Blog

Dignity in dying: when does treatment stop being care?

22 October 2026 · Echelon Medical Solutions

"When do we stop treating a dying patient?" is a question that sits uneasily with most people who work in healthcare, and with most families. UK guidance and audit evidence suggest it is better asked a different way: when does treatment aimed at prolonging life stop serving this person, and when should care aimed at comfort and dignity take the lead? Care itself never stops. This article looks at what UK regulators, national guidelines and audits say, where care has fallen short, and what the evidence cannot tell us. It is general information, not clinical advice, and every source is from the UK.

Why this matters

Each year over 600,000 people die in the UK, and many of these deaths occur in hospital, even though most people say they would prefer not to die there. About 70% of people die from long-term conditions that often follow a predictable course, so death is anticipated well in advance. The annual number of UK deaths is predicted to rise to 736,000 by mid-2035. A national confidential enquiry also found that almost a third of people admitted to hospital as an emergency are in their last year of life. For most deaths, there is time to plan.

What the GMC framework says about stopping treatment

The General Medical Council’s guidance, Treatment and care towards the end of life, is the clearest statement of the professional and ethical framework in the UK. It is addressed to doctors, physician associates and anaesthesia associates, but it also describes what patients and families can expect. Its key principles are:

A presumption in favour of prolonging life. Decisions about potentially life-prolonging treatment must start from this presumption, which normally means taking all reasonable steps to prolong life. They must never be motivated by a desire to bring about death.
But it is not absolute. The guidance states there is no absolute obligation to prolong life irrespective of the consequences for the patient, and irrespective of the patient’s views if they are known or can be found out.
Benefits are weighed against burdens and risks. For example, it may not be in a patient’s interests to receive potentially life-prolonging but burdensome treatment in the last days of life, when the focus of care is changing from active treatment to managing symptoms and keeping them comfortable.
The patient decides if they have capacity. Every adult is presumed to have capacity, and clinicians must provide help and support to maximise it. Capacity cannot be assumed to be lacking because of age, disability, a medical condition, or because the person makes a decision others consider unwise.
If the patient lacks capacity, decisions are made on the basis of overall benefit and which option would be least restrictive of their future choices. This involves checking for an advance refusal or someone with legal authority, and consulting those close to the patient, the healthcare team and the patient’s own wishes and values.
No judgements based on personal views of quality of life. The guidance warns against relying on personal views of someone’s quality of life, or on unfounded assumptions about older people and people with disabilities. It also notes that older people, disabled people and people from ethnic minorities have received poorer standards of care towards the end of life.

Dignity is written into the duty

The guidance says doctors must give patients approaching the end of life the same quality of care as all other patients, and must treat them and those close to them with dignity, respect and compassion, especially when facing difficult decisions.

What a DNACPR decision does and does not mean

The guidance is careful about CPR. It notes that CPR generally has a very low success rate, that it is invasive and carries burdens such as damaged organs and broken ribs, and that if it fails the patient may die in an undignified and traumatic manner. Where cardiac or respiratory arrest is an expected part of dying and CPR will not work, a decision made and recorded in advance can help the patient die in a dignified and peaceful way, and can help them spend their last hours at their preferred place.

It also sets out important limits:

A DNACPR decision applies only to CPR. It does not imply that other treatment will be withdrawn or withheld.
When explaining the decision to a patient who lacks capacity, the intention should be made clear: to spare the patient treatment that will be of no benefit, not to withhold any other care they need.
Where CPR might restart the heart but would lead to treatment unlikely to be appropriate afterwards, such as prolonged support for multi-organ failure, that can be part of the decision.
If a patient wants CPR and the clinician judges it would not be clinically appropriate, the clinician should explore the patient’s reasons and expectations. When the benefits, burdens and risks are finely balanced, the patient’s request will usually decide. If the clinician still considers CPR inappropriate there is no obligation to provide it, and the patient has the right to a second opinion.
In an emergency with no time to find information, no prior decision and no known wishes, CPR should be attempted unless, in the clinician’s judgement, it will not be successful.
A recorded DNACPR decision is not itself legally binding. It is a clinical assessment, and it should not override clinical judgement if the patient arrests from a potentially reversible cause that was not envisaged when it was made.

Recognising dying: the pivot point

NICE guideline NG31, Care of dying adults in the last days of life, covers the last two to three days of life. Its stated aim is to improve care by communicating respectfully, involving the person and those important to them in decisions, and maintaining their comfort and dignity. NICE’s own implementation advice makes the central point: correctly recognising that a person is in the last days of life can create opportunities for shared decision-making and prevent unnecessary interventions. It also says recognising dying can be challenging.

The guideline asks clinicians who think someone may be entering the last days of life to gather and document:

the person’s physical, psychological, social and spiritual needs
current clinical signs and symptoms
medical history and clinical context, including underlying diagnoses
the person’s goals and wishes
the views of those important to the person about future care.

It also tells clinicians to watch for the person who may stabilise or improve, even temporarily. In other words, recognising dying is a judgement that has to be revisited, not a label applied once. The guideline covers recognition, communication and shared decision-making, hydration, pharmacological symptom control and anticipatory prescribing. It draws on lessons from the independent review of the Liverpool Care Pathway, which highlighted a lack of understanding of the role of hydration at the end of life.

The wider UK framework for dying people, set out in the 2014 report One Chance to Get It Right, rests on five principles: recognising imminent death, communication with the patient, communication with the family, the needs of families, and an individual plan of care.

What the national audit shows

The National Audit of Care at the End of Life (NACEL) reviews the final hospital admission of adults who die in England, Wales and Jersey. Its most recent report, published in August 2026, draws on 19,705 case note reviews and 5,994 surveys of bereaved people.

Measure (2025 data)	Result
Patients recognised as likely to die during the final admission	84%
People expected to die with an individualised plan of care	86%
Hydration review documented in the last days of life	82%
Bereaved people rating the care excellent or good	77%
Patients reviewed by a specialist palliative or end of life care team	58% (47% in 2022)
Hospital providers with face-to-face specialist palliative care seven days a week	64% (61% in 2024)
Assessments of spiritual, religious and cultural needs recorded	49%

The report found improvement in several areas, but fewer than half of patients had documented evidence of taking part in personalised care and support planning conversations. It also found differences in care by ethnicity, with poorer outcomes where ethnicity was not recorded, and it recommends further national research into inequalities.

The audit’s earlier 2022 report gives a sense of the time available once dying is recognised. Dying was recognised in 87% of cases, and the median time from recognition to death was 47 hours. In that audit, 13% of families disagreed that staff had communicated sensitively with the dying person, and a quarter of cases had a documented discussion about how far the patient wished to be involved in their care. The report concluded that earlier action is needed once uncertain recovery is identified.

When care falls short

The Ombudsman’s findings

In 2015 the Parliamentary and Health Service Ombudsman published Dying without Dignity, drawing on 265 complaints about end of life care investigated between January 2011 and April 2015. It identified recurring themes, including not recognising that people are dying or responding to their needs, poor communication, poor planning leading to uncoordinated care, inadequate out-of-hours services and poor pain management. It estimated that end of life care could be improved for up to 355,000 people a year in England. One case involved a 74-year-old man with cancer who had 14 unnecessary attempts to reinsert a drip and spent his last days in avoidable pain.

Burdensome tests in the last days

A quality improvement project across nine medical wards on two Scottish hospital sites (reported as a conference poster) looked at patients whose deaths were expected. At baseline, 53% (46 of 86) still had blood tests and 66% (57 of 86) routine observations in the last 24 hours of life. The rate varied from 14% to 77% between wards. The ward with the lowest rates had palliative medicine consultant input almost daily and a palliative medicine consultant at its weekly team meeting, where treatment ceilings and the benefits and burdens of specific interventions were routinely discussed. Interventions such as a sticker in the notes, staff education and an electronic alert produced a significant improvement on only one ward, which is itself a useful finding: this is a hard change to achieve.

Planning earlier

The 2024 NCEPOD report Planning for the End reviewed care for adults with dementia, heart failure, lung cancer or liver disease who died in hospital between April and September 2022. It drew on 701 clinician questionnaires and 350 sets of case notes from NHS hospitals in England, Wales and Northern Ireland. Its findings show how late the conversation often starts:

Parallel planning, meaning palliative care alongside disease-modifying treatment, was used for 135 of 439 patients (30.8%).
Where it was used, 94 of 130 patients (72.3%) had specialist palliative care input, compared with 120 of 290 (41.4%) where it was not.
Preferences for care at the end of life were not recorded for 169 of 233 patients (72.5%).
A lead person was documented for 257 of 396 patients (64.9%). Where there was one, specific end of life documentation was used for 66.7%, against 32.8% where there was not.
Seven-day specialist palliative care was available in 125 of 210 hospitals (59.5%).
End of life care training was part of induction in only 137 of 214 hospitals (64.0%).

Its recommendations include giving patients with advanced chronic disease palliative care alongside treatment, normalising conversations about death and dying (for example by asking "Would you be surprised if this patient died within the next 12 months?"), naming a care coordinator, providing specialist palliative care in hospitals and the community, training all patient-facing staff, and ensuring that existing advance care plans, including ReSPECT forms, are shared between all providers. The report notes that these documents often stay in one place when a patient is transferred, which means their wishes are often not followed.

What this means for pre-hospital responders

Many pre-hospital encounters happen when someone is already near the end of life, and responders rarely have the full picture.

The GMC guidance acknowledges that in an emergency there may be no time to access information, so CPR should be attempted unless it clearly will not succeed. This is why advance decisions need to be recorded and visible.
NCEPOD lists ambulance trusts among the bodies supporting its recommendation that advance care plans be shared between all providers, because a plan is of little use if it does not travel with the patient.
A DNACPR decision is about CPR only. Comfort, symptom control, communication and respect continue regardless.
Recognising that someone may be dying, and communicating that clearly to the next clinician, is itself a form of care.

Echelon’s training is built around UK pre-hospital care frameworks.

What the evidence cannot tell us

The GMC and NICE sources are guidance, not studies, and neither can say when stopping a specific treatment is right for a specific person.
NACEL and NCEPOD describe hospital care, and mostly care for people who died in hospital. They say less about deaths at home, in care homes or in hospices.
The audit data show how often processes are documented, not how a death felt to the person.
The Scottish figures come from a small quality improvement project reported as a poster, and the Ombudsman’s report is based on complaints, which are unlikely to be representative of all care.

Key takeaways

UK guidance starts from a presumption in favour of prolonging life, but states there is no absolute duty to prolong life regardless of its consequences for the patient or their wishes.
The decision is about the balance of benefit and burden for this person, led by their wishes, or their overall benefit if they lack capacity, and never by assumptions about their quality of life.
A DNACPR decision applies only to CPR. Comfort and dignity remain the priority throughout.
Recognising dying is the pivot, and NICE reminds clinicians to keep reviewing it, because people can stabilise or improve.
Audit data show real progress, with 84% recognised as likely to die, but personalised care planning, spiritual and cultural assessment and early palliative care remain patchy.

Over to you: What would make it easier for clinicians and families to say out loud that someone is dying?

Sources

Interested in pre-hospital medical training for your team? Contact operations@echelonmedicalsolutions.com.