Residency · Residency · Geriatrics

Delirium - Prevention, Diagnosis, and Management

Introduction

Delirium is an acute, fluctuating disturbance in attention, awareness, and cognition that develops over hours to days, representing one of the most common and most dangerous medical emergencies encountered in geriatric medicine. The incidence varies substantially by clinical setting: 10 to 31 percent of hospitalized medical patients develop delirium, the rate rises to 15 to 53 percent in postoperative populations, and it reaches a staggering 50 to 80 percent among mechanically ventilated ICU patients.

The consequences of delirium are severe and far-reaching. Delirium is associated with a two- to three-fold increase in mortality, prolongs hospitalization by 5 to 10 days, carries a ten-fold increased risk of subsequent dementia, and significantly increases the likelihood of institutional placement. The annual healthcare costs attributed to delirium in the United States are estimated at $164 billion, rivaling the economic burden of diabetes and cardiovascular disease. Despite these devastating consequences, delirium is frequently underdiagnosed, with studies indicating that treating physicians miss the diagnosis in 60 to 70 percent of cases, particularly the hypoactive subtype.

Pathophysiology

Neuroinflammatory Hypothesis

The leading pathophysiological model posits that systemic inflammation, whether triggered by infection, surgery, trauma, or other acute insults, results in microglial activation and neuroinflammation within the central nervous system. Breakdown of the blood-brain barrier permits peripheral inflammatory mediators, including IL-1, IL-6, and TNF-alpha, to access the CNS, where they disrupt normal neurotransmission, particularly within cholinergic circuits. Preexisting neurodegeneration, as seen in dementia, lowers the threshold for delirium by reducing the cognitive reserve available to compensate for acute perturbations, explaining why dementia is the single strongest risk factor for delirium.

Neurotransmitter Imbalance

Cholinergic deficiency is central to delirium pathogenesis, and anticholinergic medications represent the most common pharmacological precipitant. Serum anticholinergic activity has been shown to correlate with delirium severity, providing direct evidence for this mechanism. Relative or absolute dopaminergic excess appears to contribute, which explains the partial efficacy of antipsychotic medications in managing symptoms. GABAergic dysregulation accounts for the ability of both benzodiazepines (as GABA agonists) and alcohol withdrawal (as loss of GABA agonism) to trigger delirium. Serotonergic, glutamatergic, and melatonergic pathways are also implicated, and HPA axis hyperactivation with elevated cortisol levels exerts direct neurotoxic effects, particularly on the hippocampus.

Predisposing and Precipitating Factor Model (Inouye & Charpentier, 1996)

The conceptual framework developed by Inouye and Charpentier elegantly models delirium as resulting from the interaction between baseline vulnerability (predisposing factors) and acute insults (precipitating factors). In a patient with high baseline vulnerability, such as one with advanced dementia, even minor precipitants such as a urinary tract infection or a single dose of diphenhydramine may be sufficient to trigger delirium. Conversely, in a patient with low baseline vulnerability, such as a cognitively intact young adult, a severe precipitant such as major surgery under general anesthesia, sepsis, or prolonged ICU stay is typically required. This model has profound implications for prevention, as it identifies both the patients most at risk and the modifiable factors most amenable to intervention.

Risk Factors

Predisposing Factors (Baseline Vulnerability)

Dementia stands as the single strongest predisposing risk factor for delirium, with an odds ratio of 2 to 5, and is present in 25 to 50 percent of patients who develop delirium. Age 65 and older carries an odds ratio of 2 to 3. Prior delirium history, functional impairment with ADL dependence, sensory impairment in vision or hearing, high comorbidity burden as measured by a Charlson Comorbidity Index of 3 or greater, depression, alcohol use disorder, and malnutrition all contribute to baseline vulnerability.

Precipitating Factors

Medications are among the most important and modifiable precipitating factors. Anticholinergic medications, benzodiazepines, opioids (especially meperidine), corticosteroids, and fluoroquinolones are the most frequently implicated classes. Anticholinergic burden can be quantified using the Anticholinergic Cognitive Burden (ACB) scale, with a cumulative score of 3 or greater indicating high risk.

Infection is a common precipitant, though it is important to note that urinary tract infection is frequently overdiagnosed as the cause of delirium, and clinicians should always seek additional precipitants rather than attributing delirium solely to bacteriuria. Surgery, particularly cardiac, orthopedic, and abdominal procedures, carries substantial delirium risk. Both undertreated pain and opioid exposure can precipitate delirium, creating a clinical tension that requires careful management. Urinary retention, fecal impaction, metabolic derangements (hyponatremia, hypernatremia, hypercalcemia, hepatic and renal failure, hypoglycemia), hypoxia, sleep deprivation, physical restraints (odds ratio 4.4), indwelling catheters (odds ratio 2.4), and the ICU environment itself are additional precipitants.

<image>A two-panel risk factor model for delirium. The left panel shows a balance beam or seesaw diagram illustrating the predisposing-precipitating factor interaction model. On one side, stack predisposing factors (dementia, age, functional impairment, sensory deficits, comorbidities) as weighted blocks. On the other side, stack precipitating factors (medications, infection, surgery, pain, metabolic derangements) as weighted blocks. Show that when the combined weight exceeds a threshold, the beam tips into "DELIRIUM." Show two example scenarios: (1) a patient with heavy predisposing factors needing only a small precipitant, and (2) a robust patient requiring a large precipitant. The right panel should list the top 10 medications causing delirium with their anticholinergic burden scores.</image>

Clinical Features and Subtypes

Core Features (DSM-5 Criteria)

The DSM-5 diagnostic criteria require five elements for a diagnosis of delirium. Criterion A specifies a disturbance in attention, manifested as reduced ability to direct, focus, sustain, and shift attention, along with reduced awareness of and orientation to the environment. Criterion B requires that the disturbance develops over hours to days, represents a change from baseline attention and awareness, and tends to fluctuate in severity during the course of the day. Criterion C describes additional cognitive disturbances that may include memory impairment, disorientation, language dysfunction, visuospatial impairment, or perceptual disturbances. Criterion D requires that the disturbance not be better explained by a pre-existing neurocognitive disorder and that it not occur in the context of a severely reduced level of arousal such as coma. Criterion E mandates evidence from the history, physical examination, or laboratory findings that the disturbance is a direct physiological consequence of a medical condition, substance intoxication or withdrawal, toxin exposure, or multiple etiologies.

Motor Subtypes

Delirium presents in three motor subtypes with markedly different clinical profiles and prognoses. Hyperactive delirium, accounting for approximately 25 percent of cases, is characterized by agitation, restlessness, pulling at lines and tubes, hallucinations, and delusions. This subtype is the most easily recognized but represents the minority of cases. Hypoactive delirium is the most common subtype, accounting for approximately 50 percent of cases, and manifests as lethargy, reduced spontaneous movement, flat affect, and social withdrawal. It is the subtype most commonly missed by clinical teams and carries the worst prognosis, with increased rates of pressure injuries, aspiration, pulmonary embolism, and mortality. Mixed delirium, representing approximately 25 percent of cases, features alternating episodes of hyperactive and hypoactive behavior.

Diagnostic Assessment

Screening and Diagnostic Tools

The Confusion Assessment Method (CAM) is the gold standard diagnostic instrument for non-ICU settings. It assesses four features: acute onset or fluctuating course, inattention, disorganized thinking, and altered level of consciousness. Diagnosis requires the presence of features 1 and 2 plus either feature 3 or 4. When used by trained assessors, the CAM achieves sensitivity of 94 to 100 percent and specificity of 90 to 95 percent. However, sensitivity drops precipitously to 46 percent when used by untrained bedside nurses, underscoring the critical importance of proper training in CAM administration.

The CAM-ICU is an adaptation for intubated and non-verbal patients, beginning with assessment of arousal using the Richmond Agitation-Sedation Scale (RASS). If the RASS is negative 3 or higher, the assessment proceeds to evaluate attention through the Attention Screening Examination, disorganized thinking through a series of yes/no questions and a simple command, and altered consciousness through RASS deviation from zero. The 4AT is a rapid screening tool requiring less than two minutes that does not require specialized training, with a score of 4 or greater suggesting probable delirium. The 3D-CAM is a brief three-minute structured CAM-based assessment suitable for non-ICU settings. The Delirium Rating Scale-Revised-98 (DRS-R-98), scored from 0 to 46, is the best instrument for monitoring delirium severity over time.

Workup for Underlying Cause

The investigation of delirium etiology should be systematic and thorough. A targeted history should focus on medication changes within the preceding 72 hours, new symptoms, and recent procedures. Physical examination should include vital signs with oxygen saturation, assessment of hydration status, identification of focal neurological deficits, evaluation of surgical wounds, and abdominal examination for urinary retention and fecal impaction. Standard laboratory evaluation includes complete blood count, basic metabolic panel with sodium, calcium, glucose, and renal function assessment, urinalysis, and hepatic function tests. Additional studies should be obtained based on clinical suspicion: thyroid-stimulating hormone, vitamin B12, ammonia, troponin, blood cultures, and drug levels for medications with narrow therapeutic windows.

An electrocardiogram should be obtained to evaluate for arrhythmia and to document baseline QTc interval before antipsychotic use. Neuroimaging is not routine and should be reserved for patients with focal neurological deficits, head trauma, absence of an identifiable precipitant, or suspicion of CNS pathology. Lumbar puncture is indicated only when meningitis or encephalitis is suspected. Electroencephalography may be useful when nonconvulsive status epilepticus is a concern, and characteristically shows generalized slowing in delirium.

Prevention — The Most Effective Strategy

HELP (Hospital Elder Life Program) — Inouye et al.

The Hospital Elder Life Program (HELP), developed by Sharon Inouye and colleagues, is the most extensively studied and validated multicomponent non-pharmacological intervention for delirium prevention. Targeting six established delirium risk factors, HELP reduces delirium incidence by 33 to 40 percent, corresponding to a number needed to treat of 5 to 9.

The six components of HELP are orientation protocols (including boards with date, names, and daily schedule along with systematic reorientation), cognitive stimulation (through therapeutic activities, reminiscence therapy, and word games), sleep enhancement (using warm milk or herbal tea, relaxation techniques, noise reduction, and minimization of nighttime interruptions), early mobilization (with ambulation three times daily, active range-of-motion exercises, and minimization of physical restraints), vision and hearing optimization (ensuring glasses and hearing aids are available and functioning), and hydration and nutrition management (encouraging oral intake, monitoring fluid balance, and addressing dehydration). The cost savings associated with HELP are estimated at $1,600 to $6,400 per case of delirium prevented, making it cost-effective as well as clinically effective. HELP is endorsed by the American Geriatrics Society, NICE, and the American College of Surgeons.

Pharmacological Prophylaxis

Pharmacological prophylaxis for delirium is generally not recommended, as no pharmacological agent has demonstrated robust evidence for delirium prevention. The REDUCE trial by van den Boogaard and colleagues in 2018 showed that intravenous haloperidol prophylaxis in ICU patients did not reduce delirium incidence, duration, or mortality. Melatonin and ramelteon have produced limited and conflicting evidence, with one small RCT by Al-Aama et al. in 2011 suggesting reduced delirium incidence with ramelteon 8 mg nightly. Dexmedetomidine may reduce delirium in post-cardiac surgery ICU patients (PRODUCE trial) but cannot be broadly generalized. Neither ketamine as an adjunct to general anesthesia (PODCAST trial, 2017) nor cholinesterase inhibitors (donepezil in the NICE trial) have demonstrated preventive benefit.

<image>A hospital room illustration showing evidence-based non-pharmacological delirium prevention strategies (HELP program) in action. Show a well-lit hospital room with: a large clock and orientation board visible on the wall with date, location, and care team names; a patient wearing their own glasses and hearing aids; a water pitcher and cup within reach; family photos on the bedside table; a mobility aid (walker) accessible nearby; curtains open to natural light; a "no unnecessary interruptions" sign for nighttime; a volunteer sitting with the patient doing a cognitive activity (puzzle or word game). Include callout labels for each intervention. Show the absence of physical restraints and urinary catheter. The room should appear calm, organized, and patient-centered.</image>

Management of Established Delirium

Non-Pharmacological Management (First-Line, Always)

Non-pharmacological approaches constitute the foundation of delirium management and should always be implemented regardless of whether pharmacotherapy is also required. The first priority is identifying and treating the underlying cause or causes, which are often multifactorial, with a median of two to three precipitants per delirium episode.

Medication review with discontinuation or dose reduction of offending agents, particularly anticholinergics, benzodiazepines, and opioids, is essential. Reorientation strategies should be implemented, including consistent caregivers, family presence, familiar objects from home, and visible clocks and calendars. Sleep-wake cycle restoration requires daytime light exposure, nighttime darkness and quiet, and avoidance of nighttime medications and vital sign checks when possible. Mobilization, with getting patients out of bed to a chair and ambulating as tolerated, addresses deconditioning and helps restore circadian rhythms. Sensory optimization through provision of glasses, hearing aids, and dentures is a simple but frequently overlooked intervention.

Hydration and nutrition should be maintained with adequate oral intake while avoiding intravenous fluid overload. Pain management should employ scheduled non-opioid analgesia, particularly acetaminophen, while minimizing opioid use without leaving pain untreated. Tethers, including Foley catheters, intravenous lines, and telemetry monitors, should be removed when not essential. Physical restraints must be avoided because they increase delirium duration and severity, with an odds ratio of 4.4 for delirium persistence. Family engagement, including education about delirium and encouragement of presence and familiar conversation, is a powerful non-pharmacological intervention.

Pharmacological Management (Reserved for Severe Agitation Threatening Safety)

Pharmacological management is indicated only when non-pharmacological measures have failed and the patient is at imminent risk of harm, such as pulling life-sustaining devices or experiencing severe distress. The available evidence for antipsychotic efficacy in delirium is modest at best. Haloperidol at 0.5 to 1 mg PO/IV/IM every 4 to 6 hours as needed, with a maximum of 3 to 5 mg per day in elderly patients, remains the most commonly used agent. Quetiapine at 12.5 to 25 mg PO twice daily is favored in patients with Parkinson disease or Lewy body dementia. Olanzapine at 2.5 to 5 mg PO/IM provides sedation but has anticholinergic properties. Risperidone at 0.25 to 0.5 mg PO twice daily is another option.

Importantly, the AH-HA-2 trial by Girard and colleagues in 2018 and the MIND-USA trial both demonstrated that haloperidol and ziprasidone did not reduce delirium duration compared to placebo in ICU patients, confirming that antipsychotics are symptom management tools rather than disease-modifying therapies. All antipsychotics should be used at the lowest effective dose for the shortest possible duration, with daily reassessment and prompt discontinuation when agitation resolves.

All antipsychotics carry an FDA black box warning for increased mortality in elderly patients with dementia, with a relative risk of 1.6 to 1.7 and a number needed to harm of 53 to 100. QTc monitoring is mandatory, with a baseline ECG obtained before initiation and the medication held if the QTc exceeds 500 milliseconds.

Dexmedetomidine, an alpha-2 agonist administered at 0.2 to 0.7 mcg/kg/hr intravenously, is useful in the ICU setting for delirium with agitation and offers the advantage of lacking respiratory depressant effects. Benzodiazepines should be avoided except in the specific contexts of alcohol or benzodiazepine withdrawal and hepatic encephalopathy, as lorazepam has been identified as an independent risk factor for delirium with an odds ratio of 1.2 per milligram administered.

Special Populations

Alcohol withdrawal delirium (delirium tremens) requires benzodiazepines as first-line treatment, with lorazepam or chlordiazepoxide dosed according to symptom-triggered protocols such as the CIWA scale. Hepatic encephalopathy should be managed with lactulose and rifaximin while avoiding antipsychotics, which lower the seizure threshold. Patients with Parkinson disease or dementia with Lewy bodies require strict avoidance of haloperidol due to the risk of severe neuroleptic sensitivity reactions, including profound rigidity, obtundation, autonomic instability, and a neuroleptic malignant syndrome-like presentation; quetiapine or pimavanserin are preferred alternatives. Terminal delirium, which may represent an irreversible component of the dying process, should be managed with a goal of comfort, and palliative sedation should be discussed when delirium becomes refractory to standard interventions.

Outcomes and Prognosis

The average duration of a delirium episode is 5 to 7 days, though the range extends from hours to months. Persistent delirium at hospital discharge is present in 20 to 40 percent of patients, and 70 percent of delirium survivors demonstrate cognitive impairment at 12 months. Delirium superimposed on dementia, which occurs in 22 to 89 percent of hospitalized dementia patients, accelerates the trajectory of cognitive decline. Perhaps most sobering, delirium independently increases the risk of subsequent dementia with an odds ratio of 8.7 at three years, as demonstrated by Davis and colleagues in 2012. Additionally, 15 to 25 percent of patients recall distressing delirium experiences and may develop PTSD-like symptoms.

<image>A timeline infographic showing the trajectory and outcomes of a delirium episode. Start with "baseline cognition" on the left, then show the acute decline at delirium onset (triggered by a precipitant like surgery or infection). Show the fluctuating course during the delirium episode (waxing and waning attention represented as a jagged line). Then show three possible recovery trajectories diverging: (1) full recovery to baseline (30-40%), (2) partial recovery with persistent cognitive impairment (40-50%), and (3) progressive decline to dementia/death (10-20%). Mark key time points: onset, diagnosis, treatment of precipitant, resolution of acute delirium, and 12-month follow-up. Include outcome statistics at each trajectory endpoint. Show "intervention window" highlighting the period where early recognition and treatment can shift patients toward the better trajectory.</image>

Key Clinical Pearls

• Delirium is a medical emergency — mortality rates rival STEMI and sepsis; treat the urgency accordingly
• Hypoactive delirium is the most common subtype (50%) and the most missed — maintain a high index of suspicion in any patient who is "pleasantly confused" or "just sleepy"
• The CAM has 94-100% sensitivity when used correctly, but sensitivity drops to 46% with untrained users — invest in training
• Prevention (HELP program) is more effective than any treatment — multicomponent non-pharmacological strategies reduce delirium by 33-40%
• Antipsychotics do NOT shorten delirium duration (AH-HA-2, MIND-USA) — they are symptom management only, reserved for severe agitation
• Every episode of delirium causes lasting harm — it independently increases the risk of subsequent dementia by 8-fold
• Always ask: "What medications were changed in the last 72 hours?" — medication changes are the most modifiable precipitant

References

1. Inouye SK, Westendorp RGJ, Saczynski JS. Delirium in elderly people. Lancet. 2014;383(9920):911-922.
2. Girard TD, Exline MC, Carson SS, et al. Haloperidol and ziprasidone for treatment of delirium in critical illness. N Engl J Med. 2018;379(26):2506-2516.
3. Inouye SK, Bogardus ST Jr, Charpentier PA, et al. A multicomponent intervention to prevent delirium in hospitalized older patients. N Engl J Med. 1999;340(9):669-676.
4. Davis DHJ, Muniz-Terrera G, Keage HAD, et al. Association of delirium with cognitive decline in late life: a neuropathologic study of 3 population-based cohort studies. JAMA Psychiatry. 2017;74(3):244-251.
5. Oh ES, Fong TG, Hshieh TT, Inouye SK. Delirium in older persons: advances in diagnosis and treatment. JAMA. 2017;318(12):1161-1174.

Read this lecture as Markdown