Decision-Making Capacity, Surrogates and End-of-Life

Decision-making capacity is the clinical judgement that a person can make a specific decision at a specific time. It is the gateway to valid consent: without it, a signature on a form means nothing, and operating becomes, in law, a battery. For the orthopaedic surgeon this is not an abstract concern. The typical hip-fracture patient is elderly, frequently cognitively impaired, often delirious, and is being asked to consent to an operation within hours of admission. Capacity is the single most common professionalism and ethics issue on a fracture ward.

Three ideas frame the whole topic. First, capacity is presumed — a diagnosis of dementia, delirium, intoxication or learning disability does not by itself prove incapacity, and an unwise decision does not either. Second, capacity is specific — a patient may have the capacity to consent to a hemiarthroplasty but not to a complex revision, and may have it in the morning but not in the confused evening. Third, when capacity is genuinely absent, the patient is still owed a decision, and that decision must be made in their best interests, using their values, not yours.

The ethical scaffolding is the four-principles approach of Beauchamp and Childress — respect for autonomy, beneficence, non-maleficence and justice. When a patient has capacity, autonomy leads and the other three inform the options you offer. When they do not, autonomy is exercised retrospectively through their known wishes, and beneficence and non-maleficence govern the decision you make for them.

The four principles are the shared vocabulary an examiner expects whenever you analyse an ethics scenario. Autonomy is the patient's right to self-determination and is the reason consent exists at all. Beneficence is acting in the patient's interests. Non-maleficence is the duty to avoid harm — primum non nocere. Justice is the fair distribution of resources and the equal moral worth of every patient. These principles frequently pull against each other, and ethics is the disciplined work of weighing them in a specific situation.

Capacity is what unlocks autonomy. Two linked legal milestones define the modern standard for the consent that flows from it:

• Montgomery v Lanarkshire Health Board (UK Supreme Court, 2015) replaced the old "what would a reasonable doctor disclose?" test with a patient-centred one. A clinician must disclose any material risk that a reasonable person in the patient's position would regard as significant, judged by the risk's magnitude, nature and effect and by the characteristics of this particular patient. Consent is a dialogue, not a signature. A patient cannot weigh a material risk they were never told about, so Montgomery and capacity are inseparable.
• Bolam (1957) set the professional-practice standard for the clinical care itself — a doctor is not negligent if they act in line with a responsible body of opinion — and Bolitho (1997) refined it: that opinion must withstand logical analysis. When you act in a patient's best interests without their consent, the course you choose must be logically defensible, not merely "what some surgeons do".

The practical consequence is a two-part test for every elective and most emergency decisions. Has the patient got capacity for this decision, and were they given the information a reasonable patient would want? If either is missing, the consent is not valid, and you need a different pathway.

Capacity is assessed, not assumed. In England and Wales the Mental Capacity Act 2005 gives a statutory structure that examiners in many jurisdictions expect you to know, because it is widely mirrored. It has a two-stage test and four criteria of inability.

The two-stage test asks, first, whether there is an impairment of, or disturbance in the functioning of, the mind or brain (dementia, delirium, intoxication, head injury, learning disability, severe pain, medication, hypoxia), and second, whether that impairment makes the person unable to make the decision. Both stages must be met. A patient who is perfectly able to decide but happens to have dementia does not lack capacity for that decision.

A person is unable to make a decision if they cannot do any one of four things: understand the information relevant to the decision, retain it long enough to decide, use or weigh it as part of the process of deciding, or communicate the decision by any means. That gives four criteria — 4criteria of capacity — captured by the memory hook Understand, Retain, Weigh, Communicate.

How you assess it matters as much as what you assess. Start from the presumption of capacity, and only doubt it on evidence. Make the assessment decision-specific and time-specific — test the actual operation in front of you, not some abstract global competence. Optimise the conditions: assess at the patient's best time of day, with glasses and hearing aid in, with pain and reversible contributors to delirium treated, in a quiet room, using clear language and an interpreter if needed. Apply a sliding scale — the more serious or risky the decision, the more stringent the assessment, and the more carefully you document it. A decision to refuse a life-saving transfusion is held to a higher bar of understanding than a decision to have a minor wound cleaned, because the consequences are greater.

Two reversible traps are common on the orthopaedic ward. Delirium — present in a substantial proportion of elderly hip-fracture patients — can render capacity transiently absent and then return, so reassess after treating pain, infection, hypoxia, electrolytes, constipation, urinary retention and offending medications. And communication failure (untreated deafness, no interpreter, untreated pain) can mimic incapacity; exclude these before concluding the patient cannot decide.

When a patient cannot decide, three decision-making standards operate in a defined order. The examiner wants the order and the distinction between them.

Substituted judgement asks, "what would this patient have decided if they could?" It tries to reconstruct the decision the individual would make from their known values, prior statements and way of life. It respects the person's autonomy retrospectively. It is the dominant standard in United States law and a weighty factor elsewhere.

Best interests asks, "what decision is in this patient's overall welfare?" It weighs clinical benefit, the person's wishes, feelings, beliefs and values, and the views of those close to them. It is the standard named in the Mental Capacity Act 2005, with substituted judgement folded in as one of the factors the decision-maker must weigh.

Advance directives are the patient's own decision made in advance, while they still had capacity, to take effect when they later lose it. They sit above surrogates in the hierarchy because they are the patient's autonomous voice. An advance decision to refuse treatment (England and Wales, MCA) is legally binding if it is valid (made by a capacitous adult, without coercion, sufficiently informed) and applicable (clearly covers the circumstances that have arisen). To refuse life-sustaining treatment it must be in writing, signed and witnessed, and state explicitly that it applies even if life is at risk. The orthopaedic equivalent the examiner tests is the Jehovah's Witness who has signed a refusal of blood.

Surrogate decision-makers speak for the patient when there is no binding advance directive. The authority they hold depends on the jurisdiction. In England and Wales a Lasting Power of Attorney for health and welfare (a formal, registered instrument) lets a named person consent to or refuse treatment on the patient's behalf; otherwise no relative has an automatic right to decide, and the treating clinician makes the best-interests decision, consulting family as informants about the patient's values. In the United States, a healthcare proxy or durable power of attorney for health care does the same, and where none exists a default surrogate hierarchy (usually spouse, adult children, parents, siblings) legislated by each state steps in. For the unbefriended patient — no family, no advance directive, no attorney — England and Wales provide an Independent Mental Capacity Advocate (IMCA), and many systems route such cases through a clinical ethics committee or guardianship tribunal.

Two distinctions examiners probe. The terms capacity and competence are often used interchangeably but are not identical: capacity is the clinical judgement you make at the bedside, competence is the legal determination a court makes if the case is contested. And a surrogate consulted about a patient's values is not the same as a surrogate authorised to decide — the difference turns on whether a formal instrument (LPA, proxy, statutory hierarchy) gives them that power.

Orthopaedic end-of-life decisions cluster around the patient with metastatic or terminal disease — a pathological fracture, an impending fracture, or a frail patient in whom surgery will not achieve its goal. Three principles govern this ground.

Withholding and withdrawing treatment are ethically and legally equivalent. There is no moral difference between deciding not to start an intervention and deciding to stop one already begun; both turn on whether the intervention still serves the patient's goals. The distinction some clinicians feel — that stopping is harder than not starting — is psychological, not ethical. A treatment that has become futile is withdrawn; a treatment that would be futile is withheld; the reasoning is the same.

Futility has two senses, and only one is yours to judge. Physiological futility means the intervention cannot achieve its physiological goal — cardiopulmonary resuscitation that cannot restart a heart failing from advanced metastatic disease, or internal fixation that cannot mechanically hold in bone destroyed by tumour. This is a clinical judgement you can make and defend. Quality-of-life futility is a value judgement about whether the resulting life is worth living, and that is not yours to make unilaterally — it belongs to the patient (advance directive), the surrogate, or a best-interests analysis of the patient's own values. Conflating the two is a serious error.

End-of-life care is active care, not the absence of care. When surgery is not offered, the patient is owed excellent palliation: analgesia (including regional techniques), radiotherapy for painful metastases, an orthosis or splintage for comfort, nursing care, and psychosocial and spiritual support. Withholding non-beneficial surgery must never mean abandoning the patient.

Two practical instruments belong here. A decisions-about-cardiopulmonary-resuscitation order (known variously as DNACPR, DNAR, or, increasingly, as a ReSPECT recommended-summary plan in the UK) records whether resuscitation should be attempted and, more broadly, what treatment is recommended in an emergency. CPR in advanced metastatic cancer is almost always physiologically futile, and a decision not to attempt it should be made proactively, documented, and communicated to the patient (if they have capacity and wish to discuss it) and to those close to them. For the impending pathological fracture, Mirels' criteria score the risk of fracture from site, lesion type, size and pain to guide prophylactic fixation versus radiotherapy; an established fracture in a terminal patient shifts the goal from cure to comfort and the decision toward operative stabilisation only if it controls pain or aids transfer.

Ethical decisions live or die on communication. Two structured tools keep them safe. SPIKES — Setting, Perception, Invitation, Knowledge, Empathy, Strategy and Summary — is the standard protocol for breaking serious news and leading a goals-of-care conversation with a family; it stops you dumping information and lets the family's understanding and wishes shape the exchange. ISBAR — Identify, Situation, Background, Assessment, Recommendation — is the standard handover and escalation tool; use it to communicate a capacity determination or a best-interests decision so the next team inherits the reasoning, not just the conclusion.

The systems view matters because consent and capacity failures are rarely a single person's error. Reason's Swiss-cheese model describes how harm passes through the aligned holes of several defensive layers; a missing capacity assessment, an unsigned advance directive, a handover that drops the goals-of-care plan and an unchecked consent box can align to let an operation proceed that should not. The countermeasure is the WHO Surgical Safety Checklist (Haynes 2009), whose sign-in step forces the team to confirm patient identity, site, procedure and consent out loud — a checkpoint that the capacity-and-consent process was actually completed. Gawande's Checklist Manifesto makes the same point: complex, high-stakes work is made safe by simple, team-enforced checks, not by individual vigilance alone.

Four scenarios recur in orals and on the ward, and the examiner expects you to walk through them in the right order.

The cognitively impaired hip-fracture patient is the archetype. Presume capacity, assess it decision-specifically for the operation in optimised conditions, treat delirium and reassess. If capacity is present, the patient decides. If absent, decide in best interests: fixation of a displaced femoral neck fracture in a fit-enough elderly person is usually in their interests (pain relief, mobilisation, survival), but a terminal or imminently dying patient may be better served by palliation. Document the assessment and the reasoning.

The adult who refuses blood (most often a Jehovah's Witness) tests the boundary of autonomy. A capacitous, informed, contemporaneously reaffirmed refusal — or a valid advance decision refusing transfusion — is binding, and overriding it is battery. Your duty is to optimise within the constraint: tranexamic acid, cell salvage in a continuous circuit where acceptable, acute normovolaemic haemodilution, meticulous haemostasis, and damage-control fixation to minimise blood loss. Clarify the individual's specific boundaries, because they vary. Respect the refusal even at the cost of life.

The unconscious polytrauma victim engages the emergency exception. When a patient lacks capacity, no surrogate is available, and treatment is immediately necessary to save life or limb, you may treat without consent, acting in the patient's best interests — and you must document that delay to obtain consent would have caused harm. This exception does not extend to non-emergency interventions, which wait.

The adolescent patient engages the mature-minor doctrine. In England and Wales, Gillick competence holds that a person under sixteen who has sufficient understanding and intelligence may consent to treatment without parental involvement; the principle of a mature minor is recognised, with variations, across many jurisdictions. The capacity test is the same — understand, retain, weigh, communicate — applied to the young person in front of you.

Everything below condenses the topic for revision and viva practice — the high-yield points, three memory hooks, three worked vivas, and a one-screen cheat sheet.