An older adult who is confused, withdrawn, or not themselves is one of the most common presentations in hospital and one of the most frequently misread. The three usual explanations look alike from the end of the bed, and they get conflated constantly — usually in the direction of assuming dementia, because the patient is old and the confusion is right there in front of you.

That assumption is the expensive one. Delirium is an acute medical emergency with an underlying cause, and it’s the only one of the three where finding the answer today changes the outcome.

The two questions that do most of the work

You can get a long way with two:

How fast did this happen? Delirium arrives over hours to days. Dementia takes months to years. Depression sits in between — weeks.

Can they hold attention? Inattention is the hallmark of delirium. Ask them to say the months of the year backwards, or to spell a short word backwards, and watch whether they can hold the thread. In dementia, attention is relatively preserved until late; in depression, they can usually attend if they’re willing to engage.

Onset and attention. If both point the same way, you have your answer most of the time.

	Delirium	Dementia	Depression
Onset	Hours to days	Months to years	Weeks
Course	Fluctuates, often worse at night	Slowly progressive	Fairly steady, worse in the morning
Attention	Impaired — the hallmark	Preserved until late	Variable; intact if engaged
Consciousness	Altered	Clear until very late	Clear
Typical answer	Muddled, inconsistent	Confabulates, tries to cover	“I don’t know”
Reversible	Usually, if the cause is found	No	Usually

The most useful question is asked of somebody else

“What was he like two weeks ago?” A family member, a carer, or the nursing home notes will tell you more than any bedside test. Without a baseline you cannot distinguish acute change from chronic state — and the baseline is almost never in the chart in front of you.

The delirium you’ll miss

Hyperactive delirium — agitated, pulling at lines, trying to climb out — gets noticed immediately, because it’s disruptive.

Hypoactive delirium is quiet. The patient is withdrawn, sleepy, slow to answer, not bothering anyone. It’s more common than the agitated form, it carries a worse prognosis, and it is routinely written off as tiredness, low mood, or “just their dementia.” A previously chatty patient who has gone quiet is a red flag, not a relief.

The formal way to structure this is the Confusion Assessment Method: an acute onset with a fluctuating course, plus inattention, plus either disorganised thinking or an altered level of consciousness. Use whatever screening tool your facility has adopted, and use it on schedule rather than only when someone seems off.

When it is delirium, look for why

Delirium is a syndrome, not a diagnosis. It always has a cause, and the causes are mostly ordinary:

Infection — urinary and respiratory most commonly
Drugs — especially anticholinergics, benzodiazepines, opioids, and anything newly started or newly stopped
Metabolic — sodium, glucose, calcium, kidney or liver dysfunction, dehydration
Hypoxia and low perfusion in any form
Pain, and the under-recognised classics: urinary retention and constipation
Withdrawal — alcohol and benzodiazepines
Environment — sleep deprivation, no daylight, no glasses, no hearing aids, repeated ward moves

Glasses, hearing aids, sleep, and a window

A meaningful share of hospital delirium is manufactured by the hospital. Someone who can’t see, can’t hear, hasn’t slept properly in three nights and has no idea whether it’s morning is going to be confused. Getting their glasses and hearing aids back, protecting sleep, opening the curtains, and getting them mobile is genuine treatment — not comfort care around the edges of the real work.

What this means for your shift

Establish the baseline early and write it down where the next person will find it.
Screen on schedule, not on suspicion — the hypoactive ones don’t attract attention.
Treat new confusion as acute until proven otherwise. “Known dementia” in the chart does not explain a change that started yesterday. People with dementia get delirium more easily, not less.
Chart specifics, not labels. “Oriented to person only, could not name the month, was oriented ×3 yesterday” is usable. “Confused” is not.

The bottom line

Ask how fast it happened and whether they can hold attention. If it came on quickly and attention is gone, treat it as delirium and go looking for the cause — because that’s the one where looking today changes what happens next.

Keep going

For agitation that needs managing while the cause is being found, verbal de-escalation. For charting a changing mental state so the trend is visible, neuro checks: documenting them defensibly.

Sources

References

American Geriatrics Society — delirium and older-adult care guidance, including the Beers criteria
StatPearls — Delirium — causes, assessment, and management
Hospital Elder Life Program (HELP) — the non-pharmacological prevention bundle