Nursing · Professional practice, leadership and research
Write for somebody who has never met this patient, has four minutes, and will act on what you wrote
Nurses spend an enormous proportion of a shift recording things, and how that time is spent decides both the quality of the record and how much is left for patients — so it is clinical work rather than administrative overhead. Nobody is present for a whole illness, which means everything the next person knows comes from what was written, and a gap in the record is a gap in the care. Reviews of preventable harm repeatedly find information that was known to somebody and did not reach the person who needed it. And records are read years later in complaints, inquests and inquiries by people who were not there, for whom what was written is what happened. Electronic systems improved legibility, availability and prescribing safety while introducing new errors and new burdens, so neither uncritical enthusiasm nor refusal is a professional position.
Nurses spend an enormous proportion of a shift recording things, and how that time is spent determines both the quality of the record and how much is left for patients. It is not administrative overhead; it is clinical work.
Nobody is present for a whole illness, so everything the next person knows comes from what was written. A gap in the record is a gap in the care, not merely in the paperwork.
Reviews of preventable harm repeatedly find information that was known to somebody and did not reach the person who needed it. The record and the handover are where that failure happens or is prevented.
Records are read in complaints, inquests, inquiries and court cases, sometimes a decade afterwards, by people who were not there. What was written is what happened, as far as any of those are concerned.
Electronic records improved legibility, availability and prescribing safety, and introduced new errors, new burdens and new ways to be unsafe. Neither uncritical enthusiasm nor refusal is a professional position.
The primary purpose is that the next person caring for this patient knows what has happened, what matters, and what the plan is. Every other purpose is secondary to that.
Records written primarily to protect the writer become long, defensive and unusable, and they are worse evidence as well as worse clinical documents. Writing for the next clinician produces both good care and good protection.
Recording that a patient was reviewed says nothing; recording what was found, what it was thought to mean, and what would change the plan lets the next person act. Reasoning is the valuable part.
Doctors, therapists, pharmacists, social workers and the next shift all read nursing notes, and many of them will never speak to the author. Writing for a reader who has never met the patient is the standard.
Records increasingly belong to the patient in law and in practice, and are read by them. That is a reason to write with care rather than a reason to write less.
Ate half a meal is usable; appetite poor is not. Walked twenty metres with a frame and needed to stop once is usable; mobilising well is not. Specificity is what makes an entry clinical.
A record showing that breathing is faster than this morning, or that the wound is smaller than last week, describes a trend. A single figure with no comparison describes almost nothing.
Recording what a patient actually said, in their own words, preserves information that paraphrase destroys, and it matters particularly for pain, for symptoms and for disclosures.
The observation, the action taken, who was informed, what they said, and the effect. An entry naming a concern with no record of escalation or response is an incomplete and dangerous entry.
Long entries are not read. The discipline is deciding what the next person needs to know and writing that, which takes more thought and less time.
Difficult, demanding, attention-seeking, non-compliant, pleasant, uncooperative. These record the writer's reaction rather than the patient, they are read by the patient and by lawyers, and they change how colleagues treat somebody.
Declined the medicine, saying she had felt sick after the last dose is information. Non-compliant is a label that prevents anybody finding out why. The substitution costs nothing and changes the clinical picture.
Records are not the place for criticism of colleagues or accounts of disputes, which belong in incident reporting and management processes. Doing it in the notes damages everybody including the patient.
Assuming the patient will read it, which they may, is the most reliable test of whether an entry is appropriately written. It improves accuracy as well as tone.
Ambiguous abbreviations and symbols cause medication and clinical errors, and some have killed people. Local approved lists exist precisely because informal shorthand is dangerous.
A disclosure of abuse, violence or exploitation is recorded in the person's own words, factually, without the writer's opinion about whether it is true, because that record may be read in a legal process years later. It is also recorded with thought about who can see it, since a record visible to a family member can put somebody in danger.
In an investigation, care that was given and not documented is extremely difficult to demonstrate, and the presumption tends to run against it. This is unfair and it is the reality.
An entry made at the time, or as soon as possible, carries weight that one written days later does not. It is also far more accurate, because memory of a busy shift is poor.
Amending a record is done openly — the original remains legible, the correction is dated, timed and signed, and the reason is clear. Altering or obscuring an entry destroys the credibility of everything else the author wrote.
An entry written later is recorded as such with both times stated. Writing it as though it were contemporaneous is dishonest and is discovered readily in electronic systems.
The records that withstand scrutiny are specific, contemporaneous, honest about uncertainty, and show escalation. Defensive writing reads as defensive and helps nobody.
Times establish sequence, and sequence is what determines whether deterioration was recognised and how quickly it was acted upon. An undated or untimed entry loses most of its value.
An entry that cannot be attributed cannot be clarified, questioned or defended, and electronic systems solve this only if people do not share logins. Who wrote it matters.
Called the doctor at a stated time, what was said, when they attended, and what was decided. That chain is the single most important thing in the record of a deteriorating patient.
Hours with no entry in a patient who was unwell suggest either that nothing was done or that nobody looked, whichever is true. Recording at appropriate intervals is itself evidence of monitoring.
Recording observations, medicines or care before it happens is falsification, it happens under pressure, and it has directly caused deaths when the thing was then not done. It is never acceptable.
A medicine not given, a wash declined, an observation not taken. An unexplained gap is read as done by the next person, and that is how doses are duplicated and deterioration is missed.
A patient refusing something, and nobody offering it, are entirely different facts with different consequences, and the record must distinguish them. Refusal also warrants recording what was explained.
Patients away at theatre, imaging or on leave miss medicines and observations, and recording where they were explains the gap and prompts somebody to catch up. Silence looks like an omission.
Unable to weigh because the patient could not stand, unable to assess the heels because of a cast. That records an attempt and a limitation instead of leaving a blank that implies neglect.
An omission that matters is spoken about as well as written, because the next person may not read every entry. Both channels are used for anything consequential.
A patient nobody managed to review, weigh, wash or walk during a shift is recorded as such with the reason, because the honest entry is what shows a staffing problem and the blank one hides it. Care that was not given because there were not enough people is a system finding, and it disappears entirely if every chart reads as though the shift went normally.
Observations plotted so that direction is visible are what reveal deterioration, and figures recorded in a list without plotting hide it. The visual form is the clinical instrument.
A respiratory rate written as sixteen without being counted, or a copied-forward value, produces false reassurance and is a documented factor in missed deterioration. Counting matters more than recording.
Where an early warning score is used, calculating it wrongly or calculating it and not escalating both defeat it. The score's only value is the action it triggers.
How often observations are taken is decided by the patient's condition rather than by routine, and both over-monitoring a stable patient and under-monitoring an unstable one are errors. Recording the decision and its reason is good practice.
Whether the patient was asleep, agitated, walking, on oxygen, or in pain changes the meaning of the numbers entirely. Observations without context are frequently misinterpreted.
Drinks unrecorded, flushes ignored, urine estimated, and totals never added. A chart that is not totalled is not a fluid balance, and the clinical decisions that depend on it are then made on nothing.
A daily weight, done the same way, is a more reliable measure of fluid change than any chart, and the two together are strong evidence. Rapid weight change is water rather than tissue.
Illegible entries, ambiguous abbreviations, units written so they can be misread, unclear timings and unsigned administrations cause serious harm. Medication error is among the largest sources of avoidable harm in healthcare.
Signing in advance, or signing for a batch afterwards, breaks the only mechanism that prevents duplication. It is done at the bedside, at the time, for that dose.
The list in the record and the medicines in the house are frequently different, and the difference is the clinical finding. Asking somebody to bring everything, including things bought without prescription, resolves more than any test.
A care plan produced to satisfy an audit, filed and never consulted, is a waste of the hours it took. Whether staff actually use it is the only measure of its quality.
A plan identical to every other patient's tells nobody anything. The valuable content is what is specific to this person — how they transfer, what they will eat, what frightens them, what their normal is.
Plans made with the patient's goals in them are followed by the patient; plans made about them are complied with briefly. Asking what they want to be able to do produces a better plan and better effort.
A plan describing a patient as they were on admission, two weeks later, actively misleads. Reviewing and revising it is the work, and the initial writing is the easy part.
Pressure, falls, nutrition and clot risk assessments completed and not linked to any action are a documentation exercise. The score exists to trigger something.
More is transferred at handover than through any document, and a handover that lists tasks without risk, reasoning and plan passes on the work without the understanding. What could go wrong is the most valuable sentence.
A consistent format — situation, background, assessment, recommendation, or any agreed equivalent — prevents the gaps that appear when people improvise under pressure. Structure is a safety device.
Verbal handover is forgotten and misremembered, and a written summary that the receiver can return to is what survives the shift. Both are needed and neither replaces the other.
Handover conducted amid interruptions loses information, and protecting it is a recognised safety intervention rather than inflexibility. Most omissions occur at the point of interruption.
Handover from ambulance to department, department to ward, ward to ward, hospital to community, and adult to child services are where information is lost. Each boundary needs a deliberate structured transfer.
Omitting the duration, what has been tried, the current medicines and what the patient is actually worried about produces an appointment that achieves nothing. The specialist can only work with what was written.
Discharge summaries arrive late, incomplete, or not at all, and the community team, the general practitioner and the patient then work from guesswork. Timely accurate discharge information prevents readmission.
Medicines started, stopped and altered, with reasons, is the single most important content of a discharge document. Its absence causes patients to restart stopped medicines and stop essential new ones.
A letter written in plain language to the patient, or copied to them, improves understanding and adherence and finds errors the clinician missed. Patients frequently correct facts in their own letters.
Outstanding results, uncertain diagnoses, and things somebody must follow up. Unspoken loose ends are where results are lost and diagnoses are missed for years.
Incidents, near misses and unsafe conditions that are reported can be analysed and prevented, and those that are not will recur. Under-reporting is the normal state and it is the main obstacle.
A near miss carries the same information as a harmful incident at no cost to a patient, and it is the most under-reported category by far. Reporting one is a genuine contribution.
People do not report in environments where reporting leads to punishment, and reporting then stops while the errors continue privately. A just culture distinguishes human error, at-risk behaviour and recklessness, and treats them differently.
What happened, when, what was observed, and what was done, without allocating fault. The analysis belongs to the process rather than to the report.
Being open with patients when something goes wrong, explaining it, apologising and saying what will change, is a professional and in many places legal duty. Concealment causes far more damage than the original error.
What was explained, what alternatives were discussed, what risks were mentioned, what the patient asked and what they understood. A signed form with none of that documents a signature rather than consent.
Where capacity is in question, recording how it was assessed and what conclusion was reached, for that decision at that time, is what makes subsequent care lawful and defensible.
Clinical photographs require explicit consent covering what they will be used for — the record, teaching, publication — and consent for one is not consent for another. Images of wounds and skin are extremely useful and must be handled properly.
Clinical images taken on a personal phone are a serious confidentiality breach, they persist in backups and galleries, and this has led to disciplinary and regulatory action. Approved devices and systems exist for this.
A photograph without a face may still be identifiable by a tattoo, a ring, a room or a rare condition, and removing a name does not anonymise a case. This is a common and serious error in teaching material and portfolios.
Legible entries, records available simultaneously in several places, prescribing checks, allergy alerts, and the ability to find information and to audit. These are substantial and should not be dismissed.
Selecting the wrong patient from a list, the wrong drug from a dropdown, entering data in the wrong field, and errors propagating instantly across every screen. Digital records changed the error profile rather than removing it.
Tick boxes capture what the designer anticipated and discard everything else, and a patient's situation frequently does not fit. Free text still matters, and systems that remove it lose clinical information.
Typing while a patient talks, or facing a monitor rather than a person, measurably reduces what patients disclose. Positioning the screen so it can be shared, and putting it aside at key moments, is a clinical skill.
Whether a system is safe depends on how it is designed and configured — what is default, what is mandatory, what warns and what is easy to get wrong. Clinical involvement in that design is not optional.
Carrying forward yesterday's entry, or a previous assessment, propagates errors that then appear to have been confirmed repeatedly. A wrong statement copied ten times becomes extremely hard to correct or disbelieve.
An entry largely copied from yesterday buries whatever is new, so the reader cannot find the change. The change is the reason the entry exists.
Systems that warn about everything train people to dismiss warnings without reading them, and the important alert is then dismissed with the trivial ones. Too many alerts is a safety fault, not caution.
Alerts are frequently wrong or irrelevant, and clinical judgement properly overrides them, which is why recording the reason for an override matters. Blind compliance and blind dismissal are both unsafe.
Selecting a patient, a drug or a dose from a predictive list produces a specific class of error that did not exist on paper. A deliberate check of the name on the screen against the patient is the defence.
Power cuts, network failures, upgrades, and cyber attacks take records, prescribing, results and imaging away, and this has happened to large health systems. Planning for it is not pessimism.
Blank charts, drug charts, request forms and observation charts, in a known place, printed in advance. Deciding to find them after the screens go blank wastes the first hour.
Which patients are on what, who is due a critical medicine, and who is unstable, needs to be knowable without the system. A printed list at the start of each shift is a simple and effective measure.
Contemporaneous paper records made during an outage are entered into the system afterwards, labelled with the real times, and the paper retained. Reconstructing from memory afterwards is neither accurate nor honest.
An attack on health information systems cancels operations, loses results and diverts ambulances, and it is managed as a major incident rather than as a technical problem. Clinical staff are part of that response.
Sharing credentials destroys attribution, makes the record unattributable, and makes the person whose login it is responsible for everything done under it. It is also a disciplinary matter in most places.
Electronic systems log who looked at what and when, and those logs are examined. Looking up a colleague, a neighbour, a relative or a person in the news is misconduct and it is detectable.
Having technical access to a record is not authority to read it, and the test is whether you need it to care for that person. Curiosity is not a clinical justification.
An unattended logged-in screen exposes records to anybody passing and attributes their actions to you. It is the commonest confidentiality breach in every audit.
Phones, laptops and paper handover sheets taken home or left in cars have caused very large breaches. Knowing the local rules about what may leave, and following them, is part of the job.
Information is shared where the patient consents, where the law requires it, where it is necessary for their care, and where there is serious risk of harm. Knowing the grounds prevents both wrongful disclosure and paralysed silence.
Discussing a patient with the team caring for them is necessary; discussing them with anybody else is not, and that includes colleagues who are merely curious. The test is always whether it serves this patient's care.
Most breaches are conversational rather than technical — overheard discussions, a name said in a lift, an identifiable detail posted online. Nothing about a patient goes on social media, however anonymised it feels.
Family members asking about a patient have no automatic right to information, and the patient's wishes govern it. Establishing early who may be told what avoids a great deal of conflict.
Use the minimum necessary information, for a defined purpose, kept accurate and secure, for no longer than needed, and tell people what you do with it. Every data protection regime rests on those ideas.
A young person's confidentiality may be owed to them rather than to their parent, and a record visible to a parent can prevent a young person disclosing anything at all. Knowing what a parent may see, and telling the young person plainly, is a precondition of the consultation rather than an afterthought.
Patients have a right of access to their records in most jurisdictions, and increasingly see them in real time through online portals. Writing as though they will read it is now simply realistic.
Patients who read their records find errors — wrong medicines, wrong history, wrong family details — that nobody else would have found. That correction is a genuine safety benefit.
Results not yet explained, suspected diagnoses, and third-party information require thought about timing and wording rather than being hidden. Writing plainly and arranging to explain is better than writing obscurely.
Patients who read that they were described as difficult, obese, non-compliant or a poor historian remember it permanently, and it damages the relationship irreparably. This is now a practical reason as well as an ethical one.
When patients see something alarming, the answer is accessible explanation rather than restricting access. Portals without a way to ask a question generate anxiety.
A patient's information sitting in separate systems that cannot communicate means allergies unknown, tests repeated, medicines wrong and deterioration unrecognised. Most harm at service boundaries is an information problem.
A short shared summary of diagnoses, medicines, allergies and key risks is more useful than either no sharing or dumping an entire record. Usable is better than complete.
Where systems cannot communicate, giving the patient a written list of their conditions, medicines and allergies works, and in many settings it is the only thing that works. It is low technology and highly effective.
People are entitled to know what is shared with whom, and to object in many systems, and explaining it briefly builds the trust the system depends on. Covert sharing, once discovered, destroys that trust.
The same patient existing twice in a system splits their history, hides allergies and results, and causes serious errors. Checking identity carefully at registration prevents it.
Measurement changes what staff attend to, which is why badly chosen measures produce distorted care — activity recorded rather than performed, and problems reclassified rather than solved. Choosing measures is a clinical decision.
Diagnoses and procedures recorded determine what a service is paid, what is planned, and what appears in national statistics. Inaccurate recording misdescribes an entire population's health.
Comfort, explanation, reassurance, noticing, teaching and preventing are difficult to code and therefore invisible in data, which makes nursing appear cheaper to cut than it is. That invisibility has consequences.
A rise in recorded pressure injuries may mean worse care or better detection, and a fall may mean either improvement or under-reporting. Interpreting any indicator requires knowing how it is collected.
Adjusting recording to make figures look better corrupts the data everybody later relies on, and raising it when it happens is a professional responsibility. Corrupted data produces bad decisions for years.
Audit asks whether what we do matches what we agreed to do, identifies the gap, changes something, and measures again. Without that final step it is data collection rather than audit.
Most audits are completed, reported and produce no change, which wastes everybody's effort. Deciding in advance what will change if the result is poor is what makes it worthwhile.
Audits imposed from elsewhere about things nobody cares about produce compliance and no improvement. Auditing something that irritates the ward daily produces both.
Showing a team its own results, non-punitively and promptly, measurably changes practice. Publishing a report nobody reads does not.
A handful of records checked weekly with immediate feedback improves practice more than a comprehensive annual review. Frequency is what creates attention.
Prompts, checks, calculators and reminders prevent real errors, and heavy reliance on them erodes the underlying knowledge. Both effects are documented and both matter.
A tool calculating a dose, a risk score or a warning is using assumptions that may not apply to your patient. Using it without knowing what it assumes is unsafe.
Following a system's recommendation does not transfer responsibility, and neither does overriding it. Professional judgement is exercised either way and is recorded.
A calculator fed a wrong weight, a wrong creatinine or a wrong date produces a plausible, precise and wrong answer that people act on. The input check is the safety step.
A dose, score or plan that seems wrong for the person in front of you probably is, whatever the screen says. That instinct is a safety mechanism and should be voiced rather than suppressed.
Systems that flag deterioration, read images, draft documentation, triage messages and predict risk are in clinical use. Nurses will use them and be affected by them whatever their opinion of them.
Drafting notes and letters, transcribing conversations, summarising records and flagging results are the applications with the clearest benefit, because they return time to patient care. That benefit is real.
These systems produce plausible, fluent and incorrect output, including invented details, and they do not signal uncertainty reliably. Anything clinical generated this way is checked by a person before it is used or filed.
Pasting a patient's information into a general-purpose tool sends it outside the organisation and is a serious confidentiality breach. Approved systems with agreed data handling exist for this reason.
A nurse who acts on, signs or files output from one of these systems is accountable for it. Explaining that a tool produced it is not a defence.
A tool trained on historical records learns the patterns in them, including under-treatment of particular groups, and then reproduces it while appearing objective. The bias is inherited rather than invented.
Tools that have under-identified need in minority ethnic populations, pulse oximeters that read less accurately on darker skin, and risk scores developed in populations unlike the patient in front of you. These are measured findings.
A number produced by a system is trusted more than a clinical impression, which makes a biased number harder to challenge than a biased opinion. That is why the apparent neutrality is the hazard.
Groups who received less care appear in the data as having fewer recorded problems, which a system reads as lower need. Absence of records is not absence of illness.
A nurse noticing that a tool consistently underestimates a particular kind of patient has found something worth reporting. That observation is frequently the only route by which such faults are discovered.
Remote consultation reaches people who cannot travel, take time off or find childcare, and excludes those without devices, connection, credit, privacy or confidence. Both effects are large and they affect different people.
Without the ability to look properly, touch, measure or observe how somebody walks into a room, assessment is incomplete, and knowing what you cannot assess is the skill. Deciding when remote is inadequate is a clinical judgement.
Saying explicitly what should prompt them to seek help, by when, and how, compensates for what could not be assessed. Recording that safety net is essential.
Somebody may be unable to speak freely because a family member is present, which matters enormously for abuse, mental health and sexual health. Asking whether they are alone and able to talk is a specific necessary question.
Recording that an examination could not be performed, and what was done instead, is what makes the encounter defensible and tells the next clinician what remains unknown.
Home blood pressure readings, glucose logs, wearable heart data, symptom diaries and phone recordings. Much of it is genuinely useful and it changes the consultation.
Blood pressure measured at home over days is a better predictor than a single clinic reading, and continuous glucose data reveals overnight lows that testing never showed. Dismissing patient data loses real information.
Accuracy differs between devices and conditions, some are not validated for clinical use, and heart rhythm alerts from watches generate large numbers of false alarms. Knowing what the device actually measures matters.
People monitoring themselves continuously find variations that mean nothing and seek reassurance repeatedly, and some are then investigated unnecessarily. Helping somebody decide what is worth acting on is useful work.
Patient-generated data that is looked at and not recorded is lost, and the person concludes it was not valued. Summarising it in the record honours the effort and preserves the information.
Documentation time has grown substantially, much of it duplicated across systems and forms, and it is a leading contributor to nurses' dissatisfaction and to time away from patients. That is a clinical problem.
The same information entered in several places, forms that exist for assurance rather than care, and mandatory fields that nobody reads. Removing duplication is the highest-yield improvement available.
Each new requirement is individually reasonable and nothing is ever removed, so the total becomes unmanageable. Somebody must be willing to delete things.
Recording at the bedside as it happens is more accurate and faster than reconstructing at the end of a shift, and it requires devices and time that are frequently unavailable. That is a resourcing decision.
When documentation demands mean patients are not being observed, that is a safety concern to be raised through the proper route rather than absorbed. Raising it is professional conduct.
People without devices, data, connection, digital confidence, literacy in the interface language, or a fixed address cannot use online booking, portals, apps or remote consultation. They are also the people with the greatest health needs.
Older people, poorer people, homeless people, people with learning disability, sensory impairment or dementia, people in rural areas, and those who do not read the dominant language. That is a large fraction of any caseload.
A service reachable only through an app or a website has closed itself to those groups, and their non-attendance is then recorded as disengagement. Alternative routes are an equity requirement.
Sending appointment information by text or email, and assuming it arrived, causes missed appointments and lost results. Confirming how somebody prefers to be contacted takes one question.
Offering to help somebody register, showing them the portal, or simply telephoning them instead. Small practical accommodations decide whether a service reaches the people who need it most.
Writing in the wrong patient's record, or selecting the wrong name from a list, creates a false history in one record and loses the real information from another, and both persist. Checking the name and date of birth on the page or screen against the patient in front of you, every time, is the only defence and it takes two seconds.
Duplicate and similar names on one ward are a recognised cause of serious error, including wrong blood, wrong drug and wrong procedure. Where they exist, a flag on both records and an explicit alert at handover is warranted, and relying on a first name or a bed number is how it goes wrong.
Patients move, and a record or a sample labelled by bed rather than by identity follows the bed rather than the person. This is one of the commonest mechanisms by which results and specimens are attributed to the wrong patient.
An instruction taken by telephone is written down immediately, read back to the person who gave it, and signed by them as soon as possible. Acting on a remembered verbal instruction, or one passed through a third person, is where dose and drug errors originate.
An unidentified patient is given a temporary identifier that is used consistently across notes, samples, imaging and prescriptions, and merged carefully once their name is known. Improvised descriptions instead of a consistent identifier lose people inside the system.
What makes a defensible entry; never charting in advance; recording omissions and escalation with times; confidentiality grounds for sharing; not sharing logins; and the duty of openness when something goes wrong.
Charting before care is given; recording judgements of character; altering an entry without an audit trail; leaving unexplained gaps; assuming a relative is entitled to information; sharing a login; entering patient data into an unapproved tool; accepting a system's output without checking the input.
Specific, contemporaneous, timed, signed, and showing what was found, what was done, who was told and what followed. Record refusals and omissions with reasons. Escalate and record the escalation.
Total the fluid chart. Count the respiratory rate. Sign at the bedside at the time. Write their words in quotation marks. Log out. Print the list before the system goes down. Ask whether they are alone and able to talk.
Write for somebody who has never met this patient, has four minutes, and will act on what you wrote. That single test produces better records than any policy about them.
Written in plain sentences on purpose. Nursing is studied in a second language by a large share of the students who will read this, and a sentence that has to be decoded twice is a sentence that teaches less.
These are the twelve questions written with this SUNYRA nursing manual, to check what it teaches. They are not past papers and were not written to any one examination's pattern; an examination tag means only that the examination's published blueprint covers this subject. Difficulty was not graded in the source, so every question is marked medium.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.