End-of-Life Care and Advance Care Planning

The answer first

End-of-life care is the care of a person who is likely to die within the next 12 months — not only the care delivered in the final days. The broad definition matters because most of what improves a dying patient's experience (advance care planning, the alignment of treatment with values, the avoidance of non-beneficial interventions, the timely involvement of palliative care) must be done months before death, when the patient still has capacity and the team still has options. The common failure in acute medicine is leaving these conversations until the dying phase, by which point the patient has lost the chance to direct their own care. [1]

The four rules that govern the whole topic: [1]

1. End-of-life care begins when you would not be surprised if the patient died within the next year. The Surprise Question — "Would I be surprised if this patient died in the next 12 months?" — is the trigger for advance care planning. A "no" (I would not be surprised) means: initiate the conversation now.
2. Advance care planning is a conversation, not a document. A signed directive alone changes little; a structured, repeated discussion that is documented and accessible at the point of care is what reduces non-beneficial treatment, increases concordance with patient preferences, and lowers family distress. Detering's randomised trial showed facilitated ACP delivered exactly these effects [2].
3. A ceiling of treatment is not a withdrawal of care. "Not for resuscitation" or "not for ICU escalation" means the team will continue active treatment within the agreed limits and will focus on symptom control — the phrase is do not resuscitate, but do not abandon. Withholding and withdrawing life-sustaining treatment are ethically and legally equivalent [9].
4. The dying patient has a recognisable syndrome, and anticipatory prescribing means you do not have to chase symptoms in a crisis. Reduced consciousness, reduced oral intake, changing breathing pattern, mottled skin and peripheral cooling herald the last days of life. The Just In Case Box (morphine for pain and breathlessness, midazolam for agitation, glycopyrronium or hyoscine for secretions, levomepromazine for nausea and agitation) is prescribed in advance so that the nurse can give symptom relief immediately, at any hour.

DWE high-yield trap: When asked "best management of an older inpatient with end-stage organ failure who is deteriorating and has no advance directive," the answer is almost never "escalate to ICU." The answer is recognise the end-of-life phase, hold a goals-of-care conversation, document a ceiling of treatment, and provide active symptom-focused care — with palliative care input. Escalation without a prospect of meaningful recovery is non-beneficial and often harmful [3][11].

---

Definitions — what end-of-life care includes and excludes

End-of-life care sits within palliative care but is broader in one dimension (it explicitly brackets the final 12 months) and narrower in another (palliative care extends back to the point of diagnosis of any life-limiting illness). [1]

• Palliative care (WHO): an approach that improves the quality of life of patients and families facing the problem associated with life-threatening illness, through the prevention and relief of suffering by early identification, impeccable assessment, and treatment of pain and other problems — physical, psychosocial and spiritual. Temel's trial of early palliative care in metastatic lung cancer showed it improved quality of life, reduced depression and prolonged survival [4] — palliative care is not only for the dying.
• End-of-life care: care of a person with an advanced, progressive, incurable condition who is expected to die within the next 12 months. It includes the last days of life (the "dying phase") but begins much earlier.
• Terminal care: the care of the patient in the last days to hours of life — the dying phase. This is the narrowest of the three terms.

Why the 12-month definition matters in practice: it is the trigger for advance care planning, for the discussion of preferred place of death, for involving palliative care while the patient can still participate, and for the goals-of-care conversation that sets a ceiling of treatment before a crisis forces an ICU decision at three in the morning. [1]

Exam trap: A common MRCP/FRACP question frames "palliative care" as only for patients who are imminently dying. This is wrong. Palliative care is appropriate at any stage of a life-limiting illness, alongside disease-directed therapy. The Temel trial and ASCO guidelines establish early specialist palliative care referral as standard for advanced cancer [4].

---

Advance care planning

What advance care planning is

Advance care planning (ACP) is a structured series of conversations between a patient with capacity, their family and their clinicians about the patient's values, goals and preferences for future medical care — conducted while the patient can still decide, documented so it is accessible at the point of care, and revisited as the clinical situation changes. The two products of a good conversation are (1) a record of the patient's values and goals, and (2) the appointment and preparation of a substitute decision-maker who can speak for the patient when capacity is lost. [1]

The core content of an ACP discussion: [1]

• The patient's understanding of their illness and prognosis — what they know, what they want to know, and what matters most to them now.
• Their values and goals — what makes life worth living for them, what outcomes they would find unacceptable (prolonged unconsciousness, dependence on machines, loss of the ability to recognise family).
• Treatment preferences — what they would and would not want: cardiopulmonary resuscitation, mechanical ventilation, artificial hydration and nutrition, dialysis, antibiotics in the dying phase, admission to intensive care.
• Preferred place of death — home, hospice, hospital, residential aged-care facility. Most patients say home; most still die in hospital, which is the gap ACP is meant to close.
• The substitute decision-maker — who they appoint (and whether that person accepts the role and understands the patient's values). [1]

When to initiate — the Surprise Question

The trigger for ACP is the Surprise Question: "Would I be surprised if this patient died in the next 12 months?" If the answer is no (I would not be surprised), initiate advance care planning now. The Surprise Question came from the Gold Standards Framework in primary care and is the simplest, most widely taught screening tool for end-of-life need. It is deliberately permissive — it errs toward starting the conversation. [1]

The patients who should trigger a "no" in a physician practice include: [1]

• Anyone with advanced organ failure (NYHA III–IV heart failure, oxygen-dependent COPD, stage 4–5 chronic kidney disease, decompensated cirrhosis).
• Anyone with metastatic or locally advanced incurable cancer.
• Anyone with advanced dementia, or anyone who is bed-bound, has lost the ability to perform activities of daily living, or has significant weight loss.
• Frail older patients, especially those who have fallen, been admitted, or acquired a new disability.
• Any patient in intensive care or after a critical illness where the trajectory has changed. [1]

DWE high-yield: The question "a 79-year-old with NYHA IV heart failure and an ejection fraction of 22 per cent is admitted with another decompensation — what is the most appropriate next step in his long-term management?" is testing the Surprise Question. The answer includes optimal guideline-directed medical therapy and initiating advance care planning with a goals-of-care discussion and a documented ceiling of treatment, because you would not be surprised if he died within the year. [1]

ACP should be revisited, not done once. Review it at milestones: after each hospitalisation, after each major functional decline, at each change of care setting, and whenever the patient asks. The Serious Illness Conversation Guide (Ariadne Labs) is a validated, structured script that elicits values, shares prognosis and documents goals, and it improves the frequency and quality of these conversations [7][8].

The evidence that ACP works

ACP is not a soft skill — it has hard outcome evidence: [1]

• Detering 2010 (BMJ RCT, Melbourne): facilitated ACP in medical inpatients aged 80 and over made it far more likely that end-of-life wishes were known and respected (86 per cent vs 30 per cent), reduced stress, anxiety and depression in bereaved relatives, and increased family satisfaction with care [2].
• Wright 2008 (JAMA cohort): in advanced cancer, patients who had end-of-life discussions with their clinicians received less aggressive care in the final week (less ventilation, less resuscitation, less ICU, more and earlier hospice), and there was no increase in depression or distress. More aggressive care was associated with worse quality of life near death and more major depressive disorder in bereaved caregivers [3].
• Brinkman-Stoppelenburg 2014 (systematic review): ACP was associated with reduced life-sustaining treatment at the end of life, increased hospice and palliative care use, fewer hospitalisations, and better concordance between patient preferences and the care delivered — and "complex" facilitated ACP outperformed a directive alone [1].

The practical message: a conversation that is documented and accessible changes what happens to the patient. A form that is filed in a records office does not. [1]

---

Advance directives and substitute decision-makers

Advance directives — legally binding refusals

An advance care directive (advance directive, living will) is a written statement, made by a person with capacity, about the medical treatment they would or would not want if they later lose capacity. The legally binding part is the refusal of treatment: a capacitous adult can refuse any treatment, including life-sustaining treatment, in advance, and that refusal must be respected even when the patient has lost capacity. Treating a patient in the face of a valid and applicable advance refusal is battery. [1]

What an advance directive can do: [1]

• Refuse specific treatments outright — for example refusing cardiopulmonary resuscitation, mechanical ventilation, artificial hydration and nutrition, or dialysis.
• Express values and preferences that guide the substitute decision-maker ("I would not want to be kept alive if I could not recognise my family").
• appoint a substitute decision-maker (an enduring guardian, a medical enduring power of attorney, a welfare attorney) to make decisions not covered by the directive. [1]

What an advance directive generally cannot do: [1]

• Demand a specific treatment that the clinician considers clinically inappropriate or futile — a patient cannot compel a doctor to provide non-beneficial treatment.
• Be used to request euthanasia or physician-assisted suicide outside the narrow voluntary assisted dying legislation in jurisdictions where it is legal. [1]

For a directive to be valid and applicable, it must: have been made by a person with capacity, be specific enough to apply to the current situation, and not have been revoked. The clinician's job when a directive is produced in a crisis is to confirm validity and applicability — and to honour a valid refusal even when the family disagrees. [1]

Substitute decision-makers

When a patient lacks capacity, the law provides a hierarchy of decision-makers whose role is to represent the patient's values and prior wishes (the substituted judgement standard), not to impose their own preferences. The terminology varies by jurisdiction: [1]

• Australia: an enduring guardian, a medical enduring power of attorney, or a person responsible as defined by state guardianship legislation (the hierarchy typically runs to a spouse, then an adult child, then a parent, then a sibling, then a primary unpaid carer). Where there is no person available or willing, the guardianship tribunal (state-based — for example the Victorian Civil and Administrative Tribunal, the New South Wales Guardianship Division of NCAT) can appoint a guardian or make the decision directly.
• New Zealand: a welfare guardian appointed under the Protection of Personal and Property Rights Act 1988.
• United Kingdom: a person with lasting power of attorney for health and welfare under the Mental Capacity Act 2005, or — failing that — a consultee under the best-interests process, with the Court of Protection as the final arbiter. [1]

The substitute decision-maker must be given honest prognostic information, must be asked what the patient would have wanted, and must be supported through the decision — not simply handed the choice. When the substitute decision-maker and the team genuinely cannot agree, the escalation is a second senior opinion, the hospital clinical ethics service, and finally the guardianship tribunal or court. [1]

---

Goals of care and ceilings of treatment

Aligning treatment with values and prognosis

A goals-of-care conversation is the clinical act that translates the patient's values into a treatment plan. It asks, in plain language: given where you are in your illness, what are we trying to achieve, and what are we willing to put you through to get there? The answer defines the goals — cure, life prolongation, function, comfort — and sets the ceiling and the floor of treatment accordingly. [1]

The conversation has to integrate three things: the patient's values (what matters most), the realistic prognosis (what is likely to happen), and the medical evidence (what the treatment can and cannot do). When any one of these dominates without the others, the plan goes wrong — paternalism when the prognosis dominates, futile escalation when the family's wish dominates, abandonment when the clinician gives up on a patient who still wants to fight. [1]

Ceilings of treatment

A ceiling of treatment is a documented medical decision that sets the upper limit of intervention in the event of deterioration. It states what will be done and what will not be done. A well-documented ceiling includes: [1]

• What will continue: active treatment of the current reversible problem (antibiotics, oxygen, fluids, physiotherapy), symptom control, and all comfort measures.
• What will not be escalated to: for example, no admission to intensive care, no mechanical ventilation, no cardiopulmonary resuscitation, no inotropic support.
• The reasoning: the prognosis, the goals, and the patient's or substitute decision-maker's informed agreement. [1]

The defining principle is that a ceiling is not a withdrawal of care. A patient with a "not for ICU, not for resuscitation" order still receives antibiotics for their pneumonia, oxygen for their hypoxia, fluids for their dehydration, and active symptom control — they simply will not be transferred to intensive care if these measures fail. The phrase to use with the team and family is do not resuscitate, but do not abandon. [1]

Not for resuscitation is not the same as palliative care

A common error is to conflate a "not for resuscitation" (NFR, DNAR, DNACPR) order with a decision to provide palliative care only. They are distinct: [1]

• NFR is a decision about one intervention — cardiopulmonary resuscitation — in the event of cardiac arrest. It says nothing about the rest of the treatment.
• Palliative / comfort care is a goal of care that prioritises symptom control and quality of life over disease-modifying therapy. [1]

A patient can have an NFR order and still receive full active treatment for a reversible problem (a patient with curative-intent chemotherapy who has an NFR because CPR in their state would be futile). A patient can be receiving palliative care and not have an NFR (a patient with metastatic cancer who still wants resuscitation attempted). Most patients approaching the end of life will have both, but the clinician must make each decision explicitly and explain the distinction to the patient and family. The default in many hospitals is a goals-of-care / medical goals-of-care form that records the resuscitation decision, the escalation ceiling and the focus of care in a single document, communicated to the ward, the ambulance service and the residential aged-care facility. [1]

---

Recognising the dying patient (last days of life)

The dying patient has a recognisable clinical syndrome. The cardinal signs of the last days of life are: [1]

1. Reduced consciousness — the patient becomes progressively drowsy, rouses less, and may become unresponsive. This reflects global cerebral hypoperfusion and the systemic shutdown of organ failure, not (usually) a new structural lesion.
2. Reduced oral intake — the patient stops eating and drinking, and swallowing becomes weak or absent. This is a marker of the dying process, not its cause; the team should not reflexively insert intravenous fluids to "correct" it.
3. Changes in breathing pattern — Cheyne-Stokes respiration (periods of apnoea alternating with crescendo-decrescendo tidal volumes), shallow rapid breathing, agonal breathing, and the accumulation of secretions in the upper airway producing the "death rattle."
4. Skin changes — mottling of the limbs (patchy purple discolouration), peripheral cooling, and cyanosis. These reflect failing peripheral circulation.
5. Other — urinary and faecal incontinence as sphincters relax, restlessness or terminal agitation, and (less commonly) fever. [1]

Recognising the dying phase matters because it changes the goal of care. The team switches from the curative or stabilising plan (where each deterioration triggers a new investigation and escalation) to a plan focused on symptom control and comfort, with anticipatory prescribing, the cessation of non-essential interventions, and active communication with the family. Failing to recognise the dying patient is the commonest reason for inappropriate, non-beneficial escalation in the last days of life. [1]

DCE long-case trap: The registrar who says "we will keep going, we can always make him not for resuscitation later" has it backwards. The decision about the ceiling should be made before the deterioration, not in the middle of it. A long-case candidate who fails to address the patient's end-of-life trajectory has missed a major mark. [1]

---

Management of the dying patient

Once the dying phase is recognised, the care plan is a structured bundle — documented, communicated to the team and family, and reviewed at least daily. The components are: stop non-essential medications and interventions, anticipatory prescribing, symptom control, comfort measures, and family and spiritual support. [1]

Stop non-essential medications and interventions

Review the medication chart. Stop medications that are not contributing to comfort in the immediate term: statins, antihypertensives, oral hypoglycaemics, disease-modifying drugs, antiplatelets and anticoagulants (unless being given for symptom relief), vitamins, and most preventive therapies. Continue medications that prevent distress: analgesia (converted to a subcutaneous route if the patient cannot swallow), antiemetics, anticonvulsants, anxiolytics if required, and drugs for symptom control (for example, glyceryl trinitrate for angina, furosemide for pulmonary oedema). The aim is to reduce the medication burden to what serves comfort. [1]

Stop non-essential investigations (blood tests, imaging) and non-essential monitoring (frequent vital signs, continuous oximetry) that disturb the patient without changing the plan. Continue only the monitoring that guides comfort. [1]

Anticipatory prescribing — the Just In Case Box

Anticipatory prescribing means writing the prescriptions for the common symptoms of the dying phase before they occur, so that the nurse can give the drug immediately at any hour, without waiting for a doctor. In Australia and New Zealand this is the Just In Case Box (also called a terminal care pack, or in the UK the anticipatory prescribing "just in case" medicines). NICE Guideline NG142 (Care of Dying Adults in the Last Days of Life) endorses anticipatory prescribing for dying patients. [1]

The four core subcutaneous medicines and their target symptoms: [1]

Important clinical points on anticipatory prescribing: [1]

• The subcutaneous route is preferred in the dying patient because intravenous access is hard, intramuscular injections are painful, and the patient cannot swallow.
• A continuous subcutaneous infusion (syringe driver) is used when PRN doses are required repeatedly, when the patient cannot take oral medicines, or for continuous symptom control. The total PRN dose over 24 hours is converted to the infusion, with PRN top-ups still available for breakthrough.
• Morphine for breathlessness is well supported — it reduces the ventilatory drive and the subjective distress of dyspnoea at doses that do not cause harmful respiratory depression in the opioid-tolerant or the dying patient. For the non-hypoxaemic breathless patient, supplemental oxygen offers no benefit over room air (Abernethy trial) — a fan, positioning and low-dose opioids are the tools [5].
• Noisy secretions are usually more distressing to the family than to the semi-conscious patient. Positioning (lateral, head down), reducing non-essential fluids, and an antimuscarinic early are the management. Suction is distressing and rarely helpful for the secretions of the dying phase.

Symptom control summary

The common symptoms in the last days of life, and the first-line approach: [1]

• Pain: continue the patient's usual analgesia, converted to a subcutaneous route; morphine 2.5–5 mg SC PRN (opioid-naive) or the equivalent; adjuvants (for neuropathic pain, consider a subcutaneous ketamine or methadone infusion via specialist palliative care).
• Breathlessness: treat reversible causes where appropriate (oxygen if hypoxaemic, bronchodilators, treating effusions); low-dose morphine; fan and positioning; benzodiazepine if associated anxiety is prominent.
• Agitation and terminal restlessness: exclude reversible causes (urinary retention, constipation, pain, hypoxia, drug toxicity); midazolam; an antipsychotic (haloperidol or levomepromazine) if delirium is the driver.
• Nausea and vomiting: identify the cause (chemical, gastric stasis, bowel obstruction, raised intracranial pressure) and treat cause-specifically — haloperidol or levomepromazine for chemical, metoclopramide for gastric stasis, hyoscine for obstruction, dexamethasone for raised pressure.
• Secretions: antimuscarinic (glycopyrronium or hyoscine butylbromide), positioning, fluid reduction, family explanation.
• Delirium: treat reversible precipitants; haloperidol or low-dose atypical antipsychotic for severe distress; avoid benzodiazepines unless alcohol or benzodiazepine withdrawal (cross-reference the delirium topic). [1]

Comfort measures

Mouth care is a basic dignity. The dying patient breathes through the mouth and the mucosa dries; perform regular mouth care — moisten the mouth with water on a swab, apply lip balm, clean secretions, and remove poorly fitting dentures. Artificial saliva helps xerostomia. [1]

Pressure area care: reposition every 2–4 hours for comfort (not as rigidly as for prevention in a patient who is imminently dying), use a pressure-relieving mattress, keep the skin clean and dry, and manage incontinence. The aim is comfort and dignity. [1]

Continence: manage with pads or a urinary catheter if the patient is distressed by wetness or if skin breakdown is a concern. [1]

Family, spiritual and psychological support

The dying patient is part of a family system, and the family's experience of the death shapes their bereavement. The team should: [1]

• Communicate proactively — explain the signs of dying, what the plan is, that the patient is not in pain, and that the family can be present. Update regularly.
• Offer presence — allow the family to be with the patient, to touch and speak to them, and to perform rituals that matter to them.
• Respect culture and faith — ask about religious and cultural practices around death and dying, and facilitate them (the timing of last rites, the handling of the body, the presence of a chaplain or elder).
• Screen for spiritual distress — suffering at the end of life is not only physical. Ask what gives the patient meaning and whether they want to see a chaplain or spiritual carer.
• Support the staff — caring for dying patients is hard; debrief after difficult deaths. [1]

---

Ethical issues at the end of life

Withholding and withdrawing treatment are ethically equivalent

The central ethical principle is that withholding (not starting) and withdrawing (stopping) life-sustaining treatment are ethically and legally equivalent — there is no moral difference between never starting a ventilator and taking a patient off one after a trial has failed [9]. The practical consequence is that a treatment should be started as a trial whenever there is genuine uncertainty about the prognosis: it is easier, psychologically, to start and then stop than to withhold and then regret. Clinicians (and families) often find withdrawing emotionally harder than withholding, but this asymmetry is a bias, not a principle.

The corollary is that there is no obligation to provide futile treatment. A treatment is futile when it cannot achieve its physiological purpose (for example, CPR in a patient with end-stage multisystem failure) or when it offers no realistic prospect of a meaningful outcome the patient would value. Futility is a clinical judgement informed by prognosis and by the patient's goals, and it can and should support a decision to withhold or withdraw — in consultation with the patient or substitute decision-maker. [1]

The doctrine of double effect

The doctrine of double effect distinguishes between the intended effect of a treatment (relief of suffering) and a foreseen but unintended adverse effect (for example, respiratory depression from an opioid). It provides the ethical and legal basis for giving adequate analgesia and sedation to the dying patient even when a foreseeable consequence is shortening of life — provided the primary intention is symptom relief and the dose is proportionate to the symptom. In practice, opioids and benzodiazepines correctly titrated for symptom relief at the end of life do not shorten life, and the doctrine is most often invoked to reassure anxious clinicians and families that good symptom control is not euthanasia. [1]

Voluntary assisted dying (VAD)

Voluntary assisted dying — where a competent adult with a terminal illness can request and self-administer (or, in limited circumstances, have a practitioner administer) a lethal medication — is legal in some Australian states and in New Zealand, under strictly regulated conditions, and is distinct from palliative care, from the doctrine of double effect, and from withdrawal of life-sustaining treatment. [1]

The legal status (as of 2026): [1]

• Victoria — Voluntary Assisted Dying Act 2017 (the first Australian jurisdiction; commenced June 2019).
• Western Australia — Voluntary Assisted Dying Act 2019 (commenced July 2021).
• Other Australian states subsequently legislated: South Australia, Tasmania, Queensland, and New South Wales (Acts passed 2021–2022, commenced progressively through 2023).
• New Zealand — End of Life Choice Act 2019, brought into force by referendum in 2020 and commenced November 2021.
• The Northern Territory was historically the first jurisdiction in the world to legalise assisted dying (Rights of the Terminally Ill Act 1995), but this was overridden by the Commonwealth's Euthanasia Laws Act 1997. The Territories regained the power to legislate in 2022. [1]

The eligibility criteria are broadly similar across jurisdictions: the patient must be an adult with decision-making capacity, acting voluntarily, with a terminal illness that is advanced, progressive and expected to cause death within a defined period (commonly 6 or 12 months), and experiencing suffering the patient considers intolerable. The process requires multiple requests, assessments by two independent medical practitioners, and a final self-administration (or, where the patient cannot self-administer, practitioner administration in some jurisdictions). [1]

For the physician, the key points are: (1) a request for hastened death is usually a communication about suffering — explore the reason (often unrelieved pain, depression, fear, loss of autonomy, feeling a burden), address reversible suffering, involve specialist palliative care, and do not abandon the patient; (2) where VAD is legal, a clinician who conscientiously objects does not have to participate but must not abandon the patient and should provide information and referral; (3) VAD is never a substitute for high-quality palliative care, and most patients who want to die change their minds when their suffering is relieved. [1]

---

Communication — breaking bad news and discussing prognosis

The SPIKES protocol

The SPIKES protocol (Baile 2000) is the standard structured approach to delivering bad news [6]:

The two steps most often poorly done are Perception (asking rather than assuming what the patient knows) and Empathy (sitting with the emotion rather than rushing to information or a plan). A good bad-news conversation is measured by whether the patient feels heard, not by how much information was transferred. [1]

Discussing prognosis honestly

Discussing prognosis is a clinical skill and an ethical duty. The principles: [1]

• Ask before telling. Find out what the patient wants to know — some want a number, some want a range, some want only the gist. Match your answer to their preference.
• Be honest but humane. Do not give false hope, but do not strip all hope either. The frame is "hope for the best, prepare for the worst."
• Give a range, not a point estimate. "Weeks to a few months, rather than years" is more useful and more honest than a false-precision number, and acknowledges the irreducible uncertainty of individual prognosis.
• Acknowledge uncertainty explicitly. "I cannot tell you exactly, and I will be honest with you as we learn more" maintains trust.
• Connect prognosis to goals. Once the patient knows the likely trajectory, ask what matters most with the time that is left — this is the bridge from prognosis to advance care planning. [1]

The Bernacki and Block synthesis of best practice in serious illness communication distils these principles into a structured, systematised approach that any clinician can learn [7][8].

---

Bereavement

Anticipatory grief

Anticipatory grief is the grief a patient and family experience before the death, in response to the losses that accompany a terminal illness — the loss of the future, of function, of role, of the person they knew. It is normal, and it is an opportunity: anticipatory grief, well supported, is associated with better post-death bereavement outcomes. The clinician's role is to normalise it, to listen, and to refer to psychosocial or bereavement services when the distress is disproportionate. [1]

Normal grief versus prolonged grief disorder

Normal grief is highly variable but tends to soften over weeks to months, with waves of sadness and longing that gradually integrate into a life that moves forward. Most bereaved people do not need formal treatment; they need acknowledgement, practical support, and time. [1]

Prolonged grief disorder (formerly complicated grief) is a distinct, diagnosable condition that affects an estimated 7 to 10 per cent of bereaved people and is more common after a traumatic or unexpected death, after the death of a child, and in people with a history of mood disorder or limited social support. Prigerson's validation work established the diagnostic criteria [10], and the disorder is now formally recognised in the DSM-5-TR and ICD-11.

The diagnostic features (DSM-5-TR): intense yearning or longing for the deceased, or intense preoccupation with them, present most of the day, nearly every day, for at least 12 months (6 months in children), at a level that exceeds cultural and developmental norms, and causing significant functional impairment. Accompanying features include disbelief or emotional numbness, identity disturbance, avoidance of reminders, intense emotional pain, difficulty reintegrating, and a sense that life is meaningless. [1]

Why it matters for the physician: prolonged grief disorder is not the same as major depression, it does not respond to standard antidepressants as first-line, and it responds best to specialised grief-focused therapy (complicated grief therapy). The physician who recognises it can refer appropriately. Risk factors that should prompt screening include a previous psychiatric history, the death of a child, a traumatic or violent death, social isolation, and high pre-death caregiver burden. [1]

Family support after death

The family's care does not end at the death. Good practice includes: [1]

• A dignified death — a peaceful environment, the family present if they wish, the body cared for with respect, and the cultural and religious practices around death respected.
• Communication of the death — in person where possible, with empathy and time.
• Bereavement follow-up — a condolence contact, an offer of a bereavement review, and referral to bereavement services for at-risk families. Specialist palliative care services routinely offer bereavement follow-up.
• Recognition of complicated grief — screen at-risk families at 6 to 12 months and refer for specialised therapy where prolonged grief disorder is suspected. [1]

---

High-yield DWE and DCE points

• The Surprise Question ("would you be surprised if this patient died in the next 12 months?") is the trigger for advance care planning. A "no" answer means: start the conversation.
• An advance directive's refusal of treatment is legally binding and must be honoured even after capacity is lost; ignoring it is battery.
• Withholding and withdrawing life-sustaining treatment are ethically and legally equivalent — start uncertain treatments as a trial.
• "Not for resuscitation" is a decision about CPR only; it is not the same as palliative care, and it does not mean withdrawing other active treatment.
• Anticipatory prescribing for the dying patient: morphine (pain, breathlessness), midazolam (agitation), glycopyrronium or hyoscine (secretions), levomepromazine (nausea, refractory agitation) — all subcutaneous.
• Morphine for refractory breathlessness works; routine supplemental oxygen for the non-hypoxaemic breathless patient does not (Abernethy) [5].
• A request for hastened death is usually a communication about suffering — explore and address it, involve palliative care, do not abandon.
• Prolonged grief disorder is distinct from major depression, presents at 6–12 months with intense yearning and functional impairment, and responds best to grief-focused therapy [10].
• Early palliative care improves quality of life and survival in advanced cancer (Temel) [4] — palliative care is not only for the imminently dying.
• The most common DCE long-case failure is leaving the goals-of-care and end-of-life discussion until a crisis forces it; the strong candidate raises it proactively.

---

Common exam traps and pitfalls

1. Treating advance directives as advisory rather than binding. A valid refusal of treatment must be honoured. The exam question that includes a valid directive refusing CPR and asks the "best next step" is testing whether you will honour the refusal even if the family objects.
2. Confusing "not for resuscitation" with "palliative care." They are independent decisions; a patient can have either, both, or neither.
3. Equating a goals-of-care discussion with a single form. The conversation is the intervention; the form is the record.
4. Using benzodiazepines for delirium in the dying patient outside alcohol or benzodiazepine withdrawal. They worsen delirium; an antipsychotic is the correct agent (see the delirium topic).
5. Treating the reduced oral intake of the dying phase with intravenous fluids. Artificial hydration in the last days may worsen pulmonary and peripheral oedema and respiratory secretions; the decision is made jointly and is often to reduce or stop.
6. Missing the spiritual and cultural dimension. Suffering at the end of life is not only physical; the candidate who ignores it loses marks.
7. Confusing the doctrine of double effect with euthanasia. Symptom-relieving doses of opioids correctly titrated are not euthanasia, and they do not shorten life.
8. Diagnosing prolonged grief disorder too early. Normal grief lasts months; the diagnosis requires the 12-month threshold (6 months in children) and functional impairment. [1]

---

Key references

Brinkman-Stoppelenburg A, Rietjens JAC, van der Heide A. The effects of advance care planning on end of life care: a systematic review. Palliat Med 2014;28:1000-1025 (PMID 24651708). Detering KM, Hancock AD, Reade MC, Silvester W. The impact of advance care planning on end of life care in elderly patients: randomised controlled trial. BMJ 2010;340:c1345 (PMID 20332506). Wright AA, Zhang B, Ray A, et al. Associations between end-of-life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment. JAMA 2008;300:1665-1673 (PMID 18840840). Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non-small-cell lung cancer. N Engl J Med 2010;363:733-742 (PMID 20818875). Abernethy AP, McDonald CF, Frith PA, et al. Effect of palliative oxygen versus room air in relief of breathlessness in patients with refractory dyspnoea: a double-blind, randomised controlled trial. Lancet 2010;376:784-793 (PMID 20816546). Baile WF, Buckman R, Lenzi R, et al. SPIKES — a six-step protocol for delivering bad news. The Oncologist 2000;5:302-311 (PMID 10964998). Bernacki RE, Block SD. Communication about serious illness care goals: a review and synthesis of best practices. JAMA Intern Med 2014;174:1994-2003 (PMID 25330167). Bernacki R, Hutchings M, Vick J, et al. Development of the Serious Illness Care Program. BMJ Open 2015;5:e009032 (PMID 26443662). Truog RD, Campbell ML, Curtis JR, et al. Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American College of Critical Care Medicine. Crit Care Med 2008;36:953-963 (PMID 18431285). Prigerson HG, Horowitz MJ, Jacobs SC, et al. Prolonged grief disorder: psychometric validation of criteria proposed for DSM-V and ICD-11. PLoS Med 2009;6:e1000121 (PMID 19652695). The SUPPORT Principal Investigators. A controlled trial to improve care for seriously ill hospitalized patients. JAMA 1995;274:1591-1598 (PMID 7474243). NICE Guideline NG142, Care of Dying Adults in the Last Days of Life (2015, updated 2021). Advance Care Planning Australia. Palliative Care Australia National Palliative Care Standards (5th ed, 2018). Victoria — Voluntary Assisted Dying Act 2017; Western Australia — Voluntary Assisted Dying Act 2019; New Zealand — End of Life Choice Act 2019. [1]

References

1. [1]Brinkman-Stoppelenburg A, Rietjens JAC, van der Heide A The effects of advance care planning on end-of-life care: a systematic review Palliat Med, 2014.PMID 24651708
2. [2]Detering KM, Hancock AD, Reade MC, Silvester W The impact of advance care planning on end of life care in elderly patients: randomised controlled trial BMJ, 2010.PMID 20332506
3. [3]Wright AA, Zhang B, Ray A, et al. Associations between end-of-life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment JAMA, 2008.PMID 18840840
4. [4]Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non-small-cell lung cancer N Engl J Med, 2010.PMID 20818875
5. [5]Abernethy AP, McDonald CF, Frith PA, et al. Effect of palliative oxygen versus room air in relief of breathlessness in patients with refractory dyspnoea: a double-blind, randomised controlled trial Lancet, 2010.PMID 20816546
6. [6]Baile WF, Buckman R, Lenzi R, Glober G, Beale EA, Kudelka AP SPIKES-A six-step protocol for delivering bad news: application to the patient with cancer Oncologist, 2000.PMID 10964998
7. [7]Bernacki RE, Block SD; American College of Physicians High Value Care Task Force Communication about serious illness care goals: a review and synthesis of best practices JAMA Intern Med, 2014.PMID 25330167
8. [8]Bernacki R, Hutchings M, Vick J, et al. Development of the Serious Illness Care Program: a randomised controlled trial of a palliative care communication intervention BMJ Open, 2015.PMID 26443662
9. [9]Truog RD, Campbell ML, Curtis JR, et al. Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American College [corrected] of Critical Care Medicine Crit Care Med, 2008.PMID 18431285
10. [10]Prigerson HG, Horowitz MJ, Jacobs SC, et al. Prolonged grief disorder: Psychometric validation of criteria proposed for DSM-V and ICD-11 PLoS Med, 2009.PMID 19652695
11. [11]The SUPPORT Principal Investigators A controlled trial to improve care for seriously ill hospitalized patients. The study to understand prognoses and preferences for outcomes and risks of treatments (SUPPORT). The SUPPORT Principal Investigators JAMA, 1995.PMID 7474243