Recognising the patient who is getting worse
Medical-surgical nursing is where most nurses in the world work and where most examinations put most of their marks. The competency it tests is not knowledge of one body system but the ability to hold several unwell people in mind at once and correctly identify which of them is deteriorating.
Medical-surgical nursing is the care of adults with acute and chronic illness, and it is where most nurses in the world work. Its difficulty is not that any single condition is hard, but that the field has no natural boundary: a single ward holds a patient recovering from bowel surgery, a patient in heart failure, a patient whose diabetes has decompensated and a patient waiting on a diagnosis nobody has made yet. The competency being examined is not knowledge of one system but the ability to hold several unwell people in mind at once and correctly identify which of them is deteriorating.
On NCLEX-RN, Physiological Integrity is the largest category at 43 to 67 per cent of the paper, and most of that is medical-surgical content. On NORCET Mains the emphasis is explicitly on case-scenario questions testing nursing competence rather than recall. The reason is not that examiners like this subject. It is that this is where a registered nurse will actually spend their working life, and where failure to recognise deterioration kills people who were expected to recover.
A recurring confusion in students is that medical-surgical nursing means learning to diagnose. It does not. The nurse's competency is recognition and escalation: noticing that something has changed, describing it accurately, and getting the right person to the bedside fast enough. A nurse who correctly identifies that a post-operative patient is tachycardic, cold, and producing less urine has done the whole job even if they never name the cause — because the actions that follow are the same regardless of whether it turns out to be bleeding, sepsis or dehydration.
This manual is organised around the reasoning a nurse uses at the bedside rather than around body systems, because body systems are how disease is taught and deterioration is not experienced that way. It begins with assessment, moves to the physiology of deterioration, then to fluid and electrolyte reasoning, then to the perioperative journey, then to the chronic conditions that fill most beds, and finally to handover — which is where more information is lost than anywhere else in a hospital.
Experienced nurses often say they knew a patient was unwell before the observations showed it, and that is genuinely true — but it is not teachable and it is not reliable under fatigue. A systematic assessment is what makes recognition reproducible: the same sequence, in the same order, every time, so that nothing is skipped because it did not occur to you. The sequence most widely taught is ABCDE — airway, breathing, circulation, disability, exposure — and its value is that it is ordered by how quickly each problem kills.
The airway is first because an obstructed airway kills in minutes. The assessment is direct: can the patient speak in full sentences? A patient who answers a question clearly has a patent airway, and that single exchange has completed the first step. Signs of obstruction include stridor, gurgling, snoring respiration in a drowsy patient, see-saw chest movement, and silence in a patient who is visibly distressed. Silence with distress is the most dangerous of these, because the absence of sound is easily read as calm.
Breathing is assessed by rate, effort, symmetry and saturation. Respiratory rate is the single most useful observation in the whole set and the one most often recorded inaccurately, because it is the only one requiring the nurse to stand and count rather than read a machine. A rising respiratory rate is among the earliest signs of deterioration in almost every acute illness, frequently preceding a change in blood pressure by hours. A nurse who records a rate of twenty because that is what was written last time has removed the most sensitive early warning available.
Circulation is assessed by heart rate, blood pressure, peripheral perfusion, capillary refill and urine output. The important teaching point is that blood pressure is a late sign. A young, previously well patient compensates for significant blood loss by increasing heart rate and constricting peripherally, holding blood pressure normal until compensation fails — at which point it falls suddenly and steeply. A normal blood pressure in a patient who is tachycardic, cold at the fingers and not passing urine is not reassurance; it is compensation that has not yet run out.
Disability means neurological state: level of consciousness, pupils, and blood glucose, which is checked in any patient with altered consciousness because hypoglycaemia is rapidly reversible and rapidly fatal if missed. Consciousness is described using a scale rather than a word, because 'drowsy' means different things to different people while a recorded score does not. Exposure means looking at the patient — the wound, the abdomen, the calves, the back, the skin — because a great many findings are visible and are missed simply because nobody uncovered the patient.
Most health systems now aggregate observations into an early warning score, where each observation outside a normal range scores points and the total drives a defined response. The value of these systems is that they convert individual judgement into a mandated action, which protects a junior nurse who is worried but unsure whether escalation is justified. The known weakness is the same one as barcode scanning: a score that is calculated but not acted on, or observations timed to suit the round rather than the patient, produces a documented reassurance that nobody has actually checked.
Deterioration is usually a trend rather than an event. A patient whose heart rate has climbed from 78 to 96 to 112 across three sets of observations is deteriorating even though 112 alone might not trigger anything, while a patient whose heart rate has been 110 all week is simply a patient with a heart rate of 110. This is why observations are charted rather than merely recorded: the chart exists to make the direction of travel visible, and a nurse who reads only the latest figure has thrown away most of the information the chart holds.
Sepsis is the deterioration pattern every nurse must recognise because it is common, rapidly fatal, and treatable if caught. The signs are non-specific individually and compelling together: fever or an unexpectedly low temperature, tachycardia, tachypnoea, confusion or reduced consciousness, reduced urine output, and mottled or discoloured skin. In older patients confusion may be the only early sign, and in very young children a weak or high-pitched cry, poor feeding and difficulty rousing carry the same weight. The nurse's contribution is speed: recognising the pattern and escalating it, since delays in recognition translate directly into mortality.
Internal bleeding after surgery is frequently missed because the patient looks stable until they do not. The pattern is a rising heart rate, a narrowing pulse pressure, cool and clammy peripheries, restlessness or anxiety, and falling urine output — with blood pressure maintained until late. Restlessness deserves particular attention because it is easily attributed to pain or anxiety and medicated rather than investigated, and a patient sedated for restlessness caused by hypovolaemia loses the one sign they were still able to give.
Certain groups deteriorate without producing the expected signs. Older patients may not mount a fever with infection and may present only with confusion or a fall. Patients on beta blockers cannot mount a tachycardia, so the compensatory sign the nurse is watching for never appears. Patients with diabetes may have reduced pain sensation and can have a myocardial infarction with little or no chest pain. A nurse who knows only the textbook presentation will miss these entirely, which is why the question to keep asking is not whether the signs match a pattern but whether this patient is different from how they were earlier.
Fluid and electrolyte management defeats more students than any other part of medical-surgical nursing, and the reason is that it cannot be memorised. The same sign means opposite things in different patients: a raised heart rate might mean too little fluid or too much, depending on whether the heart is failing. What has to be learned is a way of reasoning — where is the fluid, where should it be, and what is the body doing to compensate — rather than a table of values.
A fluid balance chart is only as good as the discipline behind it, and it is among the least reliably completed documents in a hospital. Intake includes oral fluids, intravenous fluids, flushes and fluids given with medications, which are routinely omitted and add up. Output includes urine, drains, vomit, stoma output and significant wound leakage, and the last three are routinely estimated generously or not at all. A chart showing a neat positive balance in a patient who is visibly dry is not evidence about the patient; it is evidence about the chart.
Daily weight, taken at the same time on the same scales in similar clothing, is the most reliable indicator of fluid gain or loss available on a ward, because a litre of fluid weighs a kilogram and body tissue does not change that fast. A patient who has gained three kilograms in two days has retained roughly three litres, whatever the fluid chart says. This is why daily weights are mandatory in heart failure and why skipping them because the patient is tired removes the single best measurement being taken.
Reduced circulating volume produces a predictable sequence: thirst, reduced urine output with concentrated urine, tachycardia, cool peripheries with delayed capillary refill, reduced skin turgor, dry mucous membranes, and only late a falling blood pressure with postural drop. The sequence matters more than any single item, and the nurse's specific skill is noticing the early items in a patient whose blood pressure is still normal — because that is the window in which correction is straightforward.
The opposite picture appears in patients whose heart or kidneys cannot handle the volume they have been given, and it is often iatrogenic — caused by the treatment. The signs are breathlessness worse on lying flat, crackles at the lung bases, raised jugular venous pressure, peripheral oedema, rapid weight gain and falling oxygen saturation. A post-operative patient who becomes breathless overnight after generous intravenous fluids is a common and dangerous presentation, and the harm is done by fluid prescribed with good intentions and not reviewed.
Potassium sits in a narrow range, and both directions are dangerous to the heart. Low potassium produces weakness, cramps, constipation and arrhythmias, and is commonly caused by diuretics, vomiting and diarrhoea. High potassium may produce almost no symptoms until it produces a fatal arrhythmia, which is why it is found on a blood test rather than at the bedside, and why a result showing it is treated as urgent rather than filed. Patients with kidney impairment are at particular risk, and any patient on a potassium-sparing medication whose renal function has worsened deserves attention before anyone else raises it.
Pre-operative preparation exists to make sure the right operation happens to the right patient on the right side, and that the patient arrives in the best condition achievable. That means confirming identity and consent, marking the site, establishing fasting status, checking allergies, reviewing which medicines are to be withheld and which continued, removing jewellery and prostheses, and ensuring baseline observations exist so that post-operative changes can be compared to something. Baseline observations are frequently treated as paperwork, and they are the reference against which every later number is judged.
This is a common examination area and a common real-world error. Some medicines are withheld before surgery because they increase bleeding or interact with anaesthesia, and others must continue because stopping them is more dangerous than the operation. This manual names no drug and no timing, because both are specific to the patient, the operation and the local protocol. What the nurse must know is that the decision is made deliberately by the prescribing team and recorded, and that a medicine withheld without a recorded decision is an error rather than a precaution.
The surgical safety checklist is performed at three points — before anaesthesia, before the incision, and before the patient leaves the theatre — and its effectiveness comes from being spoken aloud by a team that pauses. The most valuable element is the team introduction, because a scrub nurse who has said their name out loud is measurably more likely to speak up later, and the errors this checklist prevents are usually errors somebody in the room already suspected. Swab, instrument and sharps counts are performed and reconciled, and an unreconciled count stops the operation from closing.
The immediate post-operative period is about airway, breathing, circulation, pain, and the specific complications of the operation performed. Observations are frequent and then spaced as the patient stabilises. The findings that must be escalated rather than watched are: a rising heart rate with falling urine output, which suggests bleeding; increasing pain in a limb or compartment, which may be compartment syndrome and is a surgical emergency; and any drop in consciousness or oxygen saturation. Nausea, pain and shivering are common and treatable; none of them explains a rising heart rate.
Post-operative complications arrive on a rough schedule, and knowing it changes what a nurse looks for on a given day. Atelectasis and early breathlessness appear in the first day or two and are addressed by sitting the patient up, deep breathing and early mobilisation. Infection at the wound tends to appear from around the third to fifth day. Venous thromboembolism risk runs across the whole recovery and beyond discharge, which is why prophylaxis, early mobilisation and calf observation continue after the patient feels well. Anastomotic leak in bowel surgery classically presents later than expected, often with tachycardia and unease before anything localising.
Heart failure is among the commonest reasons for repeated admission, and most readmissions are preventable. The nursing contributions are specific and unglamorous: daily weights taken properly, accurate fluid balance, positioning the breathless patient upright, monitoring for the fluid overload signs described earlier, and teaching the patient to weigh themselves at home and act on a defined gain. A patient discharged without understanding their own weight threshold will return, and that teaching is nursing work rather than an optional extra.
The nursing priorities are recognising an exacerbation early, supporting breathing without removing the drive to breathe, and understanding that oxygen in this group is a drug with a target range rather than a comfort measure. Patients are often frightened, and fear worsens breathlessness in a loop that responds to calm presence and positioning as much as to medication. Smoking cessation support, inhaler technique checked rather than assumed, and vaccination status are the interventions that actually change the number of future admissions.
Almost every ward holds patients with diabetes, and hospital admission disrupts everything that normally keeps it stable: meals are delayed, patients are fasted, illness raises glucose, and steroids raise it further. The nurse's role is monitoring, recognising hypoglycaemia and treating it immediately, recognising the far slower onset of hyperglycaemic emergencies, and noticing when a fasted patient is still receiving glucose-lowering medication. That last one is the single most common serious diabetes error in hospital and it is entirely preventable by someone reading the chart and the fasting status together.
Impaired kidneys change how nearly every other problem behaves: fluid is harder to shift, potassium accumulates, and many drugs are cleared more slowly and reach toxic levels at ordinary doses. The nursing observations that matter are accurate fluid balance, daily weights, blood pressure, and awareness of which medicines are renally cleared. A nurse who notices that renal function has fallen while a renally-cleared medicine continues unchanged has prevented a harm that would otherwise have been discovered by its consequences.
Chronic wounds — pressure ulcers, diabetic foot ulcers, venous leg ulcers — are largely nursing-led and largely preventable. Pressure damage is caused by pressure, shear and moisture over time in a patient who cannot reposition themselves, and the interventions are repositioning, surface selection, skin inspection, nutrition and continence management. The most important point for a student is that a pressure ulcer is generally a failure of care rather than an unavoidable consequence of illness, which is why it is reported and investigated as harm.
The foundational principle is that pain is what the patient says it is. This sounds like a platitude and is in fact a corrective, because the documented pattern in healthcare is that some patients are believed less than others — and the groups believed least are those least able to argue. Pain is systematically under-treated in older patients, in people with dementia, in people who do not share the clinician's language, and in patients with a history of substance use. A nurse who notices that they are more sceptical about one patient's pain than another's has noticed something important about their own practice.
A numerical scale is useless in a patient who is confused, intubated, very young or severely cognitively impaired, and their pain is no less real. Behavioural assessment replaces self-report: facial expression, guarding, restlessness, changes in breathing, vocalisation, resistance to movement, and a change from the person's own baseline. The last is the most reliable and the one only a nurse who knows the patient can use — which is a strong argument for continuity of care that has nothing to do with sentiment.
Positioning, splinting a wound during coughing, heat or cold, a quiet environment, explanation, and simple presence all reduce pain measurably. They are often presented as things to try when medication is unavailable, which understates them: they work alongside medication and they address the fear component of pain, which analgesia does not touch. A frightened patient in pain is in more pain, and reassurance is a clinical intervention rather than a kindness.
Giving analgesia is half the task; the other half is evaluating whether it worked and whether it caused harm. For opioids specifically, the observation that matters is sedation level rather than respiratory rate alone, because increasing sedation precedes respiratory depression. A patient becoming progressively harder to rouse after opioid doses is showing the warning sign, and waiting for the respiratory rate to fall means waiting until the point of danger has already been reached.
Hand hygiene remains the single most effective intervention against healthcare-associated infection, and compliance remains imperfect everywhere, which is an uncomfortable pair of facts. The widely taught structure is the five moments: before touching a patient, before a clean or aseptic procedure, after body fluid exposure risk, after touching a patient, and after touching the patient's surroundings. The fifth is the one most often skipped, because touching a bed rail does not feel like touching a patient — and the organisms that cause ward outbreaks live on exactly those surfaces.
Aseptic technique protects a site that should have no organisms: a wound, a cannula insertion, a catheter, a sterile field. The principle to carry is the identification of key parts — the surfaces that will touch the patient's sterile site — and protecting those specifically, rather than attempting an impossible general sterility in a ward side room. A nurse who can name which parts of their procedure are key parts is performing aseptic technique; one following the steps without that understanding is performing a ritual that breaks the first time the situation is slightly different.
Every invasive device is a route in. Urinary catheters cause a large share of healthcare-associated infection, and the most effective intervention is not better catheter care but fewer catheter days — asking daily whether it is still needed. The same logic applies to cannulas and central lines: insert for a reason, care for the site, inspect it, and remove it when the reason ends. A device left in because nobody reviewed it is a risk being run for no benefit at all.
Source isolation protects other patients, and it has a genuine cost to the isolated one: less frequent contact, more missed observations, measurable loneliness and worse outcomes on several measures. Both things are true at once. The professional response is not to soften the precautions but to compensate deliberately for their effects — going in for reasons other than tasks, and not letting the gown and gloves become a reason to visit less often.
More clinical information is lost at handover than at any other point in a hospital stay, and most serious incidents involve a handover failure somewhere in their history. The causes are structural: handovers happen at shift ends when everyone is tired, in noisy places, under time pressure, and often without the written record present. A handover that consists of one nurse reciting what they remember to another nurse who is already thinking about their own workload transfers far less than either believes.
The widely taught structure is SBAR — situation, background, assessment, recommendation — and its value is that it forces the speaker to reach a recommendation. Without a structure, a worried nurse describes a patient at length and leaves the listener to work out what is being asked, which is precisely the point at which escalation fails. Saying plainly what you think should happen, even if you are unsure, gives the other person something to agree or disagree with. 'I think this patient needs to be seen now' is a complete communication; a list of observations is not.
Every nurse eventually escalates a concern and is not acted on, and every health system has a graded assertiveness route for exactly that situation, even where it is not named as such. The sequence is to state the concern, state it again with the reason, state the consequence you are worried about, and then go up the chain rather than letting it drop. This is the hardest skill in the subject because it requires contradicting someone more senior while uncertain, and it is the skill most often absent from the accounts of patients who died on a ward where somebody had already been worried about them.
The practical marks of a handover that works: it happens somewhere the information is available, it covers what has changed rather than everything, it names the patients the outgoing nurse is worried about and says why, it identifies what is outstanding and who is responsible, and it allows questions. A handover nobody can interrupt is a performance. A student who learns to ask one clarifying question at handover has already improved the shift they are about to work.
Malnutrition in hospital is common, under-recognised and directly worsens outcomes: wounds heal more slowly, infections are more likely, muscle is lost, and length of stay increases. It is rarely caused by an absence of food. It is caused by meals arriving while a patient is off the ward, trays placed out of reach, packaging nobody can open one-handed, fasting orders that outlive their reason, and nobody recording that three meals in a row went back untouched. Each of those is a nursing observation rather than a dietetic one.
Most systems use a screening tool on admission combining body mass index, recent unintentional weight loss and the effect of acute illness on intake. The tool is not the point; acting on the score is. A screening completed on admission, filed, and never repeated tells nobody that the patient has eaten almost nothing for five days. Re-screening at intervals and whenever the clinical picture changes is what makes the tool useful rather than administrative.
A patient who cannot swallow safely will aspirate, and aspiration pneumonia is a serious and sometimes fatal complication. Warning signs are coughing or a wet, gurgling voice after swallowing, food pocketing in the cheek, drooling, prolonged mealtimes and recurrent chest infections. After a stroke, swallow screening before the first oral intake — including before oral medication — is standard practice in most systems. Giving a tablet with a sip of water to a patient who has not been screened is a common and consequential shortcut.
Assisting someone to eat is frequently the lowest-status task on a ward and one of the most clinically significant. A patient who is helped to sit upright, given time, and not rushed will eat substantially more than the same patient with a tray left in front of them. Protected mealtimes exist for this reason, and they work only if the whole team honours them — the same cultural condition that determines whether a do-not-disturb medication round is real.
A patient loses muscle strength measurably within days of immobility, alongside bone density, cardiovascular conditioning, and gut and bladder function. An older patient admitted able to walk who spends a week in bed may leave unable to manage at home despite the original illness being cured. This is deconditioning, it is caused by care rather than disease, and early mobilisation is the treatment. Sitting a patient out of bed is a clinical intervention, not a comfort measure.
Falls in hospital cause fractures, head injuries, loss of confidence and prolonged stays. Risk factors cluster: previous falls, unsteady gait, confusion, urinary urgency, poor vision, unfamiliar surroundings and certain medicines, particularly sedatives and those lowering blood pressure. The practical interventions are unglamorous — call bell within reach, footwear that fits, a clear path to the toilet, adequate lighting, addressing urgency rather than waiting for the patient to try alone. Most inpatient falls happen on the way to the toilet.
Bed rails, chairs a patient cannot rise from and sedation are sometimes used with the intention of preventing falls, and the evidence does not support them: they increase agitation, cause different injuries, and are a restriction of liberty with its own legal framework. A confused patient trying repeatedly to get out of bed is usually trying to meet a need — the toilet, pain, thirst, fear — and meeting the need is both more effective and more lawful than preventing the movement.
Pressure damage and immobility are the same problem seen from different directions. A patient who cannot reposition themselves is accumulating pressure over bony areas, and the intervention is repositioning on a schedule, appropriate surfaces, inspecting the skin at every opportunity, and managing moisture. Heels and the sacrum are where most damage occurs, and heels are the site most often missed because they are under a blanket at the far end of the bed.
Discharge is a process rather than an event, and one that begins on the day of admission by asking what has to be true for this person to go home safely. Leaving it to the final morning produces the familiar failures: a patient waiting all day for medication, going home to a house with no food and stairs they cannot climb, with a follow-up nobody arranged and a wound nobody will dress.
The discharge medication list is where errors concentrate, because it is compiled in a hurry from a chart that has changed repeatedly. Reconciliation means comparing what the patient took before admission against what they are leaving with, and accounting for every difference: stopped, started, changed, or unintentionally lost. The commonest serious error is a medicine stopped temporarily during the admission and never restarted, which nobody notices until the condition it controlled returns.
Information given at discharge is given at the worst possible moment — the patient is tired, anxious to leave, and will retain very little. The technique with the best evidence is teach-back: rather than asking whether they understand, ask them to explain in their own words what they will do. This reliably exposes a misunderstanding that a nod would have hidden, and it takes less time than the readmission it prevents.
A patient readmitted within days has usually been failed by the discharge rather than by their illness: the teaching did not land, the medicines were wrong, the follow-up did not happen, or nobody asked whether anyone was at home. Reading readmission as a nursing outcome rather than an inevitability is what turns it into something addressable — and most of what makes it addressable happens on the ward before the patient leaves.
Transfusion is the clearest case in nursing of an error that is both catastrophic and entirely preventable by checking. Giving a unit intended for another patient can cause an acute haemolytic reaction, in which the recipient's immune system destroys the transfused cells, and it can kill within minutes. The error is almost never a failure of knowledge — it is a failure of identification, usually at the bedside, usually under time pressure, and usually by someone who was certain they had the right patient.
Everything before the bedside — the sample, the laboratory cross-match, the label — can be correct and the patient can still receive the wrong unit, because all of it depends on the sample having come from the patient whose name is on it. The bedside check is the only point where the physical patient and the physical unit are in the same place at the same time. It is performed against the patient's wristband and their stated identifiers, not against the notes, not against the bed number, and not against what the previous nurse said. Where local policy requires two people, both must check independently rather than one reading aloud while the other agrees.
Most serious reactions begin early, which is why observations are taken before starting, shortly after starting, and at intervals through the unit. The first minutes matter most and are the period during which the patient should not be left alone. The signs to act on are fever, rigors, flushing, back or loin pain, chest tightness, breathlessness, a fall in blood pressure, dark urine, or simply a patient who says they feel suddenly and badly wrong. That last one is a real clinical sign in transfusion and appears in incident reports far more often than it appears in textbooks.
Stop the transfusion first. Keep the line open with a separate infusion so there is access, assess the patient, escalate immediately, and do not discard the unit or the giving set — both are needed to work out what happened. The instinct to disconnect everything and tidy up is strong and destroys the evidence that would identify the cause. The sequence to hold is: stop, maintain access, assess, escalate, keep everything.
Transfusion is one of the few ward interventions that requires the patient to have been told what it is, why it is proposed, and what the alternatives are — and some patients decline blood on grounds that must be respected and recorded in advance rather than discovered during an emergency. The documentation carries an unusual weight too: the unit number, the start and finish times, the observations and who performed the check must all be recorded, because blood is traceable from donor to recipient by law in most jurisdictions. A unit given and incompletely documented breaks a chain that exists so that a problem found in one donation can be traced to everyone who received it.