What you may do, what you must do, and who answers for it
Eleven of the thirteen examinations in this library weight this subject, more than any other. That is not an accident of syllabus design: a regulator can remove a nurse from the register for an ethical failure in a way it rarely does for a gap in clinical knowledge, and the situations that end careers are almost never the dramatic ones.
Nurses are removed from registers far more often for conduct than for incompetence. The cases are rarely dramatic: a record written up to say what should have happened rather than what did, a confidence shared with a colleague who did not need it, a task accepted that was outside the person's scope because refusing felt rude, a concern not raised because somebody senior seemed sure. Every one of those is a decision made in a few seconds by somebody who knew better in the abstract, which is why this subject is examined so heavily and why studying it as a list of definitions wastes the opportunity.
Ethics is what you ought to do; law is what you may be compelled to do or punished for not doing. They overlap enormously and they are not identical. Something can be lawful and unethical — discharging a patient to an unsafe home because the bed is needed. Something can be ethical and unlawful — a practice permitted in one country and prohibited in the next. A nurse needs both frames, and needs to know which one they are reasoning in, because arguments that confuse them go nowhere.
It will be read in a great many countries whose law differs on consent, on the age of decision-making, on abortion, on end-of-life care, on reporting duties and on what a nurse may do without a doctor present. A manual that chose one jurisdiction and wrote as though it were universal would give confident, wrong answers to most of its readers. So this text teaches the principles, names the questions, and says plainly where the answer is national — and tells you to read your own code and your own statute, which you should have done anyway.
Four things hold across every system this manual will be read in. A competent adult's refusal is binding. Information about a patient belongs to the patient. A nurse is answerable for what they personally do, including what they accept from somebody else. And when something goes wrong, the patient is owed the truth. Everything else in this subject is elaboration on those four, and a student who can reason from them will get most examination questions right even in an unfamiliar jurisdiction.
A person has the right to decide what happens to their own body, including deciding against what you believe is best for them. Autonomy is the principle students find easiest to state and hardest to honour, because honouring it means watching somebody make a choice you think is wrong. It is not absolute — it does not extend to demanding a treatment that is not indicated, and it operates differently where a person lacks capacity — but within its range it wins. A great deal of poor nursing consists of quietly working around a decision rather than accepting it.
Doing good, and not doing harm. These are stated separately because they pull apart constantly: almost every treatment that helps also carries a risk, and the question is never whether harm is possible but whether the expected benefit justifies it and whether the patient agreed to that trade. Non-maleficence is the stronger duty of the two. A nurse who declines to act where the balance is unclear is on safer ground than one who acts because doing something feels better than doing nothing.
Like cases treated alike, and finite resources distributed by defensible criteria rather than by who complains loudest, who is known to staff, or who can pay. Justice is the principle most visibly violated in health systems everywhere and the one individual nurses have least power over — but not none. Who gets seen first in a queue, whose call bell is answered, whose pain is believed: those are allocation decisions made by nurses all day, and the evidence that they are made unevenly along lines of language, race, class and gender is substantial.
The four regularly point in different directions, and a framework that always gave one answer would not be a framework, it would be a rule. What the principles give you is a structured way to say what is actually at stake, which is the first step in any real ethical discussion. A nurse who can say that this is a conflict between the patient's autonomy and the family's understanding of beneficence has already made the conversation tractable. A nurse who can only say the situation feels wrong has not.
The International Council of Nurses publishes a code of ethics for nurses that has been adopted or adapted by national associations worldwide, and most countries' regulators publish a code of their own. These are not law in most jurisdictions, but they are the standard against which a regulator judges a nurse, which makes them binding in the way that matters to a career. They are short and they are written to be read. A student who has never read their own code has skipped the primary source of the subject they are being examined on.
Consent is valid when the person has capacity to give it, has enough information to decide, and gives it voluntarily. All three must be present. A signature on a form is evidence that a conversation happened; it is not consent, and a form signed by somebody who was not told what the procedure involves is worth nothing. The most common failure is the second element: information given in language the person did not follow, at a moment when they were frightened, with no opportunity to ask.
What the procedure is, what it is for, what it involves, what is likely to happen, what could go wrong including the risks a reasonable person in this patient's position would want to know, what the alternatives are, and what happens if nothing is done. That last one is omitted constantly and is often the piece a patient most needs. The test in many jurisdictions has moved from what a body of clinicians would disclose towards what this particular patient would want to know, which is a change in the direction of taking people seriously.
Consent obtained from somebody already on a trolley outside a theatre, or in front of a relative who has been answering for them, or from an employee whose job depends on agreeing, is questionable. Pressure need not be deliberate to invalidate it. Nurses are frequently the people who notice that a patient is agreeing because everybody in the room clearly expects them to, and raising that is a legitimate professional act rather than an obstruction.
Consent can be written, spoken or implied by conduct — a person who rolls up their sleeve for a blood pressure cuff has consented to that. Written consent is used where the intervention is significant, and its purpose is evidential. Who may obtain consent for a given procedure is set locally and usually requires somebody able to perform it or with specific training, which means a nurse asked to get a signature for a procedure they cannot explain should decline and say why.
Consent can be withdrawn at any point, including during a procedure, and the person withdrawing it does not have to justify themselves. Where stopping immediately would itself be dangerous, the correct response is to stop as soon as it is safe and to explain. Continuing because the procedure is nearly finished is a decision a nurse should be extremely reluctant to make and should never make silently.
Capacity is assessed for a particular decision at a particular moment. The functional test asks whether the person can understand the information relevant to the decision, retain it long enough to use it, weigh it against alternatives, and communicate a choice. A person may have capacity to decide about a dressing and not about surgery; may have it in the afternoon and not at three in the morning; may lose and regain it. Declaring somebody globally incapable is both clinically wrong and, in most jurisdictions, legally wrong.
An adult is assumed to have capacity unless there is reason to doubt it, and the doubt must be based on something other than the fact that the person disagrees with you. A diagnosis of dementia, learning disability, mental illness or intoxication raises the question; it does not answer it. The reverse error is equally common: a patient who is articulate and agrees readily is presumed to have understood, when they may have understood nothing at all.
Before deciding somebody lacks capacity, the obligation is to do what can be done to support them: use plain language, use their language and a proper interpreter rather than a relative, use pictures or writing, choose a time of day when they are at their best, treat pain, restore hearing aids and glasses, reduce noise, and involve somebody they trust. A very large share of apparent incapacity is unmet communication need, and a nurse is usually the person best placed to notice that.
Where a person genuinely cannot decide, jurisdictions differ substantially in who decides instead — a legally appointed representative, a family member, a court, a clinician acting in best interests — and in what weight is given to what the person would have wanted. What is common everywhere is that the decision should be the least restrictive option that meets the need, that the person's own previously expressed wishes carry weight, and that the person is still involved as far as they are able. Who holds the authority where you work is something to know before you need it.
Health information belongs to the patient. A nurse holds it because they need it to provide care, which defines both the duty and its limit: it is shared with those involved in that person's care, to the extent they need it, and with nobody else. The test worth carrying is not whether the person you are telling is a colleague but whether they need to know in order to care for this patient. A great deal of routine breach fails that test comfortably.
Corridors, lifts, handovers within earshot of other patients, phone calls at a nurses' station, screens left open, notes left face-up, and social media. Almost none of it is malicious and all of it is a breach. The cases regulators act on most often are the ordinary ones: a nurse who looked up a neighbour's record out of concern, a nurse who described a case online in enough detail to be identifiable, a nurse who told a family member something the patient had not chosen to share.
With the person's consent; where there is a serious risk to them or to an identifiable other; where the law requires it, which typically includes certain infectious diseases, some injuries and safeguarding concerns. The exceptions are narrower than students assume and the threshold is usually risk of serious harm rather than general concern. Where you disclose, you disclose the minimum necessary, to the right person, and you record what you disclosed and why.
A spouse, a parent of an adult child, an employer and a police officer are all, by default, third parties. Nurses are put under real pressure here, particularly where family involvement is culturally expected, and the professional answer is the same everywhere: ask the patient what they want shared and with whom, record the answer, and follow it. Where the patient cannot say, act on what is necessary for their care and on any previously expressed wish.
The duty of confidence generally survives the patient's death, which surprises students and matters in practice when relatives ask questions. Records remain confidential, subject to the access rights the law gives. And a record is a document that others will read: writing something in it that is speculative, judgemental or disparaging is both a professional failure and, in many systems, disclosable to the patient themselves.
It is a clinical tool for the next person caring for this patient; it is the evidence of what was done and why; and it is, eventually, the thing that will be read in an investigation. Those three purposes point the same way — contemporaneous, factual, complete — which is fortunate, because a record written for the third purpose alone reads defensively and is usually worse at the first.
A record written at the end of a shift from memory is less accurate, and everybody involved knows it. Times matter more than anything else and are the first thing lost. Where an entry is made late it is recorded as a late entry with the time of writing and the time of the events, which is honest and is expected; a late entry disguised as a contemporaneous one is falsification, and that is a register matter rather than a slip.
What you observed, what you did, what you were told and by whom, what you assessed, what you escalated and to whom and when, and what the response was. Not speculation about what a colleague was thinking; not judgements about a patient's character; not abbreviations your service has not agreed. A useful discipline is to write as though the patient will read it, because in many systems they can.
An error is corrected by a single line through it, leaving the original legible, with the correction, the date and your name. Obliterating, overwriting or removing a page destroys the record's value as evidence and reads to any investigator as concealment, regardless of the intention. Electronic records keep their own audit trail, which means an alteration is visible whether or not you intended it to be.
Students are taught that if it was not recorded it did not happen, which is a useful slogan and not quite true — evidence of care can come from elsewhere. What is true is that an undocumented action is very difficult to defend and that the absence of a record is itself evidence of something. The practical point is not to write more but to write the things that would matter later: times, assessments, escalations, responses, and the patient's own words about consent and refusal.
A nurse is personally accountable to the patient, to their employer, to their regulator and, in many systems, to a court. Those four are separate and an action can be acceptable to one and not another. The most important consequence is that following an instruction does not transfer responsibility: a nurse who carries out an order they knew or should have known was wrong is accountable for having carried it out. Every code of conduct in the world says this, and it is the single sentence in this subject that students most need to internalise.
Scope is what your qualification, your registration and your employer authorise you to do, and it differs enormously between countries and between levels of nurse. Working beyond it is a serious matter. Working needlessly below it is also a failure — a nurse who could have started an intervention within their scope and did not, because they were nervous, has harmed the patient just as surely. Knowing exactly where your boundary sits, on day one of a post, is professional basic hygiene.
Delegating a task does not delegate accountability for the decision to delegate. The person delegating must be satisfied that the task is appropriate to delegate, that the person accepting it is competent to do it, that they understand what is being asked, that supervision is available, and that the outcome will be checked. Delegating to somebody you have never worked with, because the shift is short-staffed, without any of that, is a decision you will be asked to justify.
The person accepting is accountable for accepting. A nurse who takes on something they are not competent to do, because refusing felt awkward or because the ward was under pressure, cannot later point at the person who asked. Declining is a professional act and is best done plainly and early: I have not been assessed for this, I will find somebody who has. Students find this the hardest single behaviour in the subject and it is worth rehearsing before it is needed.
In most legal systems, a claim in negligence requires a duty of care owed to the person, a breach of that duty, harm, and a causal link between the breach and the harm. All four are needed. A serious error that causes no harm is not negligence, though it may well be a regulatory matter; harm that would have occurred anyway is not caused by the breach. Understanding the four elements stops students treating every mistake as a lawsuit and every good outcome as a defence.
A nurse is judged against the standard of a reasonably competent nurse in the same role, not against the best nurse in the country and not against a specialist if you are not one. Inexperience is generally not a defence: a patient is entitled to competent care regardless of who happens to be on shift, which is why working beyond your competence and failing to ask for help are both so serious.
Medication administration; failure to monitor or to recognise deterioration; failure to escalate; failure to communicate a finding; falls; pressure damage; and record keeping. Note how many of these are failures to act rather than wrong actions. The image of negligence as a dramatic mistake is misleading; far more often it is a series of small omissions under pressure, each defensible alone, that add up to a patient nobody was watching.
Employers are generally liable for the acts of employees in the course of employment, which is why individual nurses are rarely sued personally. This is not a shield against everything: acting outside your scope, acting against instruction, or acting outside employment can fall outside it, and it never protects you from regulatory action, which is the sanction that actually ends careers. Indemnity arrangements differ by country and are worth understanding for your own situation.
When a patient has been harmed or could have been, they are owed an account of what happened, an apology, and an explanation of what will be done. Several regulators now make this an explicit professional duty. It is also the thing everybody's instinct resists, because it feels like an admission of liability — and in most jurisdictions an apology is not, and is protected by statute in some. A patient who is told the truth early is far less likely to pursue a claim than one who discovers it later.
Safety systems depend on people reporting errors and near misses, and people report only where doing so is safe. The evidence on this is unambiguous and the World Health Organization's patient safety work rests on it: services that respond to error by finding an individual to blame get fewer reports, not fewer errors, and the harm continues unmeasured. A student should know both that they must report and what a healthy response looks like, so they can recognise an unhealthy one.
Stop; make the patient safe; tell somebody senior immediately; document what happened factually; complete the incident report; cooperate with the review; and accept support. The temptation to wait and see whether the patient is affected is the one to resist, because the window in which harm can be prevented is short. Nurses who have made serious errors are at real risk of distress and of leaving the profession, and services that support them keep better nurses than services that do not.
The duty to raise a concern about unsafe practice is in every code. So is the fact that people find it extremely difficult, particularly about somebody senior, and particularly where the culture punishes it. The professional answer is to raise it anyway, through the route your service provides, in writing if informal raising fails, and to escalate outside the service if the risk continues. Protections for people who raise concerns exist in many countries and are uneven; the duty does not depend on the protection being reliable.
Children and adults at risk are harmed by people responsible for them, everywhere, and the health service is frequently the only place outside the home where they are seen by a professional. That makes recognition a duty of every nurse rather than a specialty. A nurse who decides safeguarding is somebody else's job has removed one of the few protections the person has.
Record what you saw and what was said, in the words actually used, at the time. Do not question the person closely or ask leading questions, which can distress them and damage any later process. Follow your local route promptly and tell the designated person. The duty is to pass an observation to a system designed to assess it, and it is not conditional on being certain — which is the reassurance students most need, because fear of being wrong is the commonest reason concerns go unraised.
Safeguarding is not only about children. Older people, people with disabilities, people with mental illness and people dependent on others for care are all at risk of neglect, financial abuse, coercion and physical harm, frequently from family and sometimes from paid carers or from services themselves. Institutional neglect — people left unfed, unturned, unwashed, without pain relief — is a safeguarding matter and not merely poor care, and recognising that changes what a nurse is obliged to do about it.
Where there is a safeguarding concern, the duty of confidence yields to the duty to protect, and every code and legal system provides for this. A nurse who stays silent because the patient asked them to has misunderstood both duties. Where the person has capacity and is an adult, their wishes carry substantial weight in what happens next, but the concern is still shared with whoever assesses it.
Many patients have recorded decisions about what they do and do not want, under names and legal forms that differ by country. Where such a decision is valid and applies to the situation, following it is delivering the care the person chose rather than withholding care. Where it cannot be found, is of uncertain validity, or does not match the situation, treatment proceeds while the question is resolved. Nobody should die because a document was in another folder; nobody should be treated against a clear, valid, applicable refusal because a nurse felt uncomfortable honouring it.
Deciding not to start a treatment and deciding to stop one are ethically and, in most systems, legally equivalent, though stopping feels far harder to everyone involved. Neither is the same as acting with the purpose of ending life, which is lawful in a small number of jurisdictions under strict conditions and unlawful in most. Giving treatment intended to relieve suffering that may also shorten life is a different thing again and is widely accepted. These distinctions matter; blurring them is how nurses end up frightened of providing adequate symptom relief.
When active treatment ends, nursing does not. Mouth care, positioning, cleanliness, pain and symptom relief, company, and care of the family are the whole of the work at that point and they are done well or badly. A student should understand that there is no such thing as nothing more can be done, only a change in what is being done, and that the change is towards work that is almost entirely nursing.
Expectations about who decides, what a dying person should be told, and what should happen to the body differ enormously between communities, and much of that variation is neither the nurse's business nor anything to correct. Where a family's wishes conflict with a patient's own expressed wishes, the patient's prevail. Where a family wants a patient not to be told their diagnosis, the answer is to find out what the patient wants to know rather than to accept or refuse the request outright.
Most systems allow a nurse to decline participation in certain specific procedures on grounds of conscience, and the scope of that right is narrow and national. Where it exists, it is generally conditional: the objection must be declared in advance rather than at the bedside, care cannot be abandoned, the patient must be referred to somebody who will provide it, and emergency care is never covered. A nurse who lets a patient suffer while asserting a conscience right has misunderstood it everywhere it exists.
Knowing what the right action is and being prevented from taking it — by staffing, by hierarchy, by policy, by resources — is a recognised and specific kind of harm to the person experiencing it, distinct from ordinary stress. It accumulates, it is a major driver of nurses leaving, and it is largely produced by systems rather than by individuals. Naming it accurately matters, because a nurse who believes they are simply failing to cope will not seek the right help or press for the right change.
Discharge a patient who is not ready because the bed is needed. Record an observation nobody took. Accept a task you are not competent for. Witness a signature you did not see given. These are the real ethical tests of a career and they arrive quietly, from people who are not villains, in units that are short-staffed. Deciding your answer in advance, while calm, is the single most useful preparation this subject offers.
Raise it early, factually and in writing where informal raising fails. Keep your own dated notes. Use the route your service provides and escalate if it stalls. Find one colleague you trust. Know what protections exist where you work and do not assume they are reliable. None of this makes it easy, and this manual will not pretend that in every workplace the person who raises a concern is thanked. The duty stands anyway, and so does the fact that a patient is relying on somebody saying something.
A licence to practise, granted on evidence of education and character, maintained by continuing requirements, and removable. In most countries practising without it is a criminal offence and so is claiming to hold it. Registration is what makes nursing a profession rather than a job, and the price of it is being answerable to a body that can end your career on the balance of probabilities rather than beyond reasonable doubt.
Regulators act on concerns about clinical failings, misconduct, health, and convictions — and on dishonesty above all, because a regulator that cannot believe a nurse's account of events cannot regulate them at all. Outcomes range from advice through conditions and suspension to removal. Students should understand that conduct outside work can be in scope where it bears on public confidence, and that the process is usually slower and more distressing than the outcome suggests.
Continuing professional development, revalidation or relicensure requirements, and a duty to notify the regulator of matters that affect fitness to practise. These differ by country in their detail and are identical in their purpose. A nurse who treats them as paperwork discovers their importance at the worst possible moment.
An internationally mobile nurse is subject to the law and the code of the country they are working in, not the one they trained in, and the differences are frequently large — in scope of practice, in what may be done without a doctor, in consent and in reporting duties. Reading the new code before the first shift is not optional, and a great many of the fitness-to-practise cases involving internationally recruited nurses turn on a rule nobody told them about.
Where there are fewer staff than patients, fewer beds than admissions and fewer supplies than needs, allocation decisions are being made constantly. The ethical question is not whether to ration but whether the criteria are explicit and defensible. Decisions made by default — whoever complains loudest, whoever a staff member knows, whoever arrived with a relative who can advocate — are still decisions, and they are the least just kind.
A ward nurse cannot fund a hospital. They can decide the order of the queue on stated clinical grounds, record and escalate the shortfall so it becomes visible, refuse to normalise a gap by working around it silently, and treat everybody in front of them with the same seriousness regardless of who they are. Written records of unmet need are what eventually move budgets; silent heroism is what allows a shortage to persist for a decade.
Some readers will work where care depends on payment, where certain patients are treated worse for who they are, or where instructions come from people with power and no clinical basis. The professional standard does not change, and neither does the reality that meeting it can be personally costly. What this manual can honestly say is that the duty is to the patient in front of you, that documentation is the tool available to almost everyone, and that the profession's codes exist precisely so that a nurse under pressure has something to point at that is not merely their own opinion.
Nurses in low-income settings are frequently asked to collect data, host studies or implement programmes designed elsewhere. The ethical questions are the ordinary ones — was consent genuinely informed, does the community benefit, is the same standard of care offered as would be offered elsewhere, what happens when the study ends — and the answers are not always good. A nurse is entitled to ask them and should.
Most are scenarios ending in what should the nurse do next, and the correct answer is usually the one that is within scope, puts the patient's safety and autonomy first, and involves telling somebody rather than deciding alone. Options that describe acting outside scope, keeping something quiet, overriding a competent refusal, or reassuring a patient in place of informing them are distractors. Reading the options as actions you would have to defend afterwards is a reliable filter.
Assuming a diagnosis removes capacity. Treating a signed form as consent. Believing that following an instruction transfers accountability. Thinking confidentiality is absolute, or that it has no exceptions. Confusing withholding treatment with ending life. Answering from your own country's law when the paper is set elsewhere — or, more often, from a general principle when the paper wants the local rule. Knowing which of those applies to your paper is half the preparation.
The three elements of valid consent. The four parts of the functional capacity test. The four elements of a negligence claim. The circumstances in which confidentiality may be broken. The conditions that make a delegation appropriate. These five lists are short, they recur in every paper, and they are the scaffolding for reasoning about a scenario you have never seen.
That a competent adult's refusal binds you. That information about a patient belongs to the patient. That accepting a task makes you accountable for having accepted it. That a patient harmed is owed the truth. And that the situations which end careers are quiet ones, arriving on ordinary shifts from people who are not villains — which is why the time to decide what you will do is now, while nothing is at stake.