Confusion that arrives in days is not dementia and not old age. It is usually a treatable illness wearing a frightening disguise — and the person most likely to spot it is the visitor who knew them last week.

What changed, and how fast

For each of these, say how today compares with the person's ordinary self — not with what an older person is supposed to be like. The tool reads the pattern: which parts changed, how quickly, and whether they come and go. Nothing is saved, nothing leaves the page, and this cannot diagnose anything. It exists to turn what you already know into a sentence a clinician can use.

What it is, and what defines it

• It is a change, not a state. The diagnosis rests on a decline from that person's own baseline over hours or days. Without a baseline there is nothing to compare against, which is exactly why a relative's account is not background colour — it is the diagnostic information.
• Inattention is the core feature. Not memory: attention. Losing the thread mid-sentence, needing questions repeated, drifting off while you are still talking. If attention is intact, delirium is unlikely; if it is newly impaired, it is the first thing to think of.
• It fluctuates, and that is not a contradiction. Someone can be lucid at eleven in the morning and lost by six in the evening. “He was fine when the doctor came” is not evidence against delirium; fluctuation is one of its defining features, and the ward round is a ten-minute sample.
• It is a sign of physical illness, not a psychiatric event. The brain is the organ that fails first and most visibly when something else is wrong. The confusion is a symptom in the same sense that a cough is: it points somewhere else.
• It is common, serious and frequently reversible. It lengthens hospital stays and carries a real risk of lasting decline, and at the same time a large share of it resolves once the cause is treated. Both halves of that sentence matter: it is neither trivial nor a verdict.
• It happens at home too. A urinary infection, a chest infection, dehydration in a heatwave or a new painkiller can produce it in somebody's own kitchen. It does not require a hospital, and it is not a hospital's private problem.

The quiet kind, which is most of it

• Hypoactive delirium is the commoner form and the one that gets missed. The person is drowsy, slowed, withdrawn, answering in single words, sleeping through meals. Nothing about that behaviour asks for attention, and on a busy ward it reads as a patient who is comfortable.
• “Settled” is the word to be suspicious of. If somebody who talked and joked last week is now quiet and easy to look after, that is a change, and a change is the whole point. The agitated form gets noticed because it is inconvenient; the quiet form is convenient, which is why it is dangerous.
• The outcomes are not better for being quiet. If anything, hypoactive delirium is associated with worse outcomes, largely because it is recognised later and its cause is treated later.
• Test attention, gently, rather than testing memory. Ask them to say the months of the year backwards from December, or the days of the week backwards. Faltering on that when they could have done it last month is a real finding, and it takes twenty seconds.
• Say the word. Staff respond to “I think this is delirium, he was completely himself on Sunday” in a way they do not respond to “he seems a bit confused”. Naming it puts it in the notes, and what is in the notes gets looked for.

Not dementia, and not simply old age

The short list of things to look for

1. Infection — urine and chest most often, and in an older person often without a fever and without them complaining of anything.
2. Drugs, new or newly stopped. Anything with a sedating or anticholinergic effect: some bladder tablets, older antihistamines, some antidepressants, opioids, sleeping tablets. Stopping alcohol or long-term sleeping tablets abruptly is its own cause, and it is dangerous.
3. Dehydration and constipation, and a bladder that is not emptying. Unglamorous, extremely common, and each of them fixable within hours. Urinary retention in particular can present as nothing but confusion.
4. Pain that nobody has asked about. Somebody who cannot explain that their hip hurts may show it only as agitation. Untreated pain causes delirium; so does over-sedation for it.
5. Metabolic things — low sodium, high or low blood sugar, kidney or liver trouble, low oxygen. All are blood tests, and all are treatable.
6. The brain itself, less often. A stroke, a bleed after a fall, a seizure. Sudden weakness or drooping on one side, or confusion after a bang on the head, changes the pathway entirely and needs emergency assessment.

What makes it worse, and what a visitor can actually do

• Glasses and hearing aids are treatment, not comfort items. Somebody who cannot see or hear the room they are in will misinterpret it. Getting the hearing aid out of the locker, with working batteries, does more than most single interventions available on the ward.
• Daylight, a clock and a calendar in view. Delirium is partly a failure of orientation, and orientation needs something to orient to. A dark bay with no window and no clock is an environment designed to sustain confusion.
• Sleep at night, and out of bed in the day. Night-time observations, moving somebody between bays after dark, and a whole day lying flat all worsen it. It is reasonable to ask whether the four-hourly night observations are still necessary.
• Familiar faces and familiar objects. Longer, calmer visits from people they know are among the few things that reliably help. Bring photographs, a dressing gown they recognise, their own hairbrush.
• Catheters, drips and cannulas are tethers. Each one is a thing to pull at, an infection risk and a reason to be restrained. Asking “does this still need to be in?” is a legitimate question, not an interference.
• Say who they are. Write a short baseline and give it to the nurse in charge: what they were like a week ago, what they normally manage, what they are called, what frightens them, what calms them. A page like that changes how a whole shift treats somebody.

Sedation, restraint, and the thing not to ask for

• Sedation does not treat delirium. It treats the disturbance the delirium causes. Antipsychotics are for a narrow situation — distress or danger that has not responded to everything else — at the lowest dose for the shortest time, and they carry real risks in older people including stroke and death.
• Sleeping tablets and benzodiazepines usually make it worse, except in alcohol or benzodiazepine withdrawal, where they are the treatment. That distinction matters and is worth stating out loud if somebody's drinking has stopped abruptly.
• Non-drug measures first, and they are not soft options. Reorientation, hearing and vision, hydration, pain relief, sleep, mobility, a familiar person: these are the evidence-based treatment, and the drug is what happens when they have failed.
• Physical restraint and bed rails increase harm. They raise the risk of injury and of prolonged delirium, and they escalate distress. If somebody is trying to get out of bed, the question is why — pain, a full bladder, fear, needing the lavatory — not how to stop them.
• Ask what the plan is when it is not working. “If he is no better in twenty-four hours, what changes?” is the question that turns a stalled situation into a reviewed one.

What to say, in six sentences

The first sentence does most of the work, because it supplies the one piece of information the notes do not contain. The fifth exists because a quiet patient has to be actively brought to attention.

What people are told, and what is so

The drill: 16 decisions

Sixteen ordinary moments — a father who is suddenly quiet, a mother who was lucid at the ward round, a new bladder tablet, a nurse who writes “pleasantly confused”. Most have an instinctive answer that sounds sensible and is wrong. Pick your move; every answer explains why.

The card

Print it, fill in the left-hand side while you still can, and hand it to the nurse in charge.

🖨️ Print the card