The instrument you use more than any other
A nurse uses communication more often than any piece of equipment they will ever touch, and it is the one thing patients consistently remember afterwards. It is also where most complaints originate, where most missed information originates, and where most of the difference between a frightening admission and a bearable one is made. It is treated as a soft subject, taught early and briefly, and then assumed to develop on its own. It does not develop on its own. The specific things that work — asking openly, waiting through a silence, checking what was understood rather than whether it was understood, saying the difficult sentence plainly — are learnable, and most staff who are poor at them have simply never been shown.
Almost everything a nurse needs to know about a patient arrives through conversation: what hurts, when it started, what they took, what they understood, what frightens them, what they want. A nurse who cannot obtain that information reliably is working from an incomplete assessment however good their clinical knowledge, and the gaps are invisible because nobody knows what was never said. Calling this a soft skill has done real damage, because it places it outside the category of things people practise deliberately.
Communication failure appears repeatedly in analyses of avoidable harm: information lost at handover, a concern not escalated because of how the conversation would go, a patient who did not understand their discharge instructions, consent obtained from someone who did not grasp what was proposed, and a family who were never told. Each of these is a failure of speaking or listening rather than of knowledge.
People remember remarkably little of what was said to them in hospital and a great deal about how they were treated — whether somebody sat down, whether they were spoken about rather than to, whether anybody explained what was happening, whether they felt like a nuisance. That memory shapes whether they return, whether they follow the plan, and whether they trust the next nurse.
The specific behaviours that make a conversation work are identifiable and learnable: open questions, silence, not interrupting, checking understanding by asking the person to say it back, sitting at eye level. Most people who communicate poorly are not unkind. They have never been shown what the alternative looks like, and they are copying whoever supervised them first.
Clinicians interrupt patients very early in their opening statement, typically within the first half-minute, and once interrupted most people do not return to what they were going to say. Allowing an uninterrupted opening costs less time than is generally feared and frequently produces the piece of information the rest of the conversation would have spent ten minutes chasing.
A pause of a few seconds after somebody stops speaking is the single most productive thing in this manual, and it is extremely uncomfortable to hold. People fill silences with what they had decided not to say. The instinct to fill it yourself with reassurance, a question or a change of subject is the instinct to suppress.
Patients rarely state their real fear directly. They mention it sideways, near the end, as they turn away, or as a question about something trivial. The question about whether they will be home by the weekend is frequently a question about whether they are dying. Noticing the sideways approach, and gently opening it, is the skill this chapter is really about.
Repeating back the substance of what somebody said, in your own words, does two things: it proves you were listening, and it lets them correct you before the misunderstanding becomes a plan. It takes one sentence, it is the most reliable single technique here, and it is used far less than it should be.
Sitting rather than standing, being at eye level, not blocking the door, not folding your arms, and putting down what you are carrying all change how long a person talks and what they say. Patients consistently overestimate how long a clinician sat with them compared with standing for the same duration, which is a strange finding and a useful one.
Much advice on listening assumes time that a short-staffed ward does not have, and is dismissed for that reason. What survives compression is narrow and real: sit down even briefly, ask one open question, do not interrupt the first answer, and say when you will come back — and then come back. Four attentive minutes are worth more than fifteen distracted ones, and a patient can tell the difference immediately.
An open question invites an account — tell me about the pain — and a closed one invites a yes or no. Both are necessary, and the order matters: open questions first to find out what is there, closed ones afterwards to pin down detail. Beginning with closed questions produces an answer to what you already suspected and misses whatever you did not think to ask.
You are not in any pain, are you, is not a question. Leading questions are asked constantly, usually under time pressure, and patients — particularly those who are frightened, polite, or dependent on you — agree with them. This is the mechanism by which a patient in considerable pain is recorded as comfortable.
A question with three parts gets one answer, usually to the last part. This happens most often when somebody is rushing, and it reliably loses the first two pieces of information. Asking one thing and waiting is slower per question and much faster overall.
Sex, money, alcohol, violence at home, and thoughts of suicide are all asked about badly or not at all, and the avoidance is the clinician's rather than the patient's. Asking plainly, without apology or euphemism, and framing it as routine — I ask everybody this — makes an honest answer far more likely than a hesitant, apologetic approach.
Patients forget a large proportion of what they are told almost immediately, and the proportion rises with anxiety and with how much was said. The implications are practical: say the most important thing first, say less than you want to, write it down, and repeat it on another occasion rather than saying it all once thoroughly.
Positive, negative, chronic, acute, stable and observation all carry different meanings in ordinary speech than in clinical use, and patients frequently reach the opposite conclusion from the one intended. A positive result sounds like good news. Stable sounds like fine. Using plain words costs nothing and prevents a specific, well-documented category of misunderstanding.
Asking do you understand reliably produces yes, from people who do not want to appear foolish or to take more of your time. Asking them to explain it back in their own words — so I know I have explained it clearly, can you tell me what you will do when you get home — finds the gap. Framing it as checking yourself rather than testing them is what makes it acceptable.
A leaflet supports a conversation and does not replace one, and it is useless if it is in a language the person does not read, at a reading level they cannot manage, or handed over as they leave. Writing three sentences by hand, in their own language where possible, frequently beats a professionally produced booklet.
Saying what you are about to do before you do it, particularly anything involving touch, exposure or discomfort, costs a sentence and changes the experience entirely. It is the most commonly omitted thing in practical care, and it is what separates being cared for from being handled.
Before telling somebody anything serious, ask what they already understand. People frequently know or suspect a great deal, and starting from their own account avoids both repeating what they know and contradicting what they have been told. It also reveals the misunderstanding you would otherwise have built on.
People differ enormously in how much detail they want and when. Asking directly — would you like me to go through the details now, or would you rather I talk to you and your daughter tomorrow — hands a small amount of control back to somebody who has just lost a great deal of it.
Warn briefly that the news is not good, then say it in one short clear sentence, then stop and wait. The single most common failure is to keep talking through the silence — with explanation, reassurance and next steps — while the person has heard nothing after the first sentence. The silence is where they take it in.
People respond with anger, disbelief, tears, laughter, silence, or practical questions about parking. None of these are wrong and none require managing. Attempting to move somebody towards a more appropriate reaction, or reassuring them out of it, is the thing people most often regret afterwards.
Frequently a nurse is not the person who broke the news but is the one present afterwards, and for hours. That is where most of the real work happens: answering the same question repeatedly without impatience, finding out what was actually heard, and telling whoever gave the news what has and has not landed.
A signature records that a conversation happened; it is not the consent itself. Valid consent requires that the person has capacity for this decision, has enough information about what is proposed, the alternatives and the risks, and is deciding voluntarily. A form signed by somebody who did not understand what they were agreeing to has recorded nothing.
A patient lying down, undressed, with a queue of staff waiting and a procedure already set up is not in a strong position to decline. Saying explicitly that they may say no, or stop at any point, and meaning it, is what makes the consent real. It is easy to say and rarely said.
Consent is not only for operations. Washing somebody, moving them, giving an injection, taking a blood pressure and inserting a catheter all require the person's agreement, and the fact that they are in hospital does not supply it. Proceeding because somebody did not object is not consent, particularly with patients who are frightened or confused.
An adult with capacity may refuse anything, including treatment that will save their life, and may refuse for reasons others consider poor. The correct response is to check that they have the information, explore what is behind the refusal, record it, and respect it. Persuasion is legitimate; pressure is not, and the line is crossed more often than people think.
A person may understand everything and be unable to produce words, or produce fluent speech that makes no sense, or both. Assuming that difficulty speaking means difficulty understanding is a common and humiliating error. Speak normally and at normal volume, allow far more time, ask questions answerable with yes or no, and use writing, pictures or gesture as needed.
Face the person, make sure the light is on your face rather than behind you, speak clearly without shouting, and rephrase rather than repeating the same sentence louder. Check whether they have a hearing aid, whether it is in, and whether it is working, which is frequently the whole problem and is checked surprisingly rarely.
Short sentences, one idea at a time, no complex choices, no arguing with a mistaken belief, and attention to what the person is feeling rather than to whether the facts are right. Asking a person with dementia to orient themselves repeatedly is distressing and achieves nothing; joining them where they are usually calms the situation immediately.
Speak to the child as well as the parent, at their level, honestly, without promising that something will not hurt when it will. Children tolerate difficult truths considerably better than they tolerate being deceived, and a child who has been told a needle will not hurt will not believe anything else you say.
Hearing may persist, awareness is impossible to rule out, and families are listening. Explaining what you are doing, using the person's name, and not discussing them in the third person at the bedside are professional obligations rather than sentiment, and they matter to the relatives even when they do not reach the patient.
Some patients communicate through a board, a device, gesture, a communication passport written by the people who know them, or a signal that means yes agreed long ago with a relative. The professional obligation is to find out how this person communicates before concluding that they cannot, and to allow the very long pauses that a device requires without finishing their sentences for them. Speaking to the relative instead of the patient, in front of the patient, is the default that people fall into and the one that does most damage.
Using a relative to interpret means the patient cannot disclose anything about that relative, that the relative may soften or edit what is said, and that a child may be placed in an impossible position. It happens constantly because it is quicker, and it is the source of a great deal of missed information, particularly about violence, sex and money.
Speak to the patient and not to the interpreter, in short segments, in plain language, and allow twice the time. Look at the patient while they speak even though you understand nothing. Ask the interpreter to interpret rather than summarise, and check understanding through them as you would directly.
In many settings professional interpretation is simply unavailable. What remains is written phrases, pictures, translation applications used cautiously for simple concrete things, a bilingual colleague where confidentiality permits, and honesty in the record about the limitation. Recording that the history was obtained through a family member is important information for whoever reads it next.
Guessing somebody's language, religion or literacy from their name or appearance is wrong often enough to be worth avoiding entirely. Asking what language they would prefer, and whether they can read it, takes five seconds and avoids both offence and a leaflet that will never be read.
Culture is taught badly when it is taught as a set of facts about groups, because the result is a nurse who makes assumptions with more confidence than before. The useful stance is to ask this person what matters to them, since individuals vary enormously within any community and frequently differ from what a textbook says about it.
The assumption that the individual patient is the decision maker is culturally specific. In many families decisions are made collectively, or by a particular member, and a patient may genuinely want that. The professional obligation is to establish what this person wants rather than to impose either model, and to be alert to the difference between a patient's choice to involve their family and the family speaking over them.
In several cultures there is a strong expectation that serious news goes to the family before the patient, and this collides directly with legal and professional norms in other places. It is handled by asking the patient in advance who they want told and how much they want to know, which respects both the person and the obligation.
Everybody carries assumptions about class, accent, weight, age, religion, addiction and mental illness, and they alter tone, time given and how much pain is believed. The evidence that this affects care is substantial. It is not removed by deciding one is fair-minded; it is reduced by noticing it in specific moments and correcting for it deliberately.
An interpreter is a person in the room with their own reactions, particularly when the news is bad or the subject is distressing, and in small communities they may know the patient. Asking beforehand whether they are comfortable with the topic, warning them that the conversation will be difficult, and checking on them afterwards is both decent and practical, because a distressed interpreter interprets less accurately.
Anger from patients and families is most often fear, exhaustion, grief, powerlessness, or the accumulation of small failures nobody acknowledged. Responding to the emotion rather than to the words — I can see you are furious, tell me what has happened — defuses far more than explaining why the delay was unavoidable.
Standing over somebody, arguing facts, defending the organisation, saying calm down, walking away mid-sentence, and speaking to a colleague as though the person were not present all reliably make it worse. Recognising these as escalators rather than as reasonable responses is most of de-escalation.
An apology for the experience — I am sorry you have been waiting four hours with nobody explaining anything — is not an admission of liability and is what people are usually asking for. Withholding it out of fear of legal consequence is common, and the evidence broadly suggests openness reduces rather than increases complaints.
If a person is threatening or a situation is deteriorating, leaving is correct rather than a failure. Summoning help, removing yourself, and reporting it afterwards are all legitimate, and the notion that a good nurse can talk anybody down is both false and dangerous.
Every ward has a patient who is demanding, ungrateful or unpleasant, and their care measurably deteriorates — call bells answered more slowly, less time spent, symptoms believed less readily. Noticing it happening, and saying so within the team, is one of the more valuable things anybody can do, because the deterioration is collective and invisible to each individual.
Telling a patient that you got something wrong, kept them waiting through your own error, or forgot something you promised, is uncomfortable and almost always goes better than expected. A plain account, an apology without excuses, and what you will do now is the whole of it. What damages trust is not the error but the discovery that it was concealed, and patients are considerably more forgiving of mistakes than of being managed.
Most information loss happens at transitions, and structured handover exists because unstructured handover reliably omits things. The same structure used consistently is worth more than a better structure used occasionally, and the part most often dropped is what to watch for, which is the part the incoming nurse most needs.
A concern raised hesitantly is easy to dismiss. Stating the situation, the relevant background, your assessment and what you are specifically asking for gives the conversation a shape and makes it far harder to deflect. Having that structure available matters most for junior staff, at night, when they are least confident.
Verbal and telephone instructions are a known source of error because the listener hears what they expected. Repeating it back and having it confirmed takes seconds and catches a meaningful proportion, and it is skipped because it feels like doubting the other person.
Conversations in corridors, lifts and canteens are overheard constantly, and referring to patients by bed number, by diagnosis or by a nickname shapes how the whole team regards them. What is said about a patient when they are absent affects the care they receive when they are present.
A record is a message to a stranger who will read it under pressure, possibly years later, possibly in court, and possibly to the patient themselves. Written that way it looks quite different from a record written to discharge an obligation, and the difference is mostly in whether it says what somebody should actually do.
What was observed, what was done, what was said by the patient in their own words where it matters, what was reported to whom and when, and what the response was. Opinions about a person's character, motives or reliability do not belong there, and they follow that patient through every future admission.
Records written hours later are less accurate and, where something has gone wrong, look worse than they are. Writing something brief close to the event beats writing something thorough at the end of a shift, and a note recording exactly when an escalation was made is frequently the most important line in the chart.
Electronic systems make it easy to carry yesterday's entry into today, producing a record describing a patient who no longer exists. Copied documentation is worse than absent documentation, because the next person believes it and makes decisions on it.
The therapeutic relationship is warm, personal and asymmetrical: it exists for the patient's benefit, and it ends. Treating it as a friendship harms the patient, because a friendship cannot be handed over at the end of a shift and cannot be withdrawn when they are discharged without it being a loss they did not agree to.
Mentioning something about yourself can build trust and can also move the conversation onto your needs. The test is straightforward: whose benefit is this for. A brief, relevant disclosure that helps somebody feel less alone is fine; describing your own illness at length to a frightened patient is not.
Small gifts of thanks are usually fine and locally governed; money, valuables, lending or borrowing, and continuing contact after discharge are not. Social media contact with patients is a boundary breach that appears benign and reliably causes problems, and most employers now say so explicitly.
Working far beyond your hours for one patient, feeling uniquely able to care for them, keeping things from the team, or feeling resentful of colleagues' involvement are the recognised warning signs. Noticing them in yourself and saying so to a supervisor is a professional act, not an admission of failure.
On a telephone you lose colour, effort of breathing, posture, facial expression and the whole of what a person looks like, which is a large proportion of a nursing assessment. What replaces it is asking directly about the things you would otherwise have seen — can you finish a sentence, are you able to walk to the door, what colour are your lips, who is with you — and a much lower threshold for asking somebody to be seen in person.
Identity, and the caller's right to the information, must be established before anything is disclosed, and this is where most confidentiality breaches by telephone occur. A caller who knows the patient's name, date of birth and ward has not proved they are entitled to anything, and the safe route is to take a number and call back, or to ask the patient first whether this person may be told.
Remote conversations are interrupted by noise, poor lines and distraction, and they end abruptly. Putting the most important sentence at the start rather than building to it, and repeating it at the end, is how information survives a bad connection. The same applies to messages left with a third party, which should be short enough to be passed on accurately.
Every remote encounter ends with what would make you call back, who you would call, and by when — stated specifically rather than as a general instruction to seek help if worried. This is the whole safety mechanism when you cannot see the person, and it is the part most often compressed into a vague closing sentence.
Patients frequently want to talk about dying and find that everybody changes the subject, because staff fear removing hope, fear being asked something they cannot answer, and fear their own distress. The result is that a person spends their last weeks unable to say the thing most on their mind to anyone, surrounded by people being cheerful at them.
Am I dying is usually a genuine question and is usually answered with a deflection. A reasonable response is to find out what is behind it — what has made you ask that today — and then to answer honestly within what you know, including saying that you do not know and will find out who can tell them. Honesty does not remove hope; it relocates it.
Relatives repeatedly ask the same questions, and the repetition is not a failure of explanation but how people absorb something unbearable. They also need practical things nobody thinks to say: that hearing may persist, that reduced eating is part of dying rather than neglect, what the breathing will sound like, and that they may leave the room and it is not abandonment.
A great deal of this work is sitting with somebody without fixing anything, which is genuinely difficult for people trained to act. Being present, not filling the silence, and not offering reassurance that is not true are the whole of it, and staff who feel useless during it are frequently doing the most valuable thing available.
A large proportion of people in every country cannot reliably use written health information, and it correlates poorly with intelligence, education or how articulate somebody seems. Confident, well-spoken patients frequently do not understand what they have been told, and the only way to find out is to check rather than to judge from impression.
A person going home with a new device, dressing or medicine regimen does not need the whole subject. They need the next action, the sign that something is wrong, and who to call. Teaching everything results in nothing being retained; teaching three things results in three things, and the rest can be taught at the next contact.
Demonstrating and then watching the person perform the task themselves, that day, is the only reliable way to know whether they can. It routinely reveals that somebody who nodded throughout cannot open the packaging, cannot see the markings, or cannot manage it one-handed, none of which would have been discovered by explaining more clearly.
The person who will change the dressing at home is frequently not the patient but a daughter, a spouse or a neighbour who was not in the room. Establishing who that is and teaching them directly, rather than relying on the patient to pass it on, is one of the more consequential omissions in discharge.
Confidentiality is rarely breached by a deliberate disclosure. It is broken by conversations in corridors and lifts, by curtains that are not walls, by notes left visible, by screens not locked, by discussing a patient with a colleague who is not involved, and by telling a relative something the patient never agreed to share.
A conversation at a bedside is heard by everybody in the bay, and patients frequently withhold important information for exactly that reason. Recognising that a bay is a public place, and offering to move somewhere private for anything sensitive, changes what people are willing to say and is very rarely offered.
Family members have no automatic right to information about an adult patient, however involved they are and however reasonable the request. The patient decides, and asking them early — who may we speak to, and what may we tell them — prevents an impossible conversation later when a relative telephones and the patient cannot be asked.
Confidentiality is not absolute, and it yields where there is a risk of serious harm to the patient or to an identifiable other person, or where the law requires disclosure. The thresholds differ between countries, the decision is documented, and the disclosure is the minimum necessary to the people who need it rather than a general account.
Under acute pressure people speak in fragments, skip the context the listener needs, and hear only part of what is said back. This is physiological rather than a matter of composure, and the defences are structural: short sentences, one instruction at a time, naming the person you are speaking to, and requiring confirmation that it was heard.
An instruction is not complete until it has been repeated back and the completion reported. You, please take these bloods, tell me when they have gone — the third part is the one that is dropped, and it is the one that prevents everybody assuming a task was done because somebody was asked.
In a deteriorating situation the person most likely to notice something wrong is frequently the most junior person present and the least likely to say it. Naming the concern, stating what you observe, and if unresolved saying plainly that you are worried and need someone to look, is a sequence worth having ready in advance, because inventing it in the moment rarely happens.
When an emergency ends, the account given to the family, to the patient if they survived, and to the team, matters and is frequently rushed or delegated to nobody. Somebody should say plainly what happened, what was done and what happens next, and the person who led is usually the one who should do it.
Where what somebody said matters — a refusal, a complaint of pain, a statement about harming themselves, a description of how an injury happened — record it in their words inside quotation marks rather than paraphrasing it into clinical language. A paraphrase loses the thing that made it significant, and in any later review the difference between the patient said they felt unsafe at home and the patient said he will kill me if I go back is the whole of the matter.
Notes routinely record that a patient was informed, that consent was obtained, or that a family was updated, without saying by whom, when, or what was actually said. Such an entry cannot be relied on by anybody, and where something later goes wrong it is precisely the entry that will be examined. Naming what was discussed, and by whom, takes one extra clause.
A concern written in the notes and not spoken aloud has been recorded rather than communicated. Nurses frequently document something worrying and assume the next person will read it, and the next person is reading forty other pages. Anything that requires action is said to a named person as well as written down, and the record notes who was told.
Patients increasingly have access to their own records, and in several countries have a legal right to them. Words chosen in a hurry — non-compliant, poor historian, demanding, refuses to mobilise — read very differently to the person they describe, follow them through every future admission, and shape how the next clinician treats them before meeting them.
When an encounter has gone badly, saying so at handover, factually and without characterising the patient, allows the next nurse to start freshly rather than walk into something they do not understand. Saying nothing protects nobody, and describing the person as difficult guarantees that the next nurse arrives already braced for a fight.
In practical stations, a large share of the marks is for communication rather than for the procedure: introducing yourself, explaining what you will do, gaining consent, checking comfort, and closing properly. Candidates who perform the task flawlessly in silence lose marks they never knew were available.
Breaking bad news, obtaining consent, communicating with a patient who cannot speak or does not share your language, and responding to an angry relative. The expected answers are consistent: find out what they know, ask what they want, say it plainly, allow silence, check understanding by asking them to say it back, and document.
Asking do you understand instead of asking them to explain it back; using a family member as interpreter; reassuring somebody out of a reaction; treating a signature as consent; assuming difficulty speaking means difficulty understanding. Each is common in practice, which is exactly why each is examined.
Sit down. Let the first thirty seconds run uninterrupted. Wait three seconds after they stop. Say what you are about to do before you touch anybody. Ask them to tell you what they will do at home. None of these take long, none require permission, and together they are most of this subject.