Anda di halaman 1dari 17

The n e w e ng l a n d j o u r na l of m e dic i n e

review article

Current Concepts

Delirium in Older Persons

Sharon K. Inouye, M.D., M.P.H.

elirium, an acute decline in attention and cognition, is a com- From the Department of Medicine, Har-
mon, life-threatening, and potentially preventable clinical syndrome among vard Medical School, Boston. Address
reprint requests to Dr. Inouye at the Ag-
persons who are 65 years of age or older. The development of delirium often ing Brain Center, Hebrew Senior Life,
initiates a cascade of events culminating in the loss of independence, an increased 1200 Centre St., Boston, MA 02131.
risk of morbidity and mortality, and increased health care costs.1-6 Delirium in hos-
N Engl J Med 2006;354:1157-65.
pitalized older patients has assumed particular importance because the care of such Copyright © 2006 Massachusetts Medical Society.
patients accounts for more than 49 percent of all hospital days.7 Delirium compli-
cates hospital stays for at least 20 percent of the 12.5 million patients 65 years of
age or older who are hospitalized each year and increases hospital costs by $2,500
per patient,8-10 so that about $6.9 billion (value in U.S. dollars in 2004) of Medicare
hospital expenditures are attributable to delirium. Substantial additional costs ac-
crue after hospital discharge because of the need for institutionalization, rehabili-
tation services, formal home health care, and informal caregiving.
This report examines current clinical practice in delirium, identifies areas of
controversy, and highlights areas for future research.

Epide miol o gy a nd Di agnos t ic Cr i ter i a

In direct contrast to dementia, which is a chronic confusional state, delirium is an

acute confusional state. Rates of delirium are highest among hospitalized older
patients, and the rates vary depending on the patients’ characteristics, setting of care,
and sensitivity of the detection method. The prevalence of delirium at hospital ad-
mission ranges from 14 to 24 percent, and the incidence of delirium arising during
hospitalization ranges from 6 to 56 percent among general hospital populations.11,12
Delirium occurs in 15 to 53 percent of older patients postoperatively11 and in 70 to
87 percent of those in intensive care.13 Delirium occurs in up to 60 percent of pa-
tients in nursing homes or post–acute care settings14,15 and in up to 83 percent of
all patients at the end of life.16,17 Although the overall prevalence of delirium in the
community is only 1 to 2 percent,18,19 the prevalence increases with age, rising to
14 percent among those more than 85 years old. Moreover, in 10 to 30 percent of
older patients presenting to emergency departments, delirium is a symptom11 that
often heralds the presence of life-threatening conditions. The mortality rates among
hospitalized patients with delirium range from 22 to 76 percent,20 as high as the
rates among patients with acute myocardial infarction or sepsis. The one-year mor-
tality rate associated with cases of delirium is 35 to 40 percent.21
The diagnosis of delirium is primarily clinical and is based on careful bedside
observation of key features. Although the criteria continue to evolve,22,23 the diag-
nostic algorithms that are most widely used are presented in the Supplementary
Appendix (available with the full text of this article at Delirium is
often unrecognized by the patients’ physicians and nurses,1,24 in part because of
its fluctuating nature, its overlap with dementia, lack of formal cognitive assess-
ment, underappreciation of its clinical consequences, and failure to consider the
diagnosis important.

n engl j med 354;11 march 16, 2006 1157

The n e w e ng l a n d j o u r na l of m e dic i n e

Cl inic a l Ch a r ac ter is t ic s Dopaminergic excess also appears to contrib-

ute to delirium, possibly owing to its regulatory
Because delirium remains a bedside diagnosis, influence on the release of acetylcholine.29 Do-
understanding its clinical features (Table 1 and paminergic drugs (e.g., levodopa and bupropion)
the Supplementary Appendix) is crucial to the di- are recognized precipitants of delirium, and do-
agnosis of delirium. Delirium has hypoactive and pamine antagonists (e.g., antipsychotic agents) ef-
hyperactive forms (Table 1). The hypoactive form fectively treat delirium symptoms. Perturbations
of delirium is more common among older persons of other neurotransmitters, such as norepineph-
and often goes unrecognized. rine, serotonin, γ-aminobutyric acid, glutamate,
and melatonin, may also have a role in the patho-
Etiologic and Risk Factors physiology of delirium, but the evidence is less well
The cause of delirium is typically multifactorial.25 developed.1,29,32 These neurotransmitters may ex-
In fact, the development of delirium involves the ert their influence through interactions with the
complex interrelationship between a vulnerable cholinergic and dopaminergic pathways.
patient (one with predisposing factors) (Table 2) Cytokines, including interleukin-1, interleukin-
and exposure to precipitating factors or noxious 2, interleukin-6, tumor necrosis factor α (TNF-α),
insults (Table 3).26,27 Thus, in patients who are and interferon, may contribute to delirium1,29,33
highly vulnerable to delirium, such as those with by increasing the permeability of the blood–
dementia and multiple coexisting conditions, it brain barrier and altering neurotransmission.
may develop as a result of relatively benign in- Finally, chronic stress brought on by illness or
sults, such as one dose of a sleeping medication. trauma activates the sympathetic nervous system
Conversely, in patients who are not vulnerable to and hypothalamic–pituitary–adrenocortical axis,
delirium, it develops only after exposure to mul- resulting in increased cytokine levels and chron-
tiple noxious insults, such as general anesthesia, ic hypercortisolism.34 Chronic hypercortisolism
major surgery, and psychoactive medications. Ad- has deleterious effects on hippocampal serotonin
dressing just one contributing factor is unlikely to (5-hydroxytryptamine [5-HT]) 5-HT1A receptors,
resolve delirium in an older person; they should which may contribute to delirium.29,35,36 Given
all be addressed when possible. the clinical heterogeneity and multifactorial na-
ture of delirium, it is likely that multiple patho-
Pathogenesis genic mechanisms contribute to the development
The pathophysiology of delirium remains poorly of delirium.
understood. Electroencephalographic studies have
demonstrated diffuse slowing of cortical back- Approach to Evaluation
ground activity, which does not correlate with un- A flowchart for the prevention and management
derlying causes.28 Neuropsychological and neuro- of delirium from the time of admission of an older
imaging studies reveal generalized disruption in patient is shown in Figure 1. This approach, based
higher cortical function, with dysfunction in the on current clinical guidelines and expert opin-
prefrontal cortex, subcortical structures, thalamus, ion,9,20 must be guided by the individual patient’s
basal ganglia, frontal and temporoparietal cortex, medical history, findings on physical and neuro-
fusiform cortex, and lingual gyri, particularly on logic examination, and clinical setting. Although
the nondominant side.29,30 The leading hypothe- the provision of detailed procedures is beyond the
ses for the pathogenesis of delirium focus on the scope of this report, we will highlight common
roles of neurotransmission, inflammation, and pitfalls to avoid. When a patient with confusion
chronic stress. is admitted, determining the acuity of the change
Extensive evidence supports the role of cholin- in mental status is the essential first step. Neglect-
ergic deficiency. Administration of anticholinergic ing this step is the leading reason for missing the
drugs can induce delirium in humans and ani- diagnosis of delirium. If no history can be ob-
mals, and serum anticholinergic activity is in- tained, then the patient should be assumed to be
creased in patients with delirium. Physostigmine delirious until proved otherwise. Every older hos-
reverses delirium associated with anticholinergic pitalized patient should undergo brief but formal
drugs, and cholinesterase inhibitors appear to have cognitive testing with the use of instruments such
some benefit even in cases of delirium that are not as the Mini–Mental State Examination and the
induced by drugs.15,29,31 Confusion Assessment Method, since subtle de-

1158 n engl j med 354;11 march 16, 2006

current concepts

Table 1. Clinical Features of Delirium.*

Acute onset
Occurs abruptly, usually over a period of hours or days
Reliable informant often needed to ascertain the time course of onset
Fluctuating course
Symptoms tend to come and go or increase and decrease in severity over a 24-hour period
Characteristic lucid intervals
Difficulty focusing, sustaining, and shifting attention
Difficulty maintaining conversation or following commands
Disorganized thinking
Manifested by disorganized or incoherent speech
Rambling or irrelevant conversation or an unclear or illogical flow of ideas
Altered level of consciousness
Clouding of consciousness, with reduced clarity of awareness of the environment
Cognitive deficits
Typically global or multiple deficits in cognition, including disorientation, memory deficits, and language impairment
Perceptual disturbances
Illusions or hallucinations in about 30 percent of patients
Psychomotor disturbances
Psychomotor variants of delirium
Marked by agitation and vigilance
Marked by lethargy, with a markedly decreased level of motor activity
Altered sleep–wake cycle
Characteristic sleep-cycle disturbances
Typically daytime drowsiness, nighttime insomnia, fragmented sleep, or complete sleep-cycle reversal
Emotional disturbances
Manifested by intermittent and labile symptoms of fear, paranoia, anxiety, depression, irritability, apathy, anger,
or euphoria

* Additional details are provided in Table 2 in the Supplementary Appendix.

lirium is often missed. Older patients should be underlying disease. All preadmission and current
aroused during rounds and evaluated daily for medications should be reviewed; even long-stand-
the hypoactive form of delirium, which is often ing medications can contribute to delirium and
overlooked. should be reevaluated. If changes in long-term
When clinicians search for the underlying cause medications are appropriate after the indications
of delirium, they need to be aware of the possibil- and risk–benefit ratios have been carefully weighed,
ity of occult or atypical presentations of many the hospital represents the ideal venue for making
diseases in the elderly, including myocardial in- these changes. A medical history must be meticu-
farction, infection, and respiratory failure, because lously obtained to detect occult alcohol or benzo-
delirium is often the sole manifestation of serious diazepine use, which can contribute to delirium.

n engl j med 354;11 march 16, 2006 1159

The n e w e ng l a n d j o u r na l of m e dic i n e

ma, those with fever and acute changes in mental

Table 2. Predisposing Factors for Delirium.
status in whom encephalitis is suspected, or those
Demographic characteristics with no other identifiable cause of the delirium.12
Age of 65 years or older However, neuroimaging should be considered
Male sex when the history cannot be obtained or the neu-
Cognitive status
rologic examination cannot be completed (e.g.,
when the patient is combative) so as not to miss
uncommon conditions that are life-threatening
Cognitive impairment but treatable, such as subarachnoid hemorrhage
History of delirium and encephalitis.
Functional status Pr e v en t ion a nd M a nage men t
Functional dependence
Preventing delirium is the most effective strategy
for reducing its frequency and complications. Suc-
Low level of activity
cessful preventive strategies include multicompo-
History of falls nent approaches to reduce risk factors. Because
Sensory impairment delirium has many causes, multicomponent ap-
Visual impairment proaches represent the most effective and clini-
Hearing impairment cally relevant ones. The Yale Delirium Prevention
Decreased oral intake
Trial8 demonstrated the effectiveness of interven-
tion protocols targeted toward six risk factors:
orientation and therapeutic activities for cognitive
Malnutrition impairment, early mobilization to avert immobi-
Drugs lization, nonpharmacologic approaches to mini-
Treatment with multiple psychoactive drugs mize the use of psychoactive drugs, interventions
Treatment with many drugs to prevent sleep deprivation, communication meth-
Alcohol abuse ods and adaptive equipment (particularly eye-
glasses and hearing aids) for vision and hearing
Coexisting medical conditions
impairment, and early intervention for volume de-
Severe illness
pletion. A randomized clinical trial involving pa-
Multiple coexisting conditions tients who had had hip fractures demonstrated
Chronic renal or hepatic disease the effectiveness of a multicomponent strategy
History of stroke for geriatric consultation targeted toward 10
Neurologic disease domains38: oxygen delivery to the brain, fluid
Metabolic derangements
and electrolyte balance, pain management, re-
duction in the use of psychoactive drugs, bowel
Fracture or trauma
and bladder function, nutrition, early mobiliza-
Terminal illness tion, prevention of postoperative complications,
Infection with human immunodeficiency virus appropriate environmental stimuli, and treatment
of symptoms of delirium.
Electroencephalography has a limited role in Once delirium occurs, the key steps in man-
the diagnosis of delirium, because of its false agement are to address all evident causes, pro-
negative rate of 17 percent and false positive rate vide supportive care and prevent complications,
of 22 percent12; it is most useful for detecting and treat behavioral symptoms. Because delirium
occult seizures and differentiating delirium from can be a medical emergency, the first aim of man-
psychiatric disorders. Neuroimaging studies have agement is to address predisposing and precipi-
a low clinical yield (the number of positive results tating factors promptly (Fig. 1). Supportive care
divided by the total number of studies performed) should include protecting the patient’s airway,
in the evaluation of delirium and should be re- maintaining hydration and nutrition, position-
served for patients with new focal neurologic ing and mobilizing the patient to prevent pressure
signs, those with a history or signs of head trau- sores and deep venous thrombosis, avoiding the

1160 n engl j med 354;11 march 16, 2006

current concepts

use of physical restraints, and supporting the pa-

Table 3. Precipitating Factors or Insults That Can Contribute to Delirium.
tient’s daily care needs. Nonpharmacologic ap-
proaches to managing symptoms of delirium Drugs
should be instituted in every patient. These ap- Sedative hypnotics
proaches include creating a calm, comfortable en- Narcotics
vironment with the use of orienting influences, Anticholinergic drugs
such as calendars, clocks, and familiar objects
Treatment with multiple drugs
from home; regular reorienting communication
with staff members; involving family members in Alcohol or drug withdrawal
supportive care; limiting room and staff chang- Primary neurologic diseases
es; coordinating schedules for administering drugs, Stroke, particularly nondominant hemispheric
obtaining vital signs, and performing procedures Intracranial bleeding
to allow the patient an uninterrupted period for Meningitis or encephalitis
sleep at night with low levels of noise and light-
Intercurrent illnesses
ing; and encouraging normal sleep–wake cycles
by opening blinds and encouraging wakefulness
and mobility during the daytime. Since delirium Iatrogenic complications
may take weeks or months to resolve, patients Severe acute illness
must be cared for in supervised settings. Close Hypoxia
clinical follow-up after discharge is needed, es- Shock
pecially because of the poor long-term prognosis Fever or hypothermia
associated with delirium.
Pharmacologic management should be reserved
for patients whose symptoms of delirium would Dehydration
threaten their own safety or the safety of other Poor nutritional status
persons or would result in the interruption of es- Low serum albumin level
sential therapy, such as mechanical ventilation or Metabolic derangements (e.g., electrolyte, glucose, acid–base)
central venous catheters. Pharmacologic treatment Surgery
strategies are outlined in Table 4.
Orthopedic surgery
Cardiac surgery
R el at ionship be t w een Del ir ium Prolonged cardiopulmonary bypass
a nd De men t i a
Noncardiac surgery
Delirium and dementia are highly interrelated, Environmental
yet the nature of their interrelationship remains Admission to an intensive care unit
poorly examined. Although a cause-and-effect Use of physical restraints
relation has not been established between deliri-
Use of bladder catheter
um and dementia, investigation of their intersec-
tion may yield important insights that will ad- Use of multiple procedures
vance our understanding of both conditions (see Pain
Table 3 in the Supplementary Appendix). Demen- Emotional stress
tia is the leading risk factor for delirium, and fully Prolonged sleep deprivation
two thirds of cases of delirium occur in patients
with dementia.1,39 Thus, the underlying vulnera-
bility of the brain in patients with dementia may ties of persistent delirium4,40-44 and reversible de-
predispose them to the development of delirium mentia45 blurs the boundaries between these con-
as a result of insults related to acute medical ill- ditions. Moreover, studies have shown that delirium
nesses, medications, or environmental perturba- and dementia are both associated with decreased
tions. Recent studies suggest that delirium per- cerebral metabolism, cholinergic deficiency, and
sists much longer than previously believed,4,40-44 inflammation,46 reflecting their overlapping clini-
with symptoms in many patients lasting months cal, metabolic, and cellular mechanisms. In fact,
or years. The existence of the well-described enti- delirium and dementia may represent points along

n engl j med 354;11 march 16, 2006 1161

The n e w e ng l a n d j o u r na l of m e dic i n e

a continuum of cognitive disorders, rather than two

Figure 1 (facing page). Prevention and Management
entirely separate conditions.39 of Delirium in the Older Hospitalized Patient.
Does delirium contribute to dementia? Al- Thyroid-function tests include measurement of the
though it is not likely that the delirium itself thyroxine level, thyroid index, and thyrotropin level.
causes the pathologic changes of dementia, there More information about pharmacologic treatment can
is no question that delirium contributes to wors- be found in Table 4.
ening functional status, loss of independence, and
poorer outcomes among patients with dementia. sequelae. Studies investigating the pathogenesis
The long-standing traditional view is that delir- of delirium with the use of neuropsychological
ium and dementia are two separate conditions; testing, neuroimaging methods, electrophysiologi-
however, emerging evidence has highlighted their cal methods, laboratory markers, genetic studies,
overlap. First, epidemiologic studies have docu- and neuropathological approaches are greatly
mented long-term cognitive decline in patients needed. Investigation of delirium provides an
with delirium, after controlling for relevant co- important opportunity to clarify the link between
variates.47 Second, several causes of delirium may brain pathophysiology and behavioral manifesta-
not be completely reversible, particularly those tions, which might hold broader implications for
resulting in neuronal injury and permanent cog- other cognitive and psychiatric disorders. New
nitive sequelae, such as prolonged hypoxia or hy- prospects for therapy include strategies to in-
poglycemia.39 Third, neuroimaging studies dem- crease acetylcholine activity in the brain (e.g.,
onstrate regions of hypoperfusion in patients with through the use of procholinergic agents and
delirium.48 Thus, delirium may herald the onset avoidance of highly anticholinergic drugs), the
of dementia in many instances. Fourth, dementia use of selective dopamine antagonists that affect
with Lewy bodies, which includes fluctuating cog- D1, D2, D3, and D4 receptors differently, and the
nition and visual hallucinations as core signs, il- use of drugs to enhance cerebrovascular flow (e.g.,
lustrates the overlap of delirium and dementia. antiinflammatory or antiplatelet agents). Finally,
Delirium can alter the course of an underly- targeting delirium with new therapeutic approach-
ing dementia, with dramatic worsening of the es may offer opportunities for early intervention,
trajectory of cognitive decline, resulting in more preservation of cognitive-reserve capacity, and
rapid progression of functional losses and worse prevention of permanent cognitive damage, which
long-term outcomes. This phenomenon has been may potentially delay or abate the ultimate de-
well recognized clinically in elderly patients with velopment of dementia.
dementia: clinicians and family members have
noted that the patients “never returned to base- Delirium as an Indicator of the Quality
line” after an episode of delirium. In follow-up of Health Care
studies, patients in whom delirium develops have Delirium represents one of the most common
worse outcomes than those with dementia alone,49 preventable adverse events among older persons
including worsened cognitive function and in- during hospitalization53,54 and meets Williamson’s
creased rates of hospitalization, institutionaliza- criteria for an indicator of the quality of health
tion, and death.49-52 care55: the condition is common, frequently iatro-
Delirium may serve as an important model for genic, and integrally linked to processes of care.
research by offering a unique approach to advance Although many cases of delirium may be unavoid-
our general understanding of cognitive disorders able, clinical trials8,38 provide compelling evidence
and dementias (see Table 3 in the Supplementary that at least 30 to 40 percent of cases may be pre-
Appendix). The development of delirium in cer- ventable. Many aspects of hospital care contribute
tain persons may help to identify those who are to the development of delirium, including adverse
vulnerable to cognitive decline through genetic effects of medications, complications of invasive
predisposition or through the presence of early procedures, immobilization, malnutrition, dehy-
dementia or mild cognitive impairment that may dration, the use of bladder catheters, and sleep
otherwise remain unidentified. Moreover, a bet- deprivation.9 Delirium is currently included as a
ter understanding of the pathogenesis of deliri- marker of the quality of care and patient safety
um may help elucidate factors that lead directly to by the National Quality Measures Clearinghouse
neuronal injury and, thus, to permanent cognitive of the Agency for Healthcare Research and Qual-

1162 n engl j med 354;11 march 16, 2006

Hospital admission

Monitor cognitive function Prevention of delirium

Perform formal cognitive assessment Address risk factors for delirium
Establish baseline cognitive function Provide orienting communication
and recent changes Encourage early mobilization
Monitor patient for changes in Use visual and hearing aids
mental status Prevent dehydration
Provide uninterrupted sleep time
Avoid psychoactive drugs
Change in mental status

Chronic Acute

Perform cognitive assessment

Perform dementia evaluation

n engl j med 354;11

and evaluation for delirium

Rule out depression, mania,

Delirium confirmed
acute psychosis
Identify and address predisposing Provide supportive care and
Manage symptoms of delirium
current concepts

and precipitating factors prevent complications

Initial evaluation Review medications Prevent complications All patients Patients with severe agitation
Obtain history (including alcohol Review the use of prescription Protect airway, prevent
and benzodiazepine use) drugs, as-needed drugs, aspiration

march 16, 2006

Obtain vital signs over-the-counter drugs, Maintain volume status Pharmacologic management
Nonpharmacologic treatment
Perform physical and neuro- herbal remedies Provide nutritional support Reserve this approach for patients
logic examination Identify psychoactive effects Provide skin care, prevent with severe agitation at risk
Continue delirium prevention
Order selected laboratory tests and drug interactions pressure sores for interruption of essential
Reorient patient, encourage family
Search for occult infection Use mobilization, prevent medical care (e.g., intubation)
deep venous thrombosis, or for patients who pose
Use sitters
pulmonary embolus safety hazard to themselves
Avoid use of physical restraints and
Potential contributing factor Remove or alter potentially or staff
Foley catheters
identified harmful drugs Start low doses and adjust until
Use nonpharmacologic approaches
Change to less noxious alter- effect achieved
for agitation: music, massage,
native Maintain effective dose for
relaxation techniques
Lower doses 2–3 days
Yes No Use of eyeglasses, hearing aids,
Nonpharmacologic approaches
Maintain patient’s mobility and
Further options self-care ability
Evaluate and Normalize sleep–wake cycle,
Order laboratory tests: thyroid-function tests, measurement of drug levels, toxicology screen,
treat as discourage naps, aim for
measurement of ammonia or cortisol levels, test for vitamin B12 deficiency and arterial
appropriate uninterrupted period of sleep
blood gas levels
for each at night
contributing At night, have patient sleep in
factor quiet room with low-level lighting
Lumbar puncture, electroencephalography

The n e w e ng l a n d j o u r na l of m e dic i n e

Table 4. Pharmacologic Treatment of Delirium.

Class and Drug Dose Adverse Effects Comments

Haloperidol 0.5–1.0 mg twice daily orally, with Extrapyramidal symptoms, espe- Usually agent of choice
additional doses every 4 hr as cially if dose is >3 mg per day Effectiveness demonstrated in ran-
needed (peak effect, 4–6 hr) Prolonged corrected QT interval domized, controlled trials20,37
0.5–1.0 mg intramuscularly; ob- on electrocardiogram Avoid intravenous use because of
serve after 30–60 min and re- Avoid in patients with withdrawal short duration of action
peat if needed (peak effect, syndrome, hepatic insuffi-
20–40 min) ciency, neuroleptic malignant
Atypical antipsychotic
Risperidone 0.5 mg twice daily Extrapyramidal effects equivalent Tested only in small uncontrolled
Olanzapine 2.5–5.0 mg once daily to or slightly less than those studies
Quetiapine 25 mg twice daily with haloperidol Associated with increased mortality
Prolonged corrected QT interval rate among older patients with
on electrocardiogram dementia
Lorazepam 0.5–1.0 mg orally, with additional Paradoxical excitation, respirato- Second-line agent
doses every 4 hr as needed* ry depression, oversedation Associated with prolongation and
worsening of delirium symp-
toms demonstrated in clinical
Reserve for use in patients under-
going sedative and alcohol with-
drawal, those with Parkinson’s
disease, and those with neuro-
leptic malignant syndrome
Trazodone 25–150 mg orally at bedtime Oversedation Tested only in uncontrolled studies

* Intravenous use of lorazepam should be reserved for emergencies.

ity (as explained at www.qualitymeasures.ahrq. The changes required to reduce the incidence of
gov/). After adjusting for case mix, higher deliri- delirium on a national scale would require shifts in
um rates would be expected to correlate with local and national policies and system-wide chang-
lower quality of hospital care. The Assessing Care es to provide high-quality care for older persons.9
of Vulnerable Elders Project has ranked delirium Supported in part by grants (R21AG025193 and K24AG00949)
among the top three conditions for which the from the National Institute on Aging.
No potential conflict of interest relevant to this article was
quality of care needs to be improved.56 Total na- reported.
tional costs related to preventable adverse This article is dedicated to the memory of Joshua Bryan In-
events are estimated to be between $17 billion ouye Helfand.
I am indebted to Dr. Joseph Agostini for his helpful review of
and $29 billion per year,57 and delirium may ac- the manuscript and to Sarah Dowal and Patty Fugal for their
count for at least a quarter of these costs.53,54,57,58 assistance with the preparation of the manuscript.

1. Cole MG. Delirium in elderly patients. talized patients. Arch Intern Med 1992; PA, et al. A multicomponent intervention to
Am J Geriatr Psychiatry 2004;12:7-21. 152:334-40. prevent delirium in hospitalized older pa-
2. Inouye SK, Rushing JT, Foreman MD, 5. Murray AM, Levkoff SE, Wetle TT, et tients. N Engl J Med 1999;340:669-76.
Palmer RM, Pompei P. Does delirium con- al. Acute delirium and functional decline 9. Inouye SK, Schlesinger MJ, Lydon TJ.
tribute to poor hospital outcomes? A three- in the hospitalized elderly patient. J Geron- Delirium: a symptom of how hospital
site epidemiologic study. J Gen Intern Med tol 1993;48:M181-M186. care is failing older persons and a win-
1998;13:234-42. 6. O’Keeffe S, Lavan J. The prognostic dow to improve quality of hospital care.
3. Francis J, Kapoor WN. Prognosis after significance of delirium in older hospital Am J Med 1999;106:565-73.
hospital discharge of older medical pa- patients. J Am Geriatr Soc 1997;45:174-8. 10. Department of Health and Human
tients with delirium. J Am Geriatr Soc 1992; 7. Administration on Aging. A profile of Services. 2004 CMS statistics. Washing-
40:601-6. older Americans. Washington, D.C.: De- ton, D.C.: Centers for Medicare and Med-
4. Levkoff SE, Evans DA, Litpzin B, et partment of Health and Human Services, icaid Services, 2004. (CMS publication no.
al. Delirium: the occurrence and persis- 2000. 03445.)
tence of symptoms among elderly hospi- 8. Inouye SK, Bogardus ST Jr, Charpentier 11. Agostini JV, Inouye SK. Delirium. In:

1164 n engl j med 354;11 march 16, 2006

current concepts

Hazzard WR, Blass JP, Halter JB, Ouslander 28. Romano J, Engel GL. Delirium I: elec- 43. Cole M, McCusker J, Dendukuri N,
JG, Tinetti ME, eds. Principles of geriatric troencephalographic data. Arch Neurol Han L. The prognostic significance of
medicine and gerontology. 5th ed. New Psychiatry 1944;51:356-77. subsyndromal delirium in elderly medical
York: McGraw-Hill, 2003:1503-15. 29. Trzepacz P, van der Mast R. The neu- inpatients. J Am Geriatr Soc 2003;51:754-
12. Inouye SK. Delirium in hospitalized ropathophysiology of delirium. In: Linde- 60.
older patients. Clin Geriatr Med 1998; say J, Rockwood K, Macdonald A, eds. 44. McCusker J, Cole M, Dendukuri N,
14:745-64. Delirium in old age. Oxford, England: Ox- Han L, Belzile E. The course of delirium
13. Pisani MA, McNicoll L, Inouye SK. ford University Press, 2002:51-90. in older medical inpatients: a prospective
Cognitive impairment in the intensive 30. Burns A, Gallagley A, Byrne J. Deliri- study. J Gen Intern Med 2003;18:696-704.
care unit. Clin Chest Med 2003;24:727-37. um. J Neurol Neurosurg Psychiatry 2004; 45. Clarfield AM. The reversible demen-
14. Kiely DK, Bergmann MA, Jones RN, 75:362-7. tias: do they reverse? Ann Intern Med 1988;
Murphy KM, Orav EJ, Marcantonio ER. 31. Han L, McCusker J, Cole M, Abraha- 109:476-86.
Characteristics associated with delirium mowicz M, Primeau F, Elie M. Use of 46. Eikelenboom P, Hoogendijk WJ. Do
persistence among newly admitted post- medications with anticholinergic effect delirium and Alzheimer’s dementia share
acute facility patients. J Gerontol A Biol predicts clinical severity of delirium symp- specific pathogenetic mechanisms? De-
Sci Med Sci 2004;59:344-9. toms in older medical inpatients. Arch ment Geriatr Cogn Disord 1999;10:319-
15. Roche V. Etiology and management of Intern Med 2001;161:1099-105. 24.
delirium. Am J Med Sci 2003;325:20-30. 32. Shigeta H, Yasui A, Nimura Y, et al. 47. Jackson JC, Gordon SM, Hart RP,
16. Breitbart W, Strout D. Delirium in the Postoperative delirium and melatonin lev- Hopkins RO, Ely EW. The association be-
terminally ill. Clin Geriatr Med 2000; els in elderly patients. Am J Surg 2001; tween delirium and cognitive decline:
16:357-72. 182:449-54. a review of the empirical literature. Neu-
17. Casarett D, Inouye SK. Diagnosis and 33. Broadhurst C, Wilson K. Immunology ropsychol Rev 2004;14:87-98.
management of delirium near the end of of delirium: new opportunities for treat- 48. Yokota H, Ogawa S, Kurokawa A,
life. Ann Intern Med 2001;135:32-40. ment and research. Br J Psychiatry 2001; Yamamoto Y. Regional cerebral blood
18. Folstein MF, Bassett SS, Romanosski 179:288-9. flow in delirium patients. Psychiatry Clin
AJ, Nestadt G. The epidemiology of delir- 34. Robertsson B, Blennow K, Bråne G, et Neurosci 2003;57:337-9.
ium in the community: the Eastern Balti- al. Hyperactivity in the hypothalamic- 49. Fick D, Foreman M. Consequences of
more Mental Health Survey. Int Psycho- pituitary-adrenal axis in demented pa- not recognizing delirium superimposed
geriatr 1991;3:169-76. tients with delirium. Int Clin Psychophar- on dementia in hospitalized elderly indi-
19. Rahkonen T, Eloniemi-Sulkava U, macol 2001;16:39-47. viduals. J Gerontol Nurs 2000;26:30-40.
Paanila S, Halonen P, Sivenius J, Sulkava 35. Uno H, Ross T, Else JG, Suleman MA, 50. Rockwood K, Cosway S, Carver D, Jar-
R. Systematic intervention for supporting Sapolsky RM. Hippocampal damage as- rett P, Stadnyk K, Fisk J. The risk of demen-
community care of elderly people after a sociated with prolonged and fatal stress tia and death after delirium. Age Ageing
delirium episode. Int Psychogeriatr 2001; in primates. J Neurosci 1989;9:1705-11. 1999;28:551-6.
13:37-49. 36. Meijer OC, de Kloet ER. Corticoste- 51. McCusker J, Cole M, Dendukuri N,
20. American Psychiatric Association. rone and serotonergic neurotransmission Belzile E, Primeau F. Delirium in older
Practice guideline for the treatment of pa- in the hippocampus: functional implica- medical inpatients and subsequent cogni-
tients with delirium. Am J Psychiatry 1999; tions of central corticosteroid receptor tive and functional status: a prospective
156:Suppl:1-20. diversity. Crit Rev Neurobiol 1998;12:1-20. study. CMAJ 2001;165:575-83.
21. Moran JA, Dorevitch MI. Delirium in 37. Breitbart W, Marotta R, Platt MM, et 52. Baker FM, Wiley C, Kokmen E, Chan-
the hospitalized elderly. Aust J Hosp al. A double-blind trial of haloperidol, dra V, Schoenberg BS. Delirium episodes
Pharm 2001;31:35-40. chlorpromazine, and lorazepam in the during the course of clinically diagnosed
22. Cole MG, Dendukuri N, McCusker J, treatment of delirium in hospitalized Alzheimer’s disease. J Natl Med Assoc
Han L. An empirical study of different di- AIDS patients. Am J Psychiatry 1996; 1999;91:625-30.
agnostic criteria for delirium among el- 153:231-7. 53. Rothschild JM, Bates DW, Leape LL.
derly medical inpatients. J Neuropsychiatry 38. Marcantonio ER, Flacker JM, Wright J, Preventable medical injuries in older pa-
Clin Neurosci 2003;15:200-7. Resnick NM. Reducing delirium after hip tients. Arch Intern Med 2000;160:2717-
23. Laurila JV, Pitkala KH, Strandberg TE, fracture: a randomized trial. J Am Geriatr 28.
Tilvis RS. Delirium among patients with Soc 2001;49:516-22. 54. Gillick MR, Serrell NA, Gillick LS. Ad-
and without dementia: does the diagnosis 39. Inouye SK. Delirium and cognitive de- verse consequences of hospitalization in
according to the DSM-IV differ from the cline: does delirium lead to dementia? In: the elderly. Soc Sci Med 1982;16:1033-8.
previous classifications? Int J Geriatr Psy- Fillit HM, Butler RN, eds. Cognitive de- 55. Williamson JW. Formulating priori-
chiatry 2004;19:271-7. cline: strategies for prevention: proceed- ties for quality assurance activity. JAMA
24. Inouye SK, Foreman MD, Mion LC, ings of a White House Conference on Ag- 1978;239:631-7.
Katz KH, Cooney LM Jr. Nurses’ recogni- ing. London: Greenwich Medical Media, 56. Sloss EM, Solomon DH, Shekelle PG,
tion of delirium and its symptoms: com- 1997:85-107. et al. Selecting target conditions for quality
parison of nurse and researcher ratings. 40. Rockwood K. The occurrence and du- of care improvement in vulnerable older
Arch Intern Med 2001;161:2467-73. ration of symptoms in elderly patients adults. J Am Geriatr Soc 2000;48:363-9.
25. Inouye SK, Charpentier PA. Precipitat- with delirium. J Gerontol Med Sci 1993;48: 57. Institute of Medicine (IOM) Commit-
ing factors for delirium in hospitalized M162-M166. tee on Quality of Health Care in America.
elderly persons: predictive model and in- 41. Levkoff SE, Liptzin B, Evans EA, et al. To err is human: building a safer health
terrelationship with baseline vulnerabili- Progression and resolution of delirium in system. Washington, D.C.: National Acad-
ty. JAMA 1996;275:852-7. elderly patients hospitalized for acute care. emy Press, 2000:26-48.
26. Gleason OC. Delirium. Am Fam Phy- Am J Geriatr Psychiatry 1994;2:230-8. 58. Thomas EJ, Studdert DM, Newhouse
sician 2003;67:1027-34. 42. Marcantonio ER, Flacker JM, Michaels JP, et al. Costs of medical injuries in Utah
27. Rolfson D. The causes of delirium. In: M, Resnick NM. Delirium is independent- and Colorado. Inquiry 1999;36:255-64.
Lindesay J, Rockwood K, Macdonald A, ly associated with poor functional recov- Copyright © 2006 Massachusetts Medical Society.
eds. Delirium in old age. Oxford, England: ery after hip fracture. J Am Geriatr Soc
Oxford University Press, 2002:101-22. 2000;48:618-24.

n engl j med 354;11 march 16, 2006 1165

Supplementary Appendix

This appendix has been provided by the author to give readers additional information about their work.

Supplement to: Inouye SK. Delirium in Older Persons. N Engl J Med 2006;354:1157-65.

Diagnostic and Statistical Manual (DSM-IV) Diagnostic Criteria 1

A. Disturbance of consciousness (i.e., reduced clarity of awareness of the environment) with reduced

ability to focus, sustain, or shift attention.

B. A change in cognition (such as memory deficit, disorientation, language disturbance) or the

development of perceptual disturbance that is not better accounted for by a preexisting,

established, or evolving dementia.

C. The disturbance develops over a short period of time (usually hours to days) and tends to

fluctuate during the course of the day.

D. There is evidence from the history, physical examination, or laboratory findings that the

disturbance is caused by the direct physiological consequences of a general medical condition.

ICD-10 Diagnostic Criteria 2

For a definite diagnosis, symptoms, mild or severe, should be present in each one of the following areas:

a. impairment of consciousness and attention (on a continuum from clouding to coma; reduced

ability to direct, focus, sustain, and shift attention);

b. global disturbance of cognition (perceptual distortions, illusions and hallucinations-- most often

visual; impairment of abstract thinking and comprehension, with or without transient delusions,

but typically with some degree of incoherence; impairment of immediate recall and of recent

memory but with relatively intact remote memory; disorientation for time as well as, in more

severe cases, for place and person)

c. psychomotor disturbances (hypo- or hyperactivity and unpredictable shifts from one to the other;

increased reaction time; increased or decreased flow of speech; enhanced startle reaction);

d. disturbance of the sleep - wake cycle (insomnia or, in severe cases, total sleep loss or reversal of

the sleep - wake cycle; daytime drowsiness; nocturnal worsening of symptoms; disturbing dreams

or nightmares, which may continue as hallucinations after awakening);


e. emotional disturbances, e.g. depression, anxiety or fear, irritability, euphoria, apathy, or

wondering perplexity.

The Confusion Assessment Method (CAM) Diagnostic Algorithm* 3

Feature 1. Acute onset and fluctuating course

This feature is usually obtained from a reliable reporter, such as a family member, caregiver, or

nurse and is shown by positive responses to these questions: Is there evidence of an acute change

in mental status from the patient’s baseline? Did the (abnormal) behavior fluctuate during the

day, that is, tend to come and go, or did it increase and decrease in severity?

Feature 2. Inattention

This feature is shown by a positive response to this question: Did the patient have difficulty

focusing attention, for example, being easily distractible, or have difficulty keeping track of what

was being said?

Feature 3. Disorganized thinking

This feature is shown by a positive response to this question: Was the patient’s thinking

disorganized or incoherent, such as rambling or irrelevant conversation, unclear or illogical flow

of ideas, or unpredictable switching from subject to subject?

Feature 4. Altered level of consciousness

This feature is shown by any answer other than “alert” to this question: Overall, how would you

rate this patient’s level of consciousness (alert [normal], vigilant [hyperalert], lethargic [drowsy,

easily aroused], stupor [difficult to arouse], or coma [unarousable])?

* The ratings for the CAM should be completed following brief cognitive assessment of the patient, for example,

with the Mini-Mental State Examination. The diagnosis of delirium by CAM requires the presence of features 1

and 2 and of either 3 or 4.


Acute onset Disturbance occurs abruptly, usually over hours to days, but onset over

weeks described. Because the patient is often unable to provide an accurate

history, a reliable informant is needed to ascertain the time course of onset.

Fluctuating course Symptoms tend to come and go, or increase and decrease in severity over a

24-hour period. Lucid intervals are characteristic.

Inattention Difficulty focusing, sustaining, and shifting attention. Patients are easily

distracted, have difficulty maintaining conversation or following commands.

Patients perform poorly on simple bedside tests of attention, such as digit

spans and reciting months backwards.

Disorganized thinking Manifested by disorganized or incoherent speech, reflecting abnormalities in

form and content of thinking. Patients demonstrate rambling or irrelevant

conversation, unclear or illogical flow of ideas, or unpredictable switching

between subjects. Patients may be unable to make decisions, solve

problems, or plan and sequence activities. Judgment and insight may be


Altered level of consciousness Clouding of consciousness with reduced clarity of awareness of the

environment. This is typically manifested by lethargy, where the patient is

abnormally drowsy and difficult to keep aroused. The patient may also be

hyperalert, vigilant, and unable to filter out environmental stimuli.


Cognitive deficits Global or multiple deficits in cognition are typical. Disorientation to time

and place are common; disorientation to person is present in more severe

cases. Memory deficits for immediate and short-term memory are typical,

with preserved remote memory. Language functions are often abnormal,

with tangential or slurred speech, paraphasias, and word-finding difficulty

progressing to aphasia in severe cases.

Perceptual disturbances Significant perceptual disturbances occur in up to 30% of delirious patients.

Initial changes include distortions with abnormalities in the perception of

size of objects (macropsia, micropsia), shape, position, movement, or

derealization. Illusions are frank misinterpretations of environmental stimuli

(such as mistaking a pile of laundry for a person, or hallway sounds for

gunshots). Both visual and auditory hallucinations occur with delirium;

tactile hallucinations are less common. The perceptual disturbances may be

frightening to patients and may result in significant behavioral disturbances.

Psychomotor disturbances Delirium demonstrates three psychomotor variants: hyperactive, hypoactive,

and mixed. In the hyperactive form of delirium, the patient is agitated (i.e.,

restless, picking at clothing, tapping fingers, making frequent changes of

position), often vigilant and hallucinating. In the hypoactive form of

delirium, the patient has a markedly decreased level of motor activity (i.e.,

sluggishness, staring into space), often lethargic or passive. Patients may

alternate between the hyperactive and hypoactive subtypes.


Altered sleep-wake cycle Sleep cycle disturbances are characteristic. Daytime drowsiness, nighttime

insomnia, fragmented sleep, or complete sleep-cycle reversal are typical

with delirium.

Emotional disturbances Emotional disturbances are common, manifested by intermittent and labile

symptoms of fear, anxiety, depression, irritability, apathy, anger, or

euphoria. Paranoid delusions may occur.



• Diagnosis of delirium: What is the sensitivity, specificity, clinical yield and cost-effectiveness of
various evaluation approaches to delirium? What is the optimal diagnostic approach?
• Long-term follow-up studies of delirious patients: Does delirium itself lead to permanent
cognitive impairment? How often does delirium lead to mild cognitive impairment or dementia?
• Delirium superimposed on dementia: Does delirium alter the trajectory of cognitive decline in
patients with dementia?
• Cognitive reserve capacity/recovery from delirium: Are there factors that assist in cognitive
recovery after delirium, such as educational level, exercise, health habits, medications (e.g.,
cholinesterase inhibitors)?
• Genetic factors: Are there genetic determinants of the risk of development of delirium, or lack of
recovery from delirium? What is the role of APOE-ε4?
• Identification of high-risk populations: Does delirium exert its maximal detrimental effects only
in certain high-risk subgroups? Or high-risk settings (e.g., post-operative, intensive care)?
• Impact of delirium: What are the economic and societal costs associated with delirium and
delirium superimposed on dementia?

• Neuroimaging, neuropsychological, and neuropathological studies: Does delirium lead to
permanent neurological sequelae?
• Structural and functional imaging methods: Do preexisting abnormalities (e.g., white matter
hyperintensities, volume losses) or functional changes predict the development of delirium, as
well as longer-lasting cognitive decline?
• Amyloid imaging: Does degree of amyloid pathology correlate with the risk of delirium or the
likelihood of recovery from delirium?
• Animal models: Are there pathologic changes after induced delirium (e.g., general anesthesia) in
normal and dementia models?
• Laboratory, electrophysiologic, or neuroimaging markers: Can we identify markers that will
assist in diagnosis of delirium? Can we identify markers for delirium that is likely to lead to
chronic cognitive impairment or dementia?

• Behavioral manifestations: What is the underlying pathophysiology of behavioral manifestations
of delirium (e.g., hyperactive vs. hypoactive form)? Do genetic factors or neuroreceptor subtypes
play a role?

Prevention and Treatment

• Prevention of delirium: Does prevention of delirium with proven non-pharmacologic
intervention strategies mitigate or abate future cognitive decline?
• Treatment of delirium: Randomized trials to evaluate treatment strategies for delirium in
cognitively normal and in dementia patients (e.g., nonpharmacologic approaches, cholinesterase
inhibitors, dopamine antagonists, anti-inflammatory agents, anti-platelet agents, lipid-lowering
agents, antioxidant or neuroprotective drugs). Will early recognition and treatment of delirium
mitigate or abate subsequent cognitive decline?


1. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th ed.
Washington DC: American Psychiatric Association, 1994.
2. World Health Organization. The ICD-10 classification of mental and behavioural disorders.
Diagnostic criteria for research. Geneva: WHO, 1992;57-9.
3. Inouye SK, Van Dyck CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI. Clarifying confusion: The
Confusion Assessment Method. A new method for detection of delirium. Ann Intern Med. 1990;
113: 941-948.