Delirium Prevention

Delirium is an acute disturbance in attention, awareness and cognition that develops over hours to days and fluctuates in severity through the day. It is acute brain dysfunction and a medical emergency: the confusion is a symptom demanding a cause, and the cause needs immediate investigation and treatment.

The DSM-5 criteria. Five things must be true, and the fifth is the one that turns a diagnosis into a search:

1. A disturbance in attention (reduced ability to direct, focus, sustain and shift it) and awareness (reduced orientation to the environment)
2. Acute onset over hours to days, representing a change from baseline
3. A course that fluctuates during the day
4. An additional cognitive disturbance: memory, orientation, language, visuospatial ability or perception
5. Evidence that it is the direct physiological consequence of a medical condition, substance intoxication or withdrawal, a medication side effect, or several causes together

It is not diagnosed when a pre-existing dementia or a psychiatric condition explains the picture better, nor when arousal is so reduced (coma) that attention cannot be assessed.

Who gets it. Incidence tracks the population, and hip fracture sits at the top:

• Hip fracture: 35-65%, the highest-risk orthopaedic population
• Elective hip or knee arthroplasty: 4-10%
• Spinal surgery: 10-15%
• General orthopaedic trauma: 15-25%
• ICU admission after trauma: 50-80%

Risk rises with age over 65 (each decade doubles it), pre-existing cognitive impairment or dementia, multiple comorbidities, emergency surgery, prolonged anaesthesia and postoperative complications.

Vulnerability and insult. Delirium risk is the product of predisposing factors, the patient's baseline vulnerability, and precipitating factors, the acute insult. A highly vulnerable patient needs only a minor precipitant; a robust one needs a major insult.

Predisposing factors. Some are modifiable, some are not:

• Non-modifiable: age over 65, pre-existing dementia or cognitive impairment, a history of delirium, severe illness or comorbidity burden, stroke or neurological disease
• Modifiable: sensory impairment (vision, hearing), malnutrition or dehydration, functional dependence, depression, alcohol use

Precipitating factors. These are the insults the admission adds, and several are ours to withhold:

• Surgical: hip fracture surgery, emergency surgery, anaesthesia lasting over 3 hours, significant blood loss
• Medical: infection (UTI, pneumonia), hypoxia, hypotension, metabolic disturbance, severe pain
• Iatrogenic: high-risk medications, physical restraints, a urinary catheter, sleep deprivation

Predicting risk. Each of these high-risk features increases risk 2-5 fold:

• Age over 70
• Cognitive impairment (MMSE under 24)
• Severe illness (ASA III-IV)
• Sensory impairment (vision or hearing)
• Depression
• Dehydration or malnutrition
• Alcohol use disorder

Hip fracture adds its own: an intertrochanteric rather than an intracapsular fracture, a delay to surgery over 48 hours, intraoperative hypotension, postoperative anaemia (Hb under 90 g/L) and inadequate analgesia.

View larger

Frailty. The single concept that best captures predisposing vulnerability is frailty: diminished physiological reserve across multiple systems, leaving the patient unable to withstand the stress of surgery. It operationalises the vulnerable-brain idea, since a frail patient becomes delirious after only a minor precipitant, and it is the strongest single predictor of postoperative delirium. It is measured, not estimated by eye:

• Clinical Frailty Scale (CFS, Rockwood): a judgement-based 9-point scale from 1 (very fit) through 4 (vulnerable), 5 (mildly frail) and 7 (severely frail) to 9 (terminally ill). Higher scores predict delirium, complications, loss of independence and mortality, guide treatment intensity and shared decision-making, and it is quick and the most widely used scale in fracture pathways
• Fried physical phenotype: frail if 3 or more of unintentional weight loss, exhaustion, weak grip strength, slow gait speed and low physical activity (1-2 criteria is pre-frail)
• FRAIL scale: a brief 5-item questionnaire screen - Fatigue, Resistance (stair climbing), Ambulation, Illnesses and Loss of weight

Frailty screening therefore flags exactly the patients who most need the HELP bundle and orthogeriatric co-management, and it predicts the same outcomes delirium does: mortality, loss of function and institutionalisation.

Assessment tools. The CAM is the diagnostic reference, the 4AT the ward screen, the DRS-R-98 the research instrument, and none of them replaces a collateral history.

Mechanisms. The pathophysiology remains incompletely understood, and several mechanisms are implicated at once. One is neurotransmitter imbalance: a cholinergic deficiency, from reduced acetylcholine availability or an increased anticholinergic drug burden, alongside dopaminergic excess in the mesolimbic pathways, with dysregulation of serotonin, GABA, glutamate and noradrenaline.

Neuroinflammation. Systemic inflammation from surgery or infection releases cytokines (IL-1, IL-6, TNF-alpha), which cross the blood-brain barrier, activate microglia and disrupt neurotransmission. In critical illness the barrier itself becomes more permeable and admits neurotoxic substances into the CNS.

Oxidative stress and neuroendocrine dysfunction. Reactive oxygen species and mitochondrial dysfunction injure neurons. The HPA axis is dysregulated, cortisol rises, and melatonin disruption upsets the sleep-wake cycle.

Subtypes. There are three, and the frequencies are the ones to remember: hypoactive is the commonest pure subtype, and mixed the commonest pattern overall.

Severity. Severity sets the intensity of management.

Duration. Acute delirium lasts under a week; most cases resolve quickly once the cause is treated, and the prognosis is good if it is managed. Persistent delirium lasts more than a week, is associated with dementia, resolves incompletely, carries worse long-term outcomes and may never fully clear.

The three Ds. Delirium, dementia and depression are the differential for the confused older patient, and they separate on onset, course and attention.

Delirium is acute and fluctuating, worse at night ("sundowning") and lasts days to weeks; dementia is chronic, stable from one day to the next and permanent. The two are not mutually exclusive.

The investigation is a search for a cause, targeted by the clinical assessment rather than ordered as a panel. The bedside tests come first because they identify reversible causes rapidly, and hypoxia, hypoglycaemia and retention are common.

At the bedside

• Vital signs: temperature, heart rate, blood pressure, respiratory rate (fever, hypotension, tachycardia)
• Oxygen saturation
• Finger-prick blood glucose
• Bladder scan for urinary retention
• Abdominal examination for constipation or ileus
• ECG for arrhythmia, ischaemia or a silent MI

Bloods. The standard panel is FBC (infection, anaemia), U&E (renal function, sodium, potassium), LFTs, calcium, CRP or ESR and a formal glucose. Add blood cultures if febrile, urinalysis and culture if there are symptoms or a catheter, thyroid function if indicated, B12 and folate for chronic confusion, and troponin for cardiac symptoms.

When to CT the head. Focal neurological signs, a history of a fall, anticoagulation, head trauma or an unexplained neurological change. It is not routine for every delirium: metabolic and infectious causes are far more common.

Extended workup. Reserved for the patient whose first round is unrevealing or whose history points somewhere specific.

The drug review. Read the whole chart against the BAD MEDS list, and add muscle relaxants, metoclopramide and the tricyclics to the suspects. Identify recent additions or dose changes, deprescribe where possible and use the Beers Criteria.

The target. Operate within 36 hours of admission ideally and within 48 hours at most; NICE, ANZHFR and the BOA all set that ceiling. Beyond the mortality and complication figures in Klestil's meta-analysis above, early surgery brings earlier mobilisation, less pain and a shorter stay. Be precise about delirium, though: its reduction is driven mainly by multicomponent and orthogeriatric care rather than by surgical timing alone, and the one large randomised trial of ultra-early surgery, HIP ATTACK, was negative for mortality.

Reasons for delay. A delay is only acceptable for reversible medical optimisation, and the list of causes is short:

• Medical optimisation (cardiac, respiratory)
• Anticoagulation reversal
• Theatre availability, which is a system failure
• Out-of-hours staffing

Anaesthesia. Four questions: which technique, how deep, what to control on the table, and how to manage pain afterwards.

Algorithm

View larger

The order of business. Once delirium is identified the approach has four steps, in this order: investigate and treat the underlying cause (the I WATCH DEATH workup above), then supportive care and safety, then non-pharmacological management, and only then, for severe agitation or a safety risk, a drug.

Non-pharmacological management is first-line for every delirium.

• Reassurance and reorientation: a calm manner, frequent reorientation to person, place and time, procedures explained simply, familiar objects and photos, and continuity of care with the same nurses where possible
• Environment: a well-lit room by day with natural light, quiet at night, unnecessary equipment removed, a clock and calendar visible
• Mobilisation: early, even if only sitting in a chair, with physiotherapy, no bed rest and minimal restraint
• Sensory aids: glasses and hearing aids in, background noise down
• Sleep: night-time interruptions minimised, care activities grouped, excessive daytime sleep avoided
• Family: presence reduces agitation, familiar voices help, and the family need information

When a drug is justified. Only for severe agitation endangering the patient or staff, a risk of harm (pulling lines, falling), distressing hallucinations or delusions, or after non-pharmacological measures have failed. Two principles govern the prescription: no medication has proven efficacy for delirium, they only manage symptoms; and there is no role for prophylactic antipsychotics, HOPE-ICU and MIND-USA having shown no benefit. Use the lowest effective dose for the shortest duration and reassess the need frequently.

Hypoactive delirium. Often no medication is needed at all: treat the cause, mobilise and stimulate, and engage the family.

Hyperactive delirium with safety concerns. Try redirection and one-to-one observation first; if a drug is needed, haloperidol 0.5 mg. Physical restraints are the absolute last resort, because they increase both delirium and injury.

Delirium in ICU. Incidence is higher (50-80%), and it is tied to mechanical ventilation, sedation and critical illness. The ABCDEF bundle (Awakening, Breathing, Coordination, Delirium monitoring, Early mobility, Family engagement), light sedation targets and daily sedation interruption are the response.

Delirium in dementia. Harder to diagnose, because the change from baseline is harder to see; the risk is higher, the patient more prone to medication side effects, and the family's account of baseline function is essential.

On the ward. Falls and injuries (a 2-3 fold increased risk during the admission), self-removal of lines, drains and catheters, aspiration pneumonia, pressure injuries from immobility, and injuries to staff from agitated patients. Medically, the stay lengthens by a mean of 2.5 days, postoperative complications are more frequent, and nosocomial infection, venous thromboembolism from immobility, malnutrition, dehydration and medication-related adverse events all rise.

Mortality. Delirium multiplies mortality 3-5 fold, and the effect persists at a year:

• In-hospital: 10-26% against 5% without delirium
• 30 days: 15-30%
• 6 months: 25-40%
• 1 year: 35-50%

Function. Recovery is delayed or incomplete: 60% have declined in ADL function at discharge and 40% still have functional impairment at 6 months; fewer return home, new nursing-home placement is 2-3 times more likely, and caregiver burden rises. The cost to the family is real: 75% report significant stress, and post-traumatic stress affects patients and families both. The healthcare cost is driven by the longer stay, the complications and institutionalisation.

Cognition. Delirium accelerates cognitive decline in those with dementia and doubles the risk of developing dementia in those without. 30-40% have persistent cognitive impairment at 6 months, and some never return to their baseline.

Delirium as brain injury. Recent evidence suggests delirium is not simply transient and fully reversible; it may cause permanent injury. The proposed mechanisms are neuroinflammation with neuronal apoptosis, blood-brain barrier disruption, amyloid deposition of the Alzheimer type, white matter change on MRI and hippocampal atrophy. Clinically, delirium may unmask a subclinical dementia or may directly cause one, and either way the case for prevention hardens.

Delirium versus postoperative cognitive dysfunction. Persistent cognitive impairment after surgery overlaps with, but is distinct from, postoperative cognitive dysfunction (POCD), now folded into the 2018 consensus nomenclature of perioperative neurocognitive disorders (PND) (Evered et al). The PND umbrella spans a pre-existing preoperative neurocognitive disorder; postoperative delirium, the acute, fluctuating, attention-disturbed bedside syndrome of days 0-7; delayed neurocognitive recovery, cognitive decline up to 30 days; and postoperative neurocognitive disorder (POCD), decline detected from about 30 days up to 12 months.

How POCD differs. POCD is subtle and detected only on formal neuropsychological testing of memory and executive function, with no disturbance of attention or consciousness and no acute fluctuation: the patient looks normal at the bedside. Postoperative delirium is itself a risk factor for subsequent POCD and long-term decline; the two lie on one spectrum of perioperative brain injury, which is why preventing delirium, and avoiding deliriogenic drugs, intraoperative hypotension and hypoxia, plausibly also reduces longer-term cognitive harm.

The first 24-48 hours. Reorient frequently to person, place and time; assess and control pain without oversedating; mobilise as soon as it is safe; return glasses and hearing aids at once; encourage oral intake. Avoid unnecessary catheterisation, excessive fluid restriction and prolonged fasting.

On the ward. Consistent nursing staff, family involvement, a promoted sleep-wake cycle, daily cognitive stimulation and monitoring of bowel and bladder function, in an environment with natural light by day, quiet at night, a visible clock and calendar and familiar objects from home.

Before discharge. Delirium should be resolved or resolving, the underlying cause treated, the cognitive baseline documented, the discharge environment safe and support services arranged. The GP letter carries a delirium summary, a medication reconciliation and a documented follow-up plan.

What the family are told. Delirium can take weeks to resolve fully, it may recur with future illness or surgery, new confusion should prompt a call, and cognitive rehabilitation strategies help. Give a written delirium fact sheet with emergency contact numbers and the follow-up appointments.

Cognitive follow-up matters most when symptoms persist beyond the expected recovery, because delirium raises the risk of later cognitive decline.

Prevention is far more effective than treatment. Across the trials, multicomponent intervention removes roughly a third of cases, 30-40%: HELP cut incidence from 15.0% to 9.9% (matched OR 0.60), Marcantonio's hip fracture trial from 50% to 32%, and the Cochrane pooled estimate is RR 0.69. Note what that does and does not mean: it is a third of cases prevented, not a third of severity reduced, and no trial has shown benefit once delirium is established.

HELP. The Hospital Elder Life Program is the gold-standard evidence-based intervention for delirium prevention. It targets the key risk factors with non-pharmacological measures delivered by trained volunteers and staff; the components are the mnemonic below.

The details that get asked. Mobilisation means out of bed within 24 hours, walking or a wheelchair three times daily and range-of-motion exercises; physical restraints double delirium risk, and urinary catheters come out as soon as possible. Orientation means a communication board with the date and the day's schedule, orientation to person, place and time, familiar objects from home, and structured therapeutic activities three times daily (reminiscence, word games, discussion of current events).

Sleep is protected without drugs: fewer night-time interruptions, warm milk or herbal tea, relaxation music, unit-wide quiet time, lights down at night and a normal circadian rhythm. Hydration means 1500-2000 mL of fluid a day with help feeding and supplements if needed, and no prolonged fasting, dehydration or electrolyte imbalance. Vision and hearing mean clean glasses to hand, adequate lighting and large-print materials, hearing aids in place and working, amplification devices and less background noise. Pain is managed with regular paracetamol, regional anaesthesia where possible and the lowest effective opioid dose, avoiding pethidine and excessive opioids in the opioid-naive.

What HELP achieves, and its precise limits. In the original trial delirium days fell (105 vs 161) and episodes fell (62 vs 90), but severity and recurrence once delirium had developed were not reduced. That is the single most important thing to know about HELP: it prevents onset and does very little for the patient who is already delirious, which is exactly why prevention is the priority. The absolute reduction was 5.1%, an NNT of about 20 in a general medical population; in hip fracture, where baseline risk is far higher, Marcantonio's trial gives an NNT of 5.6. Health-economic analyses find it cost-effective through fewer delirium days and complications, and the effect is confirmed across settings by Cochrane (multicomponent prevention RR 0.69, PMID 26967259).

Medication review. Every drug on the chart is a candidate precipitant, and a handful of classes do most of the damage.

The Beers criteria. The American Geriatrics Society Beers Criteria list potentially inappropriate medications in older adults. The deliriogenic ones to avoid are benzodiazepines (especially the long-acting diazepam and flurazepam), anticholinergics (diphenhydramine, hydroxyzine, promethazine), tricyclic antidepressants (amitriptyline), first-generation antipsychotics in high doses, pethidine and pentazocine.

Orthogeriatric co-management. Proactive orthogeriatric care means dedicated geriatricians working alongside the orthopaedic team to optimise perioperative medical management, rather than being called when the patient becomes confused. The Van Heghe 2022 meta-analysis (PMID 34591127) found it cuts delirium by about 19% (RR 0.81, 95% CI 0.71-0.92), shortens stay by about 1.55 days, and lowers in-hospital mortality by 28% and 1-year mortality by 14%; it is broadly cost-effective and recommended in major hip fracture guidelines and registries internationally. The service comprises:

• Daily geriatrician review, starting within 24 hours of admission
• Comprehensive geriatric assessment (CGA)
• Medication optimisation
• Delirium screening and prevention protocols
• Management of comorbidities
• Discharge planning and rehabilitation coordination

The registry standard. The Australia and New Zealand Hip Fracture Registry (ANZHFR) asks for orthogeriatric care within 24 hours, surgery within 48 hours of admission, standardised delirium screening and multimodal delirium prevention.

View larger