Delirium in older adults: quick reference

Delirium in older adults: quick reference

A quick reference for ward and on-call physicians: recognise hypoactive delirium, stabilise the patient, work systematically through the possible causes, apply several non-pharmacological measures at the same time, and use drugs only in severe agitation or marked distress.

Delirium is an acute neuropsychiatric syndrome and a clinical diagnosis, often described as acute brain failure. It is precipitated by somatic illness, drugs, poisoning, withdrawal, surgery or several concurrent factors. A diagnosis of delirium is therefore the beginning of the investigation, not the end. In a frail older patient an apparently small insult — sleep deprivation, urinary retention or a new sedative drug, for example — may be enough to precipitate the condition [1,2].

Delirium is common but often missed. On medical and geriatric wards a substantial proportion of older patients are affected, and the incidence is particularly high after hip fracture, major surgery and intensive care [1,2]. The condition is associated with falls, aspiration, pressure sores, loss of function, longer length of stay, institutional care, readmission, cognitive decline and increased mortality [3,4].

Do this first

1. Assess vital signs and level of consciousness immediately.
2. Check oxygen saturation and capillary glucose.
3. Exclude immediately treatable threats without delay: hypoxia, hypoglycaemia, circulatory failure, sepsis, stroke, intracranial haemorrhage, seizure, poisoning and severe withdrawal.
4. Establish whether the condition is new or clearly worse than baseline. Contact relatives, home care services, the residential care home or the previous care provider.
5. Confirm delirium with a structured instrument, for example the 4AT or the CAM according to local protocol.
6. Look for several concurrent causes. In older patients delirium is often multifactorial.
7. Start non-pharmacological treatment immediately, in parallel with the investigation.
8. Document the diagnosis and communicate it at every handover.

Recognise it

Core criteria

Inattention is central. A simple check is to ask the patient to recite the months of the year backwards. A normal result does not exclude delirium, however, and an abnormal result is not specific to delirium.

Symptoms are often worse in the evening and at night, but diurnal variation is not a diagnostic requirement. Hallucinations, usually visual, disturbed sleep rhythm, anxiety, fear and delusions are common but not obligatory.

Psychomotor subtypes

It is wrong to assume that hyperactive delirium is always detected, or that hypoactive delirium always carries the worst prognosis in every population. Hypoactive delirium is, however, consistently underdiagnosed, and several studies have linked the subtype to high mortality, particularly in intensive care [5].

Suspect hypoactive delirium when an older patient

• suddenly becomes quieter or more passive
• stops eating, drinking or taking their medication
• needs more help than usual
• falls asleep during conversation
• answers slowly or not at all
• can no longer follow instructions
• develops new incontinence or a tendency to fall
• is described as "not being themselves"

Treat a new, unexplained fall in activity and attention as possible delirium until proven otherwise.

Establish the baseline

Do not simply ask whether the patient "is usually confused". Ask concrete questions:

• How was the patient functioning a week ago?
• Did the patient recognise relatives?
• Could the patient hold a coherent conversation?
• Could the patient manage to eat, use the toilet and take their medication?
• When was the patient last seen in their usual state?
• Have the symptoms varied over the course of the day?
• Is there a previous diagnosis of dementia or a previous delirium?

Dementia and delirium often coexist. Dementia is a strong predisposing factor, but acute deterioration in a person with dementia must not be dismissed as "the underlying disease". Delirium in people with dementia is common, underdiagnosed and associated with further loss of function and cognitive decline [6].

Use a structured instrument

4AT

The 4AT usually takes less than two minutes and assesses:

1. alertness
2. orientation
3. attention
4. acute change or fluctuation

In a meta-analysis of 17 studies the 4AT had a pooled sensitivity and specificity of 0.88 each [7]. The instrument is suitable for rapid initial assessment but does not replace history-taking, clinical examination or investigation of the cause. Specificity is lower in patients with dementia, so a positive assessment must always be interpreted together with information about an acute change from baseline.

CAM

The CAM is based on:

1. acute onset or a fluctuating course
2. inattention
3. disorganised thinking
4. altered level of consciousness

A positive CAM requires criteria 1 and 2 plus either criterion 3 or criterion 4. The CAM works best when the assessor has first carried out a structured interview and cognitive examination. Completing the CAM from general observation alone gives poorer sensitivity [1].

Clinical intuition can identify obvious cases but is insufficient to exclude delirium. In emergency care studies, physicians have missed a large proportion of cases. The 4AT has the broadest validation in the emergency department setting, while some CAM-based instruments have high specificity but fewer validation studies [8].

Repeat the assessment

A single negative test does not exclude a fluctuating delirium. Repeat the assessment whenever there is clinical suspicion, a change in behaviour or a change in the level of consciousness. Document the time and whether the patient had their glasses, hearing aid and appropriate language support.

Predisposition and precipitating factors

Think in two stages:

1. How vulnerable was the brain before the current event?
2. What acute insults have been added?

The greater the predisposition, the smaller the acute insult needed to precipitate delirium [1,2].

Common predisposing factors

• dementia or mild cognitive impairment
• previous delirium
• advanced age
• frailty
• reduced ability to perform activities of daily living
• visual or hearing impairment
• undernutrition
• multimorbidity
• renal or hepatic failure
• previous stroke or other neurological disease
• depression
• alcohol dependence
• polypharmacy

Frailty is not merely a marker of advanced age. In a meta-analysis, frail older patients had approximately double the risk of postoperative delirium compared with non-frail patients [9].

The list of causes: work through it systematically

Delirium rarely has a single cause. Find and treat every plausible contributing factor.

The medication review is mandatory

Go through the prescription chart, as-required medication, recently discontinued drugs, herbal remedies and over-the-counter preparations. Compare this with what the patient has actually been taking at home.

Particularly important drug classes are:

• anticholinergic drugs, for example certain bladder drugs, tricyclic antidepressants and sedating antihistamines
• benzodiazepines and other hypnotics
• opioids
• antiepileptics
• corticosteroids
• dopaminergic drugs
• antipsychotics
• lithium
• digoxin
• drugs causing hypoglycaemia
• drugs causing hyponatraemia
• drugs that accumulate in renal impairment
• combinations of several sedating preparations

Antipsychotics can themselves contribute to sedation, urinary retention, orthostatic hypotension and other complications. Do not reflexively add a drug before you have considered what can be stopped or reduced. Postoperative guidelines specifically recommend avoiding benzodiazepines, anticholinergic drugs and other strongly sedating preparations in at-risk patients [10].

Do not stop chronic benzodiazepine treatment abruptly. Withdrawal can precipitate an abstinence syndrome and worsen the delirium.

Infection and urine testing

Urinary tract infection is associated with delirium in older people, but asymptomatic bacteriuria is very common and has not been shown to be a sufficient explanation for delirium [11,12]. A positive urine dipstick, pyuria or bacterial growth therefore does not automatically mean a symptomatic urinary tract infection.

Assess:

• new urinary tract symptoms
• fever or hypothermia
• circulatory compromise
• flank or suprapubic pain
• the inflammatory picture
• any other likely source of infection

Avoid ending the search for a cause when the urine dipstick is positive. This is a common reason both for overdiagnosis of urinary tract infection and for missing pneumonia, an adverse drug effect, urinary retention or other illness.

Pain

Untreated pain is an important and often underestimated precipitant. In a meta-analysis, pain at rest was associated with approximately double the risk of delirium and severe pain with more than triple the risk [13]. Assess pain even in hypoactive delirium and in dementia. Use an observational pain assessment tool if the patient cannot self-report.

Avoid both undertreatment and unnecessarily high opioid doses. Monitor the depth of sedation, respiration, bowel function and urinary retention. Regional analgesia may be relevant for certain fractures and postoperative states.

Urinary retention and constipation

Examine the abdomen and ask actively about the last void and the last bowel movement. Have a low threshold for bladder scanning in:

• unexplained agitation or hypoactivity
• suprapubic pain
• small frequent voids
• overflow incontinence
• known prostatic enlargement
• drugs with anticholinergic effects
• opioid treatment
• recent removal of a urinary catheter
• the postoperative state

Bladder scanning need not be performed routinely in every patient regardless of the clinical picture, but urinary retention should be actively considered in everyone with delirium. Avoid an indwelling catheter if intermittent measures are sufficient.

Baseline investigation

The investigation should be guided by the history, examination, vital signs and clinical probability. An extensive standard panel without a clinical question burdens the patient and produces incidental findings that are difficult to interpret.

History

• the exact time course and the last time the patient was certainly normal
• current symptoms from the respiratory tract, urinary tract, abdomen, heart and nervous system
• pain, falls or trauma
• food and fluid intake
• the last void and bowel movement
• alcohol and other substances
• a complete drug history
• previous cognition and ability to perform activities of daily living
• previous delirium
• recent surgery, anaesthesia or change of care setting

Examination

• ABCDE
• temperature, blood pressure, heart rate, respiratory rate and oxygen saturation
• level of consciousness and attention
• a complete neurological examination including focal signs
• cardiac and respiratory examination
• abdominal examination and assessment of bladder filling
• skin, wounds, pressure sores and signs of trauma
• the mouth and signs of dehydration
• pain assessment
• signs of withdrawal or intoxication

Common initial tests and investigations

The following are often reasonable, but should be adapted to the situation:

• full blood count
• CRP
• sodium, potassium and calcium
• creatinine and estimated renal function
• glucose
• liver function tests
• ECG
• bladder scan where suspected, or with a low threshold as above
• blood gas in hypoxia, suspected hypercapnia, acid-base disturbance or systemic illness

Consider according to the clinical picture:

• troponin
• blood cultures
• chest radiograph
• urine dipstick and urine culture
• TSH and free T4
• vitamin B12 and folate
• drug levels, for example lithium or digoxin
• toxicology screen
• creatine kinase
• ammonia
• assessment for thiamine deficiency
• EEG
• lumbar puncture
• CT head

TSH, vitamin B12, urine dipstick and urine culture need not be taken routinely in everyone. Thyroid disturbance and vitamin deficiency are possible causes or contributing factors, but are rarely the sole explanation for a dramatically fluctuating condition that has developed over a few hours.

When should a CT head be performed?

A CT head should not be performed routinely in uncomplicated delirium with a clear extracerebral cause. Consider urgent imaging in:

• new focal neurological signs
• head trauma or a suspected fall
• anticoagulant treatment together with trauma or neurological suspicion
• new severe headache
• a reduced level of consciousness without explanation
• suspected stroke, intracranial haemorrhage or subdural haematoma
• persistent unexplained delirium despite initial investigation

In unselected groups the diagnostic yield of cerebral imaging is low. The transfer itself, the immobilisation and any sedation may also worsen the condition [2].

When should an EEG or lumbar puncture be considered?

Consider an EEG in:

• suspected non-convulsive status epilepticus
• episodic stereotyped symptoms
• unexplained fluctuation with subtle motor phenomena
• persistent impairment of consciousness after a seizure

Consider a lumbar puncture where meningitis, encephalitis or other neuroinflammation is suspected, particularly with fever, neck stiffness, seizures, new focal neurological signs or unexplained persistent impairment of consciousness.

Treatment: address the cause and provide support at the same time

Identifying and treating the precipitating cause is central to treatment, but it is rarely enough to treat a single factor. The patient simultaneously needs protection against complications and support for orientation, sleep, nutrition, mobility and communication.

A practical package of measures

Multicomponent programmes are more effective than isolated single measures. In a Cochrane review such interventions reduced the incidence of delirium from 18.4 to 10.5 per cent, corresponding to a relative risk of 0.57 [14]. The evidence is stronger for prevention than for treatment of established delirium, but the same measures are reasonable for limiting deterioration and complications.

The Hospital Elder Life Program (HELP) combines, among other things, orientation, sleep support, mobilisation, visual and hearing support, nutrition and hydration. In a meta-analysis the programme was associated with a lower incidence of delirium and fewer falls [15]. Overviews of multiple systematic reviews confirm that multicomponent interventions reduce the risk, while the effect of isolated single interventions and the effect on the duration of established delirium are more uncertain [16,17].

Sleep

Try first:

• daylight and physical activity during the day
• limiting long daytime naps
• reducing noise and light at night
• coordinating observations and blood tests
• toileting and analgesia before the night
• glasses and hearing aid during waking hours
• avoiding caffeine late in the day

Do not routinely start a hypnotic in order to "turn the day around". The drug may produce sedation without treating the delirium and may increase the risk of falls.

Physical restraint

Avoid belts, lap trays and other restraints where possible. They can increase fear, agitation, injury and the duration of delirium [1,2]. Use instead:

• close supervision
• relatives
• a sitter
• a low bed
• assisted mobilisation
• removal of unnecessary equipment
• treatment of pain, retention and other discomfort

If an acute restraint is judged absolutely necessary it must be proportionate, time-limited, prescribed and reviewed in accordance with the applicable regulations and local protocol.

Drugs: the last resort

No drug has been shown to cure delirium or reliably shorten its course. Antipsychotics should not be used routinely, should not be used as hypnotics, and should not be given merely because the patient is disorientated.

A Cochrane review found no evidence that antipsychotics reduce the severity of delirium, hasten resolution of symptoms or reduce mortality in patients outside intensive care [18]. A later systematic review likewise found no improvement in the duration of delirium, length of stay or mortality compared with placebo [19].

When drugs may be considered

Consider short-term symptomatic relief only when:

• despite de-escalation, the patient poses an immediate risk of serious harm to themselves or others
• life-saving treatment cannot otherwise be delivered
• pronounced hallucinations or delusions are causing severe distress
• the agitation is so severe that investigation of the cause or necessary treatment is otherwise impossible

The aim is calm, safe cooperation, not deep sedation.

Before an antipsychotic is given

1. Check glucose, oxygen saturation and vital signs.
2. Assess pain, urinary retention, constipation, thirst and withdrawal.
3. Attempt verbal de-escalation, environmental adaptation and support from relatives.
4. Check the current medication list.
5. Assess the QTc, electrolytes, arrhythmia risk and interactions.
6. Ask about Parkinson's disease and dementia with Lewy bodies.
7. Define which target symptom is to be treated.
8. Prescribe a review and a stop date.

Choice of drug

Clozapine requires specific monitoring and is not a routine drug for acute out-of-hours treatment. Quetiapine can also cause sedation, orthostatic hypotension and QT prolongation.

Risks of haloperidol and other antipsychotics

• QT prolongation and arrhythmia
• extrapyramidal symptoms
• akathisia
• dysphagia and aspiration
• sedation
• orthostatic hypotension and falls
• urinary retention
• neuroleptic malignant syndrome
• an increased risk of cerebrovascular events and mortality in older people with dementia

Where haloperidol is used, the QTc, electrolytes, interactions and other risk factors must be assessed. Intravenous administration carries particular risks and must follow local protocol. If the drug does not clearly improve the defined target symptom it should be stopped.

Benzodiazepines

Benzodiazepines should be avoided in ordinary delirium because they can cause:

• deeper sedation
• paradoxical agitation
• falls
• respiratory depression
• worse attention
• prolonged or worsened delirium

The Cochrane review provides no support for routine benzodiazepine treatment of delirium outside intensive care and outside a withdrawal indication [20]. The exceptions are principally alcohol withdrawal, benzodiazepine withdrawal and certain specific palliative situations.

Do not do this

• Do not give an antipsychotic merely because the patient is disorientated.
• Do not treat hypoactive delirium with stimulant or antipsychotic drugs.
• Do not start a benzodiazepine for sleep or anxiety without a specific indication.
• Do not interpret a positive urine dipstick as a definite cause.
• Do not perform a routine CT head without a clinical question.
• Do not insert a permanent urinary catheter simply to make nursing easier.
• Do not restrain the patient in bed if safer alternatives exist.
• Do not wake the patient repeatedly at night for observations that could be coordinated.
• Do not move the patient between rooms and wards without good reason.
• Do not discharge the patient on a temporary antipsychotic without an active decision.

Red flags

The following require immediate targeted assessment and often urgent action:

• an oxygen saturation that is low or falling
• hypoglycaemia
• hypotension or other circulatory failure
• fever, hypothermia or suspected sepsis
• new focal neurological signs
• sudden severe headache
• a seizure or suspected non-convulsive status
• head trauma, particularly on anticoagulant treatment
• marked neck stiffness
• a rapidly falling level of consciousness
• suspected poisoning
• alcohol or benzodiazepine withdrawal
• chest pain, a new arrhythmia or suspected myocardial infarction
• severe abdominal pain, ileus or urinary retention
• delirium with no identifiable cause that persists despite initial investigation

Older people may present with serious illness atypically. Myocardial infarction, sepsis, pneumonia and stroke can all present with delirium without classic local symptoms.

Common pitfalls

Missing hypoactive delirium

A quiet patient creates less work but may be more seriously ill. New passivity, refusal of food or somnolence must be actively investigated.

Calling delirium dementia

Dementia usually develops over months or years. Delirium develops over hours or days. Dementia with Lewy bodies can fluctuate, but a clear acute deterioration should nonetheless be managed as possible delirium.

Calling delirium depression

Hypoactivity, reduced speech and poor food intake can resemble depression. Acute onset, fluctuation, inattention and an altered level of consciousness point to delirium.

Investigating only one cause

A pneumonia does not exclude concurrent dehydration, an opioid adverse effect, pain, constipation or retention. Continue the search for causes even once a possible precipitant has been found.

Making the medication list longer

Look first for drugs that can be stopped, paused or reduced in dose. Adjust doses to renal and hepatic function.

Over-sedating

The fact that agitation has stopped does not mean the delirium has been treated. Deep sedation can convert a visible hyperactive state into a hard-to-detect hypoactive one and increase the risk of aspiration and immobilisation.

Forgetting the stop date

Antipsychotics started during an acute delirium often follow the patient to the next care setting even though the indication has ceased. Postoperative guidelines describe this as an important iatrogenic harm [10].

Documentation and handover

Document:

• that the patient has delirium or suspected delirium
• the 4AT, CAM or equivalent result
• the previous cognitive and functional baseline
• the likely causes and which have been investigated
• the non-pharmacological measures taken
• which drugs have been stopped or changed
• for any antipsychotic drug, the target symptom, the effect and adverse effects
• the planned review and stop date
• the need for follow-up after discharge

Do not write only "agitated", "confused" or "demented". The diagnosis of delirium must be clearly stated if the measures are to continue throughout the care pathway.

Course and follow-up

Delirium is not always short-lived. Symptoms may persist at discharge and recovery may take weeks or months [1]. Inform relatives that:

• fluctuations may continue after the cause has been treated
• the patient may need more help than before the illness
• sleep, attention and memory may recover slowly
• a fresh deterioration may indicate relapse or new physical illness

Delirium is associated with long-term cognitive decline. A meta-analysis of 23 studies found poorer cognitive function at least three months after the episode in patients who had had delirium [21]. A larger later review also found an increased risk of dementia, disability, institutional care, readmission and death after discharge [4]. These associations are partly explained by the fact that patients with delirium are sicker and frailer from the outset, but they are strong enough to justify structured follow-up.

Plan before discharge

• Compare cognition and function with the previous baseline.
• Ensure medication reconciliation.
• Stop temporary antipsychotics where possible.
• If an antipsychotic must be continued, document the precise indication, the planned duration and who is responsible for review.
• Assess falls risk, nutrition, mobility and the need for aids.
• Inform primary care, municipal care services and relatives.
• Plan follow-up of persisting symptoms.

A formal diagnosis of dementia should generally not be made during an ongoing delirium. Cognitive assessment should be carried out after the acute condition has stabilised, often after a few weeks, and repeated if uncertainty remains. The timing should be adapted to recovery, previous function and clinical need.

Special situations

Postoperative delirium

Postoperative delirium must not be dismissed as a normal consequence of anaesthesia. Look for:

• haemorrhage and anaemia
• hypoxia
• infection
• pain
• opioid adverse effects
• urinary retention
• constipation or ileus
• fluid and electrolyte disturbance
• withdrawal
• a surgical complication

Optimise analgesia, mobilise early, and remove catheters and lines as soon as they are no longer needed. Multicomponent interventions are recommended for at-risk patients [10].

Intensive care

In the intensive care unit the CAM-ICU or another validated critical care instrument is usually used. Assess the depth of sedation, analgesia, mechanical ventilation, sleep, immobilisation and exposure to sedating drugs. Haloperidol has not been shown to prevent delirium in intensive care or to improve mortality [22]. Intensive care management, including the sedation strategy and any use of dexmedetomidine, follows specific local protocols and cannot be transferred directly to a general ward.

Parkinson's disease and dementia with Lewy bodies

Patients with dementia with Lewy bodies may have marked sensitivity to antipsychotics, with severe rigidity, reduced consciousness and autonomic involvement. Avoid haloperidol. Treat precipitating causes and environmental factors, and discuss any pharmacological treatment with a geriatrician, neurologist or psychiatrist.

Alcohol withdrawal

Alcohol withdrawal can cause a hyperactive delirium with tremor, sweating, tachycardia, hypertension, hallucinations and seizures. This is a distinct diagnosis in which benzodiazepines are first-line treatment according to the local withdrawal protocol. Assess at the same time the need for thiamine and for treatment of electrolyte disturbances.

Palliative care

At the end of life the aim may be relief rather than complete reversal. Even so, assess treatable causes such as urinary retention, constipation, pain and adverse drug effects. The choice of drug and the acceptable depth of sedation should follow the goals of care and the palliative care plan.

Summary for the on-call physician

Sources

1. Marcantonio ER. Delirium in Hospitalized Older Adults. New England Journal of Medicine 2017. PMID: 29020579
2. Inouye SK, Westendorp RG, Saczynski JS. Delirium in elderly people. Lancet 2014. PMID: 23992774
3. Iglseder B, Frühwald T, Jagsch C. Delirium in geriatric patients. Wiener Medizinische Wochenschrift 2022. PMID: 35006521
4. Tesfaye Y et al. Long-term clinical outcomes of delirium after hospital discharge: a systematic review and meta-analysis. Age and Ageing 2025. PMID: 40627690
5. Krewulak KD et al. Risk factors and outcomes among delirium subtypes in adult ICUs: A systematic review. Journal of Critical Care 2020. PMID: 31986369
6. Fong TG, Inouye SK. The inter-relationship between delirium and dementia: the importance of delirium prevention. Nature Reviews Neurology 2022. PMID: 36028563
7. Tieges Z et al. Diagnostic accuracy of the 4AT for delirium detection in older adults: systematic review and meta-analysis. Age and Ageing 2021. PMID: 33951145
8. Carpenter CR et al. Delirium detection in the emergency department: A diagnostic accuracy meta-analysis of history, physical examination, laboratory tests, and screening instruments. Academic Emergency Medicine 2024. PMID: 38757369
9. Gracie TJ et al. The Association of Preoperative Frailty and Postoperative Delirium: A Meta-analysis. Anesthesia and Analgesia 2021. PMID: 34257192
10. American Geriatrics Society Expert Panel on Postoperative Delirium in Older Adults. American Geriatrics Society abstracted clinical practice guideline for postoperative delirium in older adults. Journal of the American Geriatrics Society 2015. PMID: 25495432
11. Shenvi C et al. Managing Delirium and Agitation in the Older Emergency Department Patient: The ADEPT Tool. Annals of Emergency Medicine 2020. PMID: 31563402
12. Krinitski D et al. Associations of delirium with urinary tract infections and asymptomatic bacteriuria in adults aged 65 and older: A systematic review and meta-analysis. Journal of the American Geriatrics Society 2021. PMID: 34448496
13. White N et al. Understanding the association between pain and delirium in older hospital inpatients: systematic review and meta-analysis. Age and Ageing 2024. PMID: 38610062
14. Burton JK et al. Non-pharmacological interventions for preventing delirium in hospitalised non-ICU patients. Cochrane Database of Systematic Reviews 2021. PMID: 34280303
15. Hshieh TT et al. Hospital Elder Life Program: Systematic Review and Meta-analysis of Effectiveness. American Journal of Geriatric Psychiatry 2018. PMID: 30076080
16. Zhao Q et al. Non-pharmacological interventions to prevent and treat delirium in older people: An overview of systematic reviews. International Journal of Nursing Studies 2023. PMID: 37826889
17. Kim YH, Kim NY, Ryu S. Effects of non-pharmacological interventions for preventing delirium in general ward inpatients: A systematic review and meta-analysis of randomized controlled trials. PLoS One 2022. PMID: 35522654
18. Burry L et al. Antipsychotics for treatment of delirium in hospitalised non-ICU patients. Cochrane Database of Systematic Reviews 2018. PMID: 29920656
19. Nikooie R et al. Antipsychotics for Treating Delirium in Hospitalized Adults: A Systematic Review. Annals of Internal Medicine 2019. PMID: 31476770
20. Li Y et al. Benzodiazepines for treatment of patients with delirium excluding those who are cared for in an intensive care unit. Cochrane Database of Systematic Reviews 2020. PMID: 32108325
21. Goldberg TE et al. Association of Delirium With Long-term Cognitive Decline: A Meta-analysis. JAMA Neurology 2020. PMID: 32658246
22. Herling SF et al. Interventions for preventing intensive care unit delirium in adults. Cochrane Database of Systematic Reviews 2018. PMID: 30484283