asked, and using death-record categories that separate hospices and care homes would let services see whether choice is being honoured^[5][9].

Language: this piece says "people with cancer" and "people living with and after cancer", and describes the illness and its treatment plainly, without combat metaphors. Clinical terms such as "stage", "incidence" and "net survival" are used in the sense the cited source defines them.

Why we're publishing this

Hospices and palliative care teams have long argued that where and how people die is shaped by the services around them, not just by illness. The evidence gathered here supports that case: the same wish to die at home is met far more often where home palliative care, visiting nurses and doctors, round-the-clock help and support for carers are in place. We have brought it together, with sources, so that hospices, palliative care organisations and the charities that work with them can use it in planning, funding bids and public campaigns. It is part of our cancer research library.

How to read this data

Place-of-death figures come from death registrations, which are close to complete but are classified differently from country to country, so cross-country comparisons show the scale of differences rather than exact rankings. Preference figures come from surveys of the public (answering a hypothetical question), of patients (sometimes asked once, sometimes repeatedly) and of bereaved relatives (reporting someone else's wish after a death); these answer different questions. Many studies of preferences leave out people whose wishes were never recorded. The factors linked to dying at home come mostly from observational studies, which show associations, not causes: people who die at home may differ from those who do not in ways the data cannot measure. The Cochrane review pools randomised trials and gives the firmest evidence on home palliative care. Several studies are from high-income countries, and some figures cover people with any serious illness rather than only cancer; we say so where that applies. Odds ratios are reported as odds, not risks.

Use this data

Free to cite with attribution to Better Health For Everyone. Charities, campaigners and journalists are welcome to reuse these figures and charts. For the underlying figures as a spreadsheet, get in touch.

Important: This article summarises published research for general information. It is not medical advice and cannot be used to diagnose, treat or predict anything about an individual. If you have concerns about your own or someone else's health, please speak to a qualified health professional.

Sources

1. Gomes B, Calanzani N, Gysels M, Hall S, Higginson IJ. Heterogeneity and changes in preferences for dying at home: a systematic review. BMC Palliative Care, 2013;12:7. europepmc.org/article/MED/23414145
2. Gomes B, Calanzani N, Curiale V, McCrone P, Higginson IJ. Effectiveness and cost-effectiveness of home palliative care services for adults with advanced illness and their caregivers. Cochrane Database of Systematic Reviews, 2013;(6):CD007760. europepmc.org/article/MED/23744578
3. Gomes B, Higginson IJ, et al. Preferences for place of death if faced with advanced cancer: a population survey in England, Flanders, Germany, Italy, the Netherlands, Portugal and Spain. Annals of Oncology, 2012;23(8):2006-2015. europepmc.org/article/MED/22345118
4. Office for National Statistics. National Survey of Bereaved People (VOICES): England, 2015. ONS statistical bulletin, 2016. ons.gov.uk/.../nationalsurveyofbereavedpeoplevoices/england2015
5. Hoare S, Morris ZS, Kelly MP, Kuhn I, Barclay S. Do patients want to die at home? A systematic review of the UK literature, focused on missing preferences for place of death. PLoS One, 2015;10(11):e0142723. europepmc.org/article/MED/26555077
6. van de Beek SH, Eckels KA, Olet DA, et al. How preferences and reality on where we die unfold: a four-country longitudinal qualitative study (EOLinPLACE). Health Expectations, 2026;29(4):e70732. europepmc.org/article/MED/42400232
7. Dannenberg A, Cohen J, Deliens L, et al. Place of death of people with cancer in 12 Latin American countries: a population study using national death registers. Cancer Medicine, 2025;14(12):e70996. europepmc.org/article/PMC/PMC12175479
8. Öhlén J, Nyblom S, Ozanne A, et al. Influence of palliative care policy on place of death for people with different cancer types: a nationwide register study. PLoS One, 2025;20(3):e0320086. europepmc.org/article/PMC/PMC11949374
9. Delalibera M, Lopes S, Dias da Silva I, Gomes B. International differences in the technical qualities of place of death classifications and data: a survey of researchers' perspectives. Palliative Medicine, 2026;40(9):1420-1426. europepmc.org/article/MED/42544535
10. Bell CL, Somogyi-Zalud E, Masaki KH. Factors associated with congruence between preferred and actual place of death. Journal of Pain and Symptom Management, 2010;39(3):591-604. europepmc.org/article/MED/20116205
11. Yang CH, Wu CY, Cheng SY, et al., on behalf of the EASED investigators. Reasons for and congruence between preferred and actual place of death among cancer patients receiving end-of-life care: a cross-cultural multicenter prospective cohort study in East Asia. Cancers, 2025;17(13):2062. europepmc.org/article/MED/40647364
12. Gomes B, Higginson IJ. Factors influencing death at home in terminally ill patients with cancer: systematic review. BMJ, 2006;332:515-521. europepmc.org/article/MED/16467346
13. Gomes B, Calanzani N, Koffman J, Higginson IJ. Is dying in hospital better than home in incurable cancer and what factors influence this? A population-based study. BMC Medicine, 2015;13:235. europepmc.org/article/PMC/PMC4599664
14. Seow H, Brazil K, Sussman J, et al. Impact of community based, specialist palliative care teams on hospitalisations and emergency department visits late in life and hospital deaths: a pooled analysis. BMJ, 2014;348:g3496. europepmc.org/article/MED/24906901
15. Lo YT, Li CY, Chuang TJ, Huang YT, Wu YL, Yang YC. Bridging policy and practice: determinants of place of death for palliative care recipients in Taiwan. Supportive Care in Cancer, 2026;34(3):177. europepmc.org/article/MED/41661347
16. Rautakorpi L, Nåhls NS, Saarto T, Nuutinen M, Akrén O, Carpén T. Assessing the impact of specialist palliative care on healthcare utilisation at the end of life among patients with pancreatic cancer: a nationwide register-based cohort study. Cancer Medicine, 2026;15(5):e71908. europepmc.org/article/MED/42057456