In theClinic

In the Clinic

Delirium
Screening and Prevention page ITC6-2

Diagnosis page ITC6-5

Treatment page ITC6-8

Practice Improvement page ITC6-13

Tool Kit page ITC6-14

Patient Information page ITC6-15

CME Questions page ITC6-16

Section Editors The content of In the Clinic is drawn from the clinical information and education
Deborah Cotton, MD, MPH resources of the American College of Physicians (ACP), including PIER (Physicians’
Darren Taichman, MD, PhD Information and Education Resource) and MKSAP (Medical Knowledge and Self-
Sankey Williams, MD Assessment Program). Annals of Internal Medicine editors develop In the Clinic
from these primary sources in collaboration with the ACP’s Medical Education and
Physician Writer Publishing divisions and with the assistance of science writers and physician writ-
Edward R. Marcantonio, MD SM ers. Editorial consultants from PIER and MKSAP provide expert review of the con-
tent. Readers who are interested in these primary resources for more detail can
consult http://pier.acponline.org, http://www.acponline.org/products_services/
mksap/15/?pr31, and other resources referenced in each issue of In the Clinic.

CME Objective: To review current evidence for the prevention, diagnosis, and
treatment of delirium.

The information contained herein should never be used as a substitute for clinical
judgment.

© 2011 American College of Physicians

 elirium is an acute state of confusion marked by sudden onset, fluctuat-
1. Inouye SK. Delirium in
older persons. N Engl
J Med.
2006;354:1157-65.
D ing course, inattention, and at times an abnormal level of consciousness
(1). Delirium is extremely common but can be challenging to diagnose.
Its manifestations range from acute agitation, which accounts for less than 25%
[PMID: 16540616]
2. Marcantonio ER,
of all cases (1), to the much more common but less frequently recognized hy-
Goldman L, Man- poactive, or quiet, variant. Delirium, which is an acute change in mental status,
gione CM, Ludwig LE,
Muraca B, Haslauer must be differentiated from dementia, which is usually characterized by a slower
CM, et al. A clinical
prediction rule for progression. However, delirium and dementia frequently coexist. Approximately
delirium after elective
noncardiac surgery.
one third of patients aged 70 years or older admitted to the general medical
JAMA. 1994;271:134- service of an acute care hospital experience delirium: One half of these are
9. [PMID: 8264068]
3. Marcantonio ER, delirious on admission to the hospital; the other half develops delirium in the
Flacker JM, Wright RJ,
Resnick NM. Reduc-
hospital (1). Postoperative delirium rates among seniors range from 15% to 25%
ing delirium after hip after elective surgery (2), such as total joint replacement, to over 50% after
fracture: a random-
ized trial. J Am Geriatr high-risk procedures, such as hip fracture repair and cardiac surgery (3, 4).
Soc. 2001;49:516-22.
[PMID: 11380742] Among patients of any age admitted to intensive care units (ICUs), the preva-
4. Rudolph JL, Jones RN,
Levkoff SE, Rockett C,
lence of delirium may exceed 75% (5), and the cumulative incidence of delirium
Inouye SK, Sellke FW, at the end of life is reported to be as high as 85% (6).
et al. Derivation and
validation of a preop-
erative prediction Mounting evidence indicates that delirium is strongly and independently as-
rule for delirium after
cardiac surgery. Cir-
sociated with poor patient outcomes. In the hospital, delirium has been asso-
culation. ciated with a 10-fold increased risk for death and a 3- to 5-fold increased
2009;119:229-36.
[PMID: 19118253] risk for nosocomial complications, prolonged length of stay, and greater need
5. Ely EW, Shintani A, for nursing home placement after discharge (1). Even after discharge, a pa-
Truman B, Speroff T,
Gordon SM, Harrell FE tient who had delirium in the hospital is more likely to have poor functional
Jr, et al. Delirium as a
predictor of mortality
and cognitive recovery and is at increased risk for death for up to 2 years.
in mechanically ven-
tilated patients in the
intensive care unit. A recently published meta-analysis that may play an important role in its associa-
JAMA. 2004;291:1753- included almost 3000 patients followed tion with poor long-term outcomes (8). In
62. [PMID: 15082703]
6. Casarett DJ, Inouye for a mean 22.7 months showed that this review, delirium persistence rates at
SK; American College delirium was independently associated discharge, 1, 3, and 6 months were 44.7%,
of Physicians-Ameri-
can Society of Inter- with an increased risk for death (odds ra- 32.8%, 25.6%, and 21%, respectively, and
nal Medicine End-of- tio [OR], 2.0 [95% CI, 1.5 to 2.5]), institu- outcomes (mortality, nursing home
Life Care Consensus
Panel. Diagnosis and
tionalization (OR, 2.4, [CI, 1.8 to 3.3]), and placement, function, cognition) of pa-
management of dementia (OR, 12.5, [CI, 11.9 to 84.2]) (7). tients with persistent delirium were con-
delirium near the end
of life. Ann Intern
Further, another recent systematic review sistently worse than those of patients who
Med. 2001;135:32-40. found that the persistence of delirium recovered.
[PMID: 11434730].
7. Witlox J, Eurelings LS,
de Jonghe JF, Kalis-
Screening and
vaart KJ, Eikelen-
boom P, van Gool
WA. Delirium in elder-
Prevention Which patients are at risk for from precipitating factors—the
ly patients and the
risk of postdischarge delirium and what are the former are chronic factors that in-
mortality, institution- common precipitating factors? crease a patient’s vulnerability to
alization, and demen-
tia: a meta-analysis. Delirium is best understood as a delirium, whereas the latter are acute
JAMA. 2010;304:443-
51. [PMID: 20664045]
multifactorial geriatrics syndrome, conditions or events that initiate
8. Cole MG, Ciampi A, although patients of any age can be delirium. Several large epidemiologic
Belzile E, Zhong L.
Persistent delirium in affected. Geriatrics syndromes, in- studies and systematic reviews have
older hospital pa-
tients: a systematic
cluding delirium, falls, incontinence, defined predisposing and precipitat-
review of frequency and failure to thrive, share several ing factors for delirium. Based on
and prognosis. Age
Ageing. 2009;38:19- characteristics (9). Most notably, risk this model, an individual’s risk for
26. [PMID: 19017678] factors are multifactorial and often delirium is defined by the sum of
9. Inouye SK, Studenski
S, Tinetti ME, Kuchel lie outside of the presenting organ predisposing and precipitating fac-
GA. Geriatric syn-
dromes: clinical, re- system, which in the case of delirium tors; the more predisposing factors
search, and policy is the central nervous system. present, the fewer precipitating
implications of a core
geriatric concept. J events required to cause delirium
Am Geriatr Soc. A common risk factor model for (10) (Box: Common Risk Factors for
2007;55:780-91.
[PMID: 17493201] delirium distinguishes predisposing Delirium). For example, a young,

© 2011 American College of Physicians ITC6-2 In the Clinic Annals of Internal Medicine 7 June 2011
 Screening methods are similar to
Common Risk Factors for those for diagnosis. The briefest
Delirium
screening methods use the Confu- 10. Inouye SK, Charpen-
Predisposing sion Assessment Method (CAM) tier PA. Precipitating
• Advanced age factors for delirium
• Preexisting dementia
diagnostic algorithm (13) (Box: in hospitalized elder-
ly persons. Predictive
• History of stroke Confusion Assessment Method Di- model and interrela-
• Parkinson disease agnostic Algorithm), which exam- tionship with base-
• Multiple comorbid conditions line vulnerability.
ines 4 key features of delirium: acute JAMA. 1996;275:852-
• Impaired vision
• Impaired hearing
change in mental status and fluctuat- 7. [PMID: 8596223]
11. Kales HC, Kamholz
• Functional impairment ing course, inattention, disorganized BA, Visnic SG, Blow
FC. Recorded deliri-
• Male sex thinking, and abnormal level of con- um in a national
• History of alcohol abuse sciousness. Diagnosis of delirium by sample of elderly in-
patients: potential
Precipitating CAM requires the presence of fea- implications for
• New acute medical problem recognition. J Geriatr
• Exacerbation of chronic medical tures 1 and 2 and either 3 or 4. Al- Psychiatry Neurol.
problem though CAM is considered to be an 2003;16:32-8.
[PMID: 12641371]
• Surgery/anesthesia accurate approach for diagnosis of 12. Cole MG, McCusker
• New psychoactive medication delirium, sensitivity varies depending J, Bellavance F,
• Acute stroke Primeau FJ, Bailey RF,
• Pain on the assessment methods used (14, Bonnycastle MJ, et
al. Systematic detec-
• Environmental change 15). The literature suggests that tion and multidisci-
• Urine retention/fecal impaction completing CAM by using routine plinary care of deliri-
• Electrolyte disturbances um in older medical
observations from clinical care is inpatients: a ran-
• Dehydration
• Sepsis probably not sufficient and that a domized trial. CMAJ.
2002;167:753-9.
standardized mental status assess- [PMID: 12389836]
13. Inouye SK, van Dyck
ment should be done to improve its CH, Alessi CA, Balkin
sensitivity (16). S, Siegal AP, Horwitz
RI. Clarifying confu-
sion: the confusion
otherwise-healthy person may be- The CAM-ICU is a specific appli- assessment method.
come delirious after being subjected cation of the CAM algorithm that
A new method for
detection of deliri-
to severe sepsis, respiratory failure, uses nonverbal responses from the um. Ann Intern Med.
and mechanical ventilation in the patient to assess attention, thinking,
1990;113:941-8.
[PMID: 2240918]
ICU. In contrast, a frail older adult and level of consciousness (17). The 14. Inouye SK, Foreman
MD, Mion LC, Katz
with cognitive impairment may be- CAM-ICU is valid, reliable, and can KH, Cooney LM Jr.
come delirious after taking a low be completed in a few minutes. Pa-
Nurses’ recognition
of delirium and its
dose of acetaminophen with diphen- tients in the ICU are at such high symptoms: compari-
son of nurse and re-
hydramine for sleep. searcher ratings.
Arch Intern Med.
Should clinicians screen 2001;161:2467-73.
[PMID: 11700159]
hospitalized patients for delirium, 15. Lemiengre J, Nelis T,
and if so, how? The Confusion Assessment Method Diagnostic Joosten E, Braes T,
Foreman M, Gast-
Algorithm
Delirium is a common, morbid con- mans C, et al. Detec-
tion of delirium by
dition; however, 50% to 80% of cases Feature 1. Acute change in mental status and bedside nurses us-
go unrecognized and undocumented fluctuating course ing the confusion
• Is there evidence of an acute change in cognition assessment method.
by the treating clinical team (11). from baseline?
J Am Geriatr Soc.
2006;54:685-9.
Trials that have assessed the effec- • Does the abnormal behavior fluctuate during the [PMID: 16686883]
tiveness of systematic programs to day? 16. Wong CL, Holroyd-
Leduc J, Simel DL,
improve case findings and treatment Feature 2. Inattention Straus SE. Does this
of delirium have shown significantly • Does the patient have difficulty focusing attention patient have deliri-
(e.g., easily distracted, has difficulty keeping track of um?: value of bed-
improved detection rates and modest what is being said)?
side instruments.
JAMA. 2010;304:779-
improvements in outcomes (12). Feature 3. Disorganized thinking 86. [PMID: 20716741]
Therefore, it seems prudent to screen 17. Ely EW, Inouye SK,
• Does the patient have rambling or irrelevant conver- Bernard GR, Gordon
hospitalized patients who are at risk sations, unclear or illogical flow of ideas, or unpre- S, Francis J, May L, et
for delirium (either due to predis- dictable switching from subject to subject? al. Delirium in me-
chanically ventilated
posing factors or the acute situation), Feature 4. Abnormal level of consciousness patients: validity and
• Is the patient anything besides reliability of the con-
including those with preexisting cog- alert—hyperalert, lethargic, stuporous, or comatose? fusion assessment
method for the in-
nitive impairment or multiple co- tensive care unit
The diagnosis of delirium requires features 1 and 2 and
morbid conditions or those admitted either 3 or 4.
(CAM-ICU). JAMA.
2001;286:2703-10.
to the ICU. [PMID: 11730446]

7 June 2011 Annals of Internal Medicine In the Clinic ITC6-3 © 2011 American College of Physicians
 risk for delirium that they should be undergoing hip fracture repair. In a ran-
screened daily, if not more frequent- domized, controlled trial, the proactive
ly. There are several options for as- geriatrics consultation group achieved a
sessing mental status before com- 36% reduction [CI, 10% to 62%], NNT = 6)
in the incidence of delirium (3). Consulta-
pleting the CAM diagnostic
tion began before surgery and continued
algorithm in non-ICU patients. The throughout the duration of hospitaliza-
Mini-Mental State Examination tion. Daily recommendations were based
(which requires a license for use) on a structured protocol that addressed
(18) takes up to 15 minutes and pro- 10 key risk factors for delirium, such as
vides a limited assessment of atten- limitation of psychoactive medications
tion and level of consciousness. The and timely removal of indwelling urinary
CAM-ICU can be used, but sensi- catheters. Co-management of hip frac-
18. O’Keeffe ST, Mulker- tivity may be lower than that of ture patients and other vulnerable surgi-
rin EC, Nayeem K,
CAM in verbal patients (19). Alter- cal populations by surgeons and geriatri-
Varughese M, Pillay I.
Use of serial Mini- natively, screeners can assess level of cians and/or internists is increasing in
Mental State Exami- popularity.
nations to diagnose consciousness by using a standard-
and monitor deliri-
um in elderly hospi-
ized measure, such as the RASS Recently, several preventive ap-
tal patients. J Am (Richmond Agitation Sedation proaches have not reduced the in-
Geriatr Soc.
2005;53:867-70. Scale) (20), and attention by using 1 cidence of delirium but have re-
[PMID: 15877566]
19. Fong TG, Jones RN,
or more additional items (see Box in duced severity and/or duration.
Rudolph JL, Yang Diagnosis section). These trials have tested new mod-
FM, Tommet D,
Habtemariam D, et els of care that involve reorganiza-
al. Development and A recent review that summarized multiple
Validation of a Brief tion of nursing care and the hos-
assessment methods concluded that CAM
Cognitive Assess- pital environment to focus on
ment Tool: The has the best available supportive evidence
Sweet 16. Arch In- patient-centered care and the re-
tern Med. 2010. as a bedside delirium instrument (16).
[PMID: 21059967]
duction of factors that commonly
20. McNicoll L, Pisani
Are there effective strategies for precipitate delirium (23, 24). A
MA, Ely EW, Gifford
D, Inouye SK. Detec- prevention? very different approach involved
tion of delirium in
the intensive care Among all the interventions for delirium, administration of low-dose
unit: comparison of the strongest evidence supports the effec- haloperidol (0.5 mg 3 times daily
confusion assess-
ment method for tiveness of prevention. The Yale Delirium for 3 days) in high-risk hip sur-
the intensive care Prevention Trial tested the effectiveness of gery. This strategy also reduced
unit with confusion
assessment method the Hospital Elder Life Program (HELP), the severity and duration (but not
ratings. J Am Geriatr which targeted 6 risk factors for delirium: incidence) of postoperative deliri-
Soc. 2005;53:495-
500. cognitive impairment, sleep deprivation, um (25). It is important to note
[PMID: 15743296] immobility, visual impairment, hearing that all of these models were pre-
21. McDowell JA, Mion
LC, Lydon TJ, Inouye impairment, and dehydration. Risk factors ventive—that is, they were started
SK. A nonpharmaco- were assessed at admission, and patients
logic sleep protocol before the onset of delirium.
for hospitalized old- with one or more of these risk factors re-
er patients. J Am ceived targeted interventions to address One methodological issue may lim-
Geriatr Soc.
1998;46:700-5. them. The interventions were delivered by
it the clinical utility of prevention
[PMID: 9625184] specially trained teams of volunteers under
22. Inouye SK, Bogardus
the guidance of specialized staff. Among trials that have reduced delirium
ST Jr, Charpentier PA,
Leo-Summers L, these, the most creative and successful was severity alone. Delirium severity
Acampora D, Holford
TR, et al. A multi- a nonpharmacologic sleep protocol that scales often disproportionately
component inter- involved trained volunteers offering pa- weight “positive” symptoms, such as
vention to prevent
delirium in hospital- tients warm milk, back rubs, and soothing agitation, perceptual disturbances,
ized older patients. music at bedtime; this intervention sub- or delusions (26, 27). Reduction of
N Engl J Med.
1999;340:669-76. stantially reduced the use of sedative– these symptoms through interven-
[PMID: 10053175] hypnotic medication (21). Delirium in the tions that use sedating or antipsy-
23. Milisen K, Foreman
MD, Abraham IL, De intervention group was significantly re- chotic medication may result in an
Geest S, Godderis J, duced (OR, 0.60, [CI, 0.39-0.92], number apparent reduction in delirium
Vandermeulen E, et
al. A nurse-led inter- needed to treat [NNT] = 19) (22). This ap- severity. However, studies suggest
disciplinary interven- proach has been implemented widely.
tion program for that patients with hypoactive deliri-
delirium in elderly
hip-fracture patients. Another approach with proven benefit for um may have equal or worse out-
J Am Geriatr Soc. prevention of delirium is proactive geri- comes than those with hyperactive
2001;49:523-32.
[PMID: 11380743] atrics consultation in elderly patients delirium (28).

© 2011 American College of Physicians ITC6-4 In the Clinic Annals of Internal Medicine 7 June 2011
 Screening... The most effective, proven approach to prevent delirium is proactive,
multifactorial, nonpharmacologic interventions, such as the HELP program or
proactive geriatrics consultation for high-risk surgical patients. Risk factors for
delirium should be assessed on admission to the hospital. High-risk patients, in-
cluding those in high-risk settings, such as the ICU, should also be screened for
delirium on admission and at least daily thereafter using proven methods, such as
CAM.

CLINICAL BOTTOM LINE

Diagnosis
When should clinicians consider a The key aspect of the physical exam-
diagnosis of delirium? ination is evaluation of mental status,
Delirium should be considered in and the most important aspect of
any confused hospitalized patient this evaluation is determining the
24. Lundström M, Ed-
and in high-risk patients with level of consciousness and attention lund A, Karlsson S,
Brännström B, Bucht
confusion in any setting. When in (Box: Commonly Used Tests of At- G, Gustafson Y. A
doubt, it is always better to rule tention). Any abnormal level of con- multifactorial inter-
vention program re-
out delirium first than to attribute sciousness, as well as the presence of duces the duration
confusion to an underlying chron- inattention, is supportive of delirium. of delirium, length
of hospitalization,
ic disorder, such as dementia, and For patients with a normal level of and mortality in
delirious patients. J
fail to recognize the presence of consciousness, assessing their think- Am Geriatr Soc.
delirium. ing, specifically whether there is evi- 2005;53:622-8.
[PMID: 15817008]
dence of rambling and incoherent 25. Kalisvaart KJ, de
What elements of the history and speech, is also important. Jonghe JF, Bogaards
MJ, Vreeswijk R, Eg-
physical examination indicate berts TC, Burger BJ,
delirium? et al. Haloperidol
Once the history is obtained from a prophylaxis for eld-
The diagnosis of delirium is based family member and/or caregiver erly hip-surgery pa-
tients at risk for delir-
entirely on the history and physical and the mental status examination ium: a randomized
examination. No laboratory tests, is completed, the diagnosis of delir- placebo-controlled
study. J Am Geriatr
imaging studies, or other tests are ium can be determined using the Soc. 2005;53:1658-
66. [PMID: 16181163]
more accurate than clinical assess- CAM Diagnostic Algorithm (see 26. Breitbart W, Rosen-
ment (1). The history and physical the Box) (13). feld B, Roth A, Smith
MJ, Cohen K, Passik
examination have 2 roles in the S. The Memorial
evaluation of delirium: confirma- Another important element of the Delirium Assessment
Scale. J Pain Symp-
tion of the diagnosis, and identifi- history and physical examination is tom Manage.
cation of potential causes and evaluating for underlying causes. 1997;13:128-37.
[PMID: 9114631]
contributors. These include a medication history, 27. Trzepacz PT, Mittal D,
Torres R, Kanary K,
vital signs, and general medical Norton J, Jimerson
Obtaining the history for delirium examination. Table 1 summarizes N. Validation of the
Delirium Rating
differs from that of other medical the key history and physical exami- Scale-revised-98:
conditions in that it is primarily nation components for both the di- comparison with the
delirium rating scale
obtained from caregivers or family agnosis and evaluation of delirium. and the cognitive
test for delirium. J
members rather than the patient. Neuropsychiatry Clin
One key element is determining Neurosci.
2001;13:229-42.
the timeline of the mental status [PMID: 11449030]
changes—acute (sudden) onset is Commonly Used Tests of Attention 28. Kiely DK, Jones RN,
Bergmann MA, Mar-
most consistent with delirium. • Digit span—up to 5 forwards and 4 backwards cantonio ER. Associ-
• Days of the week, months of the year backwards ation between psy-
Whether there have been fluctua- • Continuous performance task—raise hand when chomotor activity
tions in mental status, such as the hears a certain letter in a list delirium subtypes
and mortality
patient seeming normal at some • Attention screening examination—show pictures; among newly admit-
times and very confused at others, ask patient to remember and recall ted post-acute facili-
• Recite a list of serial 7s or 3s ty patients. J Geron-
is also an essential element of the • Spell W-O-R-L-D backwards
tol A Biol Sci Med
Sci. 2007;62:174-9.
history. [PMID: 17339642]

7 June 2011 Annals of Internal Medicine In the Clinic ITC6-5 © 2011 American College of Physicians
 What is the role of laboratory presenting sign of a stroke (30), so
testing, brain imaging, and if the risk factors, history, and
electroencephalography in the physical examination are suggestive,
diagnosis and evaluation of cerebral imaging may be warranted.
delirium? The Box: Laboratory, Imaging, and
Laboratory testing, brain imaging, Other Studies summarizes the use
and electroencephalography (EEG) of laboratory, imaging, and EEG
do not substitute for history and studies in delirium.
physical examination in the diagno-
sis of delirium. However, these What other disorders should
studies can be useful to identify clinicians consider in patients with
possible causes of delirium and suspected delirium?
correctable contributing factors, Major differential diagnoses of delir-
particularly when they are carefully ium are dementia; depression; other
selected on the basis of history and acute psychiatric syndromes; and
physical examination. When or- subsyndromal delirium, also known
dered as an unselected “delirium as the partial syndrome of delirium
battery,” the yield of these tests and (1). In many cases, it is not truly a
procedures is apt to be low. In par- “differential” diagnosis, since these
ticular, cerebral imaging and EEG syndromes can coexist and indeed
are usually not helpful in the evalu- are risk factors for one another. In-
ation of delirium, unless there is stead, it should be considered a series
strong evidence of an intracranial of independent questions: Does this
cause based on the history (such as patient have delirium? . . .
change in mental status after a dementia? . . . depression?
blow to the head) or if focal neuro-
logic signs or seizure activity is de- The most common diagnostic is-
29. Hirano LA, Bogardus tected on physical examination sue is whether a newly presenting
ST Jr, Saluja S, Leo-
Summers L, Inouye (29). Notably, delirium can be a confused patient has dementia,
SK. Clinical yield of
computed tomogra-
phy brain scans in
older general med-
ical patients. J Am Table 1. Key History and Physical Examination Elements for Delirium
Geriatr Soc.
2006;54:587-92. Variable Notes
[PMID: 16686867] History
30. Sheng AZ, Shen Q,
Cordato D, Zhang Time course of the mental status changes Relatively abrupt onset, fluctuating course suggestive of delirium;
YY, Yin Chan DK. usually obtained from family member or caregiver, not from the
Delirium within
three days of stroke
patient
in a cohort of elderly Association of mental status changes with Obtained from review of the medical record or from a family
patients. J Am Geri- other “events,” including medication changes member or caregiver
atr Soc.
2006;54:1192-8. and development of physical symptoms
[PMID: 16913984] Medication history, including over-the-counter A “brown bag” test, in which all medicines, including over the
31. Fick DM, Agostini JV,
Inouye SK. Delirium
medications. Key medication classes on which counter, are brought in for review; may be helpful in the outpatient
superimposed on to focus: sedative-hypnotics, barbiturates, setting
dementia: a system- alcohol, antidepressants, anticholinergics,
atic review. J Am opioid analgesics, antipsychotics
Geriatr Soc.
2002;50:1723-32. Sensory deprivation Absence of glasses or hearing aids normally worn by the patient
[PMID: 12366629] Pain assessment Delirium has been associated with severe pain as a contributing
32. Boland RJ, Diaz S,
Lamdan RM, Ram- factor
chandani D, McCart- Pain may be manifested only by agitation
ney JR. Overdiagno- Physical examination
sis of depression in
the general hospital. Vital signs, including oxygen saturation Provide clues to underlying causes of delirium
Gen Hosp Psychiatry.
1996;18:28-35. General medical examination, focusing Provides clues to underlying causes of delirium
[PMID: 8666210] on cardiac and pulmonary examination
33. Cole M, McCusker J,
Dendukuri N, Han L. Neurologic examination, including mental Intracranial events rarely present with delirium in elderly patients
The prognostic sig- status examination and examination for
nificance of subsyn- focal findings
dromal delirium in
elderly medical inpa- Cognitive examination, including attention Inttention is the hallmark cognitive deficit in delirium
tients. J Am Geriatr testing (see the Box:Commonly Used Tests
Soc. 2003;51:754-60. of Attention)
[PMID: 12757560]

© 2011 American College of Physicians ITC6-6 In the Clinic Annals of Internal Medicine 7 June 2011
 change in mental status from
Laboratory, Imaging, and Other
baseline is not consistent with de-
Studies to Consider in the
Evaluation of Delirium mentia and suggests delirium. In
• Complete blood count: Infection
addition, a rapidly fluctuating
and severe anemia course (over minutes to hours)
• Serum electrolytes: Electrolyte and an abnormal level of con-
disorders, especially hyper- and sciousness are highly suggestive of
hyponatremia
• BUN, creatinine: Dehydration and delirium. Diffuse Lewy body de-
occult renal failure (rare) mentia, which has among its fea-
• Glucose: Hypoglycemia, severe hy- tures a fluctuating course and an
perglycemia, hyperosmolar state abnormal level of consciousness,
• Albumin, bilirubin, international
normalized ratio: Only if liver fail- may be particularly difficult to
ure and hepatic encephalopathy distinguish from delirium. Even in
are suspected (rare) these patients, however, acute
• Urinalysis, culture: Urinary tract
infection (common in very frail changes should be evaluated and
elders) managed as delirium. Notably, the
• Chest x-ray: Pneumonia or conges- incidence of delirium in hospital-
tive heart failure if there is fever or ized dementia patients exceeds
physical findings
• Electrocardiogram: Myocardial in- 65% (31), so a diagnosis of de-
farction and arrhythmia mentia does not rule out delirium
• Arterial blood gases: Hypercarbia in these patients—indeed, it
in chronic obstructive pulmonary
disease makes it more likely.
• Drug levels: Delirium can occur
with “normal” serum levels of some Depression may also be confused
drugs. with hypoactive delirium. In 1
• Toxic screen: If ingestion is sus- study, a third of patients undergo-
pected, more common in younger
patients. ing psychiatric consultations for de-
34. Marcantonio E, Ta T,
• Cerebral imaging with CT, MRI: pression in the acute care setting Duthie E, Resnick
High suspicion of stroke or hemor- actually had hypoactive delirium NM. Delirium severi-
rhage based on history and physi- ty and psychomotor
cal examination or if delirium is
(32). Certain acute psychiatric syn- types: their relation-
ship with outcomes
persistent, unexpected, unex- dromes, such as mania or acute after hip fracture re-
plained, or occurs in younger psychosis, have a presentation simi- pair. J Am Geriatr
patients Soc. 2002;50:850-7.
lar to that of hyperactive delirium. [PMID: 12028171]
• Lumbar puncture: High suspicion of
meningitis or subarachnoid based Initially, it is best to evaluate and 35 Leff B, Burton L, Mad-
er SL, Naughton B,
on history and physical examina- manage hyperactive patients as if Burl J, Inouye SK, et
tion or if delirium is persistent, un- al. Hospital at home:
they have delirium rather than at- feasibility and out-
expected, unexplained, or occurs in
younger patients
tributing the presentation to psy- comes of a program
to provide hospital-
• Electroencephalography: Often chiatric disease and missing a seri- level care at home
for acutely ill older
shows diffuse slow wave activity ous underlying medical disorder. patients. Ann Intern
but is rarely helpful in evaluation Med. 2005;143:798-
and treatment of reversible causes Patients who present with some, 808.
[PMID: 16330791]
BUN = blood urea nitrogen; CT = but not all, diagnostic features of 36. Agostini JV, Leo-
Summers LS, Inouye
computed tomography; INR = delirium have so-called subsyndro- SK. Cognitive and
international normalized ratio; MRI = mal delirium. There is a growing other adverse effects
magnetic resonance imaging. of diphenhydramine
body of evidence that these patients use in hospitalized
older patients. Arch
may have a spectrum of outcomes Intern Med.
similar to patients who meet the 2001;161:2091-7.
[PMID: 11570937]
full diagnostic criteria for delirium 37. Han L, McCusker J,
delirium, or both. To make this Cole M, Abra-
and should be evaluated and man- hamowicz M,
determination, the physician must
aged similarly (33, 34). Primeau F, Elie M.
know the patient’s baseline status. Use of medications
with anticholinergic
In the absence of documentation When should subspecialty effect predicts clini-
cal severity of deliri-
of a baseline evaluation, informa- consultation be considered for um symptoms in
tion from family members, care- patients with delirium? older medical inpa-
tients. Arch Intern
givers, or others who know the The primary care physician usual- Med. 2001;161:1099-
105.
patient is essential. An acute ly is best suited to guide the [PMID: 11322844]

7 June 2011 Annals of Internal Medicine In the Clinic ITC6-7 © 2011 American College of Physicians
 diagnosis and evaluation of deliri- intensive care may be helpful with
um, because knowledge of the pa- the differential diagnosis of deliri-
tient’s baseline mental status is um and to evaluate contributing
key in the diagnosis and because factors. These consultants may
the spectrum of contributing also be helpful in guiding delirium
38. Marcantonio ER,
Juarez G, Goldman L, causes is broad and rarely involves treatment. No data have currently
Mangione CM, Lud- the central nervous system. How- evaluated whether particular spe-
wig LE, Lind L, et al.
The relationship of ever, given the rapid pace of pri- cialists are superior to others in
postoperative deliri-
um with psychoac- mary care practice, a consultant delirium evaluation and manage-
tive medications. may have more time to review the ment. Therefore, decisions about
JAMA.
1994;272:1518-22. data required to diagnose and which discipline to consult are
[PMID: 7966844]
39. Rudolph JL, Salow
evaluate delirium. Therefore, if based primarily on patient charac-
MJ, Angelini MC, needed, consultation with experts teristics (geriatric medicine for
McGlinchey RE. The
anticholinergic risk in geriatric medicine, psychiatry, very old patients) or setting
scale and anticholin-
ergic adverse effects
neurology, or medical/surgical (intensivists for ICU patients).
in older persons.
Arch Intern Med.
2008;168:508-13.
[PMID: 18332297]
40. Riker RR, Shehabi Y, Diagnosis... For hospitalized patients with altered cognition, assess for delirium
Bokesch PM, Ceraso first, followed by subsyndromal delirium, depression and other acute psychiatric
D, Wisemandle W,
Koura F, et al; SED-
syndromes, and then dementia. This order is based both on the reversibility and
COM (Safety and Ef- treatability of each disorder, as well as the acute consequences of missing the
ficacy of diagnosis.
Dexmedetomidine
Compared With Mi-
dazolam) Study
Group. Dexmedeto-
midine vs midazo- CLINICAL BOTTOM LINE
lam for sedation of
critically ill patients:
a randomized trial.
JAMA. 2009;301:489-
99. [PMID: 19188334]
41. Liptzin B, Laki A,
Garb JL, Fingeroth R,
Treatment
Krushell R. Donepezil When should clinicians consider management may be particularly
in the prevention
and treatment of hospitalization for suspected appropriate for nursing home resi-
post-surgical deliri-
um. Am J Geriatr
delirium? dents, for whom diagnostic, thera-
Psychiatry. The decision to hospitalize a pa- peutic, and monitoring resources
2005;13:1100-6.
[PMID: 16319303] tient with suspected delirium re- are greater than outpatients and
42. van Eijk MM, Roes
KC, Honing ML,
quires consideration of multiple for whom hospitalization may be
Kuiper MA, Karakus factors, including timeliness of the particularly traumatic.
A, van der Jagt M, et
al. Effect of rivastig- diagnostic evaluation, clinical sta-
mine as an adjunct bility, and social support. Not all However, hospitalization may be
to usual care with
haloperidol on dura- patients with delirium require necessary when delirium is associ-
tion of delirium and
mortality in critically hospitalization, and in fact hospi- ated with a destabilizing medical
ill patients: a multi- talization may exacerbate the con- illness, such as sepsis or myocardial
centre, double-blind,
placebo-controlled dition. Outpatient management infarction, or because of inadequate
randomised trial.
Lancet.
may be appropriate if the diagnos- support at home. In making this
2010;376:1829-37. tic workup can be done in a time- decision, it is important to consider
[PMID: 21056464]
43. Bergmann MA, Mur- ly and thorough fashion; patient the high risk for nosocomial com-
phy KM, Kiely DK, safety can be assured; and treat- plications and that disorientation
Jones RN, Marcanto-
nio ER. A model for ment of the conditions causing caused by a new, unfamiliar envi-
management of
delirious postacute delirium is straightforward, such ronment may exacerbate the deliri-
care patients. J Am as reversal of drug side effects or um. A study of carefully selected
Geriatr Soc.
2005;53:1817-25. treatment of simple infections. It patients with acute conditions
[PMID: 16181185]
44. Flaherty JH, Tariq SH, is also important that such outpa- managed in a “home hospital” had
Raghavan S, Bakshi tients have a dedicated family a lower rate of delirium than simi-
S, Moinuddin A,
Morley JE. A model member who can alert the primary lar patients managed in a tradi-
for managing deliri-
ous older inpatients.
care physician promptly if the pa- tional hospital setting (35). How-
J Am Geriatr Soc. tient’s status does not improve or ever, because adequate home
2003;51:1031-5.
[PMID: 12834527] acutely worsens. Nonhospital clinical and social supports are

© 2011 American College of Physicians ITC6-8 In the Clinic Annals of Internal Medicine 7 June 2011
often not available, many patients delirium, recent trials of cholinesterase in-
with delirium are hospitalized. hibitors have not shown favorable results Key Steps in the Supportive Care of
(41, 42). Delirious Patients
What nonpharmacologic measures • Minimize indwelling catheters and
are useful in treatment? In addition to searching for and ad- other “tethers,” such as intravenous
dressing contributing factors, man- lines, electrocardiography leads
Nonpharmacologic measures are • Eliminate physical restraints and
the cornerstone of delirium treat- agement of delirious patients re- mobilize the patient as soon as
ment. First and foremost, manage- quires attentive supportive care to possible
meet their needs and to avoid iatro- • Monitor urinary and bowel output;
ment involves identification and avoid urine retention and fecal im-
treatment of underlying disease genic complications (43, 44) (Box: paction, which can contribute to
processes as well as removal and re- Key Steps in Supportive Care). delirium
duction of associated contributing • Address nutritional needs, including
Randomized trials of intervention pro- assistance with meals and possible
factors. Such factors include psy- grams that have adopted some or all of hand-feeding—delirious patients
choactive medications, fluid and the measures described in the Box have may have difficulty attending to
electrolyte abnormalities, severe shown a significantly reduced duration of food and are at risk for acute mal-
pain, hypoxemia, severe anemia, in- nutrition
delirium, length of stay for delirious pa- • Provide adequate sensory input, in-
fections, sensory deprivation, and tients, and a trend toward reduction in cluding use of glasses and hearing
significant immobility. Particularly hospital mortality (45, 46). aids, provision of clocks, calendars,
in elderly patients, it may not be and adequate lighting
When nonpharmacologic • Provide frequent orientation and
possible to identify a single cause structured interpersonal contact to
for delirium. Because there is a cu- interventions are insufficient, facilitate cognitive “reconditioning”
mulative effect of many vulnerabili- should clinicians consider drug • Adopt healthy sleep–wake cycles,
ty factors at baseline and acute pre- therapy? If so, which drugs are encouraging night sleeping by
useful and what are the adverse reducing environmental stimuli,
cipitating factors, small gains in including minimizing staff noise,
several factors may yield impressive effects of drug therapy? using vibrating (silent) pagers, elim-
results overall (1). The primary therapy for delirium inating waking for vital signs ex-
involves the identification and cept if essential, reducing hospital
Psychoactive medications are among treatment of its causative factors. ward lighting, and turning off tele-
visions and radios.
the most important and reversible Moreover, there are no Food and
contributors to delirium and there- Drug Administration–approved
fore warrant particular attention. drugs to treat delirium. However,
Medication classes associated with drug therapy (off-label) is frequent-
the highest risk include the benzodi- ly used for such symptoms of delir-
azepines, sedative-hypnotics, med- ium as delusions or hallucinations
ications with strong anticholinergic that are frightening to the patient
45. Lundström M, Edlund
properties, opioid analgesics, and or for patient behaviors that are A, Karlsson S,
pro-dopaminergic agents (36–39). In Brännström B, Bucht
dangerous to him- or herself or G, Gustafson Y. A mul-
many cases, good alternatives exist others. Even in these cases, verbal tifactorial interven-
tion program reduces
and should be used if possible (Table comfort and reassurance by the the duration of deliri-
2). Meperidine has the strongest an- hospital staff and provision of a sit- um, length of hospi-
talization, and mor-
ticholinergic properties of any opioid ter or family companion may be tality in delirious
patients. J Am Geriatr
analgesic and also has active metabo- preferable to drug therapy. Pharma- Soc. 2005;53:622-8.
lites that accumulate in the blood- cologic intervention must be used [PMID: 15817008]
46. Pitkälä KH, Laurila JV,
stream; it has been associated with a cautiously, as it may prolong deliri- Strandberg TE, Tilvis
RS. Multicomponent
high risk for delirium, and an alter- um and increase the risk for com- geriatric interven-
native opioid should be used when- plications by converting a hyperac- tion for elderly inpa-
tients with delirium:
ever possible (38). Benzodiazepines tive patient into a stuporous one a randomized, con-
are another commonly used psy- whose risk for a fall or aspiration is trolled trial. J Geron-
tol A Biol Sci Med
choactive medication class with increased. Sci. 2006;61:176-81.
[PMID: 16510862]
strong associations with delirium. 47. Campbell N, Bous-
Recently, several meta-analyses have exam- tani MA, Ayub A, Fox
Recently, the α-adrenergic agonist dex- ined pharmacologic treatment of agitation GC, Munger SL, Ott
C, et al. Pharmaco-
medetomidine has been identified as a less in delirium (47–50), although they con- logical management
delirium-causing alternative to benzodi- tained few studies. One of these studies, of delirium in hospi-
talized adults—-a
azepines for sedation in critically ill patients done in young patients with AIDS, estab- systematic evidence
(40). On the other hand, despite the associa- lished the superiority of haloperidol to ben- review. J Gen Intern
Med. 2009;24:848-
tion of anticholinergic medications with zodiazepines (51). All of the other studies in 53. [PMID: 19424763]

7 June 2011 Annals of Internal Medicine In the Clinic ITC6-9 © 2011 American College of Physicians
 the meta-analyses showed that the new In older patients with mild deliri-
generation of antipsychotics were equiva- um, low doses of haloperidol (0.5
lent to haloperidol. None of the studies used to 1 mg orally or 0.25 to 0.5 mg
a placebo-control group. Based on this limit- parenterally) should be used initial-
ed evidence, high-potency antipsychotics ly, with careful reassessment before
are considered the treatment of choice for
increasing the dose. In more severe
agitation in delirium because of their low
anticholinergic potency and minimal risks delirium, somewhat higher doses
for hypotension or respiratory depression may be used initially (0.5 to 2 mg
(49, 50). parenterally), with additional

Table 2. Drugs That May Cause Delirium and Potential Substitutes

Agent Potential Mechanism Potential Substitute (Reference) Notes
Leading to Delirium
Benzodiazepines, especially Central nervous system Nonpharmacologic sleep management, Associated with delirium in medical,
long-acting, including sedation intermediate agents (lorazepam); surgical, and intensive care unit
diazepam, flurazepam, dexmedetomidine can be used as a patients
chlordiazepoxide sedative in the intensive care unit (40)
Benzodiazepines:
ultra–short-acting, Central nervous system Nonpharmacologic sleep management, Associated with delirium in case
including triazolam, sedation and withdrawal intermediate agents (lorazepam) (21) reports and series
alprazolam, midazolam
Barbiturates Severe withdrawal Avoid inadvertent discontinuation,
syndrome or substitute benzodiazepine
Choral hydrate Central nervous system Nonpharmacologic sleep protocol (21) No better for delirium than
sedation benzodiazepines
Alcohol Central nervous system If history of heavy intake, carefully Alcohol history is imperative
sedation; withdrawal monitor and use benzodiazepines if
syndrome withdrawal symptoms occur
Antidepressants, especially Anticholinergic toxicity Secondary amine tricyclics: nortriptyline, Secondary amines as good as tertiary
the tertiary amine tricyclic desipramine. Selective serotonin reuptake for adjuvant treatment of chronic
agents: amitriptyline, inhibitors or other agents. pain
imipramine, doxepin
Antihistamines, including Anticholinergic toxicity Nonpharmacologic protocol for sleep Obtain history of over-the-counter
diphenhydramine (21); decongestants for colds drug use

Anticholinergics: oxybutynin, Anticholinergic toxicity Lower dose, behavioral measures Even newer agents may have central
other bladder nervous system side effects
antispasmodics
Opioid analgesics, especially Anticholinergic toxicity; Consider using local, regional, and Higher risk in patients with renal
meperidine central nervous system nonopioid analgesics as opioid-sparing insufficiency. Must balance the risks
sedation; fecal impaction strategies. Use scheduled rather than from opioid use against the risks from
PRN dosing when possible. Use pain.
prophylactic strategies to prevent opioid
side effects, such as fecal impaction.
Antipsychotics, especially Anticholinergic toxicity; Eliminate, or if necessary use
low-potency agents central nervous system low-dose, high-potency agents
sedation
Anticonvulsants, especially Mysoline is converted to Consider alternative agent Toxic reactions can occur despite
primidone (Mysoline), phenobarbital, which is a “therapeutic” drug levels
phenobarbital, phenytoin sedating long-acting
barbiturate
Histamine-2 blocking Anticholinergic toxicity Lower dosage. Consider antacids or Usually seen with high-dose and/or
agents proton-pump inhibitors. intravenous use
Antiparkinsonian agents: Dopaminergic toxicity Lower dose. Adjust dosing schedule. Usually with end-stage disease and
levodopa-carbidopa, high doses.
dopamine agonists,
amantadine
Antiarrhythmics Interfere with neuronal Lower dose; consider need for Highly lipophilic and cross
metabolic physiology alternative agent blood–brain barrier
Almost any medication Consider risks and benefits of all
if time course is medications in the elderly
appropriate

© 2011 American College of Physicians ITC6-10 In the Clinic Annals of Internal Medicine 7 June 2011
dosing every 60 minutes as required restraints should be removed as
for symptom management. One soon as possible. Physical restraints
must be careful to assess for probably do not decrease the rate of
akathisia (motor restlessness), falls by confused patients and may
which may be an adverse effect of increase the risk for injury. Re-
high-potency antipsychotic med- straints may also prolong delirium;
ications and can be confused with their use has been shown to be an
worsening delirium. Haloperidol independent predictor of persist-
should be avoided in older persons ence of delirium to hospital dis-
with parkinsonism and Lewy body charge (53).
disease—an atypical antipsychotic
with less extrapyramidal effects When should clinicians consider
may be substituted. The pharmaco- specialty consultation?
logic management of agitation in Similar to diagnosis, most cases of
delirium is summarized in Table 3. delirium can be managed by the
primary care physician or hospital-
Higher-dose intravenous haloperi- based generalist. Consultation may 48. Lacasse H, Perreault
MM, Williamson DR.
dol may be the drug of choice for be helpful for persons whose cause Systematic review of
antipsychotics for
critically ill patients in the ICU of delirium remains obscure or who the treatment of
setting. For such patients, the do not improve after obvious causes hospital-associated
delirium in medically
risk–benefit ratio of medication ad- have been addressed. Medical/sur- or surgically ill pa-
verse effects versus the removal of gical intensivist consultation may tients. Ann Pharma-
cother.
lines and devices often favors phar- be indicated for delirious critical 2006;40:1966-73.
[PMID: 17047137]
macologic treatment. Such therapy care patients. Geriatrics consulta- 49. Lonergan E, Britton
must be used with special caution tion may be particularly helpful for AM, Luxenberg J,
Wyller T. Antipsy-
in older persons. In addition to ex- management of frail, older persons chotics for delirium.
Cochrane Database
trapyramidal effects, the potential with several medical problems, Syst Rev.
for QT interval prolongation and medications, or complex rehabilita- 2007:CD005594.
[PMID: 17443602]
torsade de pointes, neuroleptic ma- tive needs. Psychiatric consultation 50. Seitz DP, Gill SS, van
Zyl LT. Antipsy-
lignant syndrome, and withdrawal is particularly helpful for younger chotics in the treat-
dyskinesias are important concerns patients with extreme agitation or ment of delirium: a
systematic review. J
(52). In all cases where such “phar- other life-threatening behavioral Clin Psychiatry.
2007;68:11-21.
macologic restraints” are used, the disorders requiring high-dose or al- [PMID: 17284125]
health care team must clearly iden- ternative sedating medications 51. Breitbart W, Marotta
R, Platt MM, Weis-
tify the target symptoms necessitat- (Table 3). These patients are vul- man H, Derevenco
ing their use, frequently review the nerable to dangerous side effects, M, Grau C, et al. A
double-blind trial of
efficacy of these drugs in control- and such drug administration haloperidol, chlor-
promazine, and lo-
ling the target symptoms, and as- should be guided by physicians razepam in the treat-
sess the patient for adverse effects experienced in the use of these ment of delirium in
hospitalized AIDS
and complications. medications at high doses. Trials of patients. Am J Psy-
chiatry.
specialty consultation for delirium 1996;153:231-7.
Are physical restraints ever have been mixed, with some show- [PMID: 8561204]
52. Lawrence KR, Nasr-
appropriate? ing no benefit, and others showing away SA. Conduc-
Physical restraints are always objec- a reduction in the delirium severity tion disturbances as-
sociated with
tionable but may be required to and duration (12, 46). While such administration of
control violent behavior or to pre- mixed evidence should not preclude
butyrophenone an-
tipsychotics in the
vent the removal of important de- consultation in complex cases, it critically ill: a review
of the literature.
vices, such as endotracheal tubes, does suggest that prevention is Pharmacotherapy.
intra-arterial devices, and catheters, more effective than treatment, even
1997;17:531-7.
[PMID: 9165555]
particularly in the ICU. In these when done by specialists. 53. Inouye SK, Zhang Y,
Jones RN, Kiely DK,
cases, calm reassurance provided by Yang F, Marcantonio
a sitter or family member may be What is the risk for recurrence ER. Risk factors for
delirium at dis-
more effective than the use of re- after an initial episode and how charge: develop-
ment and validation
straints. Whenever restraints are should clinicians follow patients? of a predictive mod-
used, the indicators for use should An increasing body of evidence el. Arch Intern Med.
2007;167:1406-13.
be frequently reassessed, and the suggests that patients with delirium [PMID: 17620535]

7 June 2011 Annals of Internal Medicine In the Clinic ITC6-11 © 2011 American College of Physicians
 remain vulnerable, even after the further modifications of the
confusion clears. Clinicians should treatment plan, hospitalization,
develop both a short-term and or increased support services.
long-term monitoring plan for pa-
tients with delirium. Medical conditions contributing
to delirium may require follow-
In the short term, delirious pa- up testing, such as ensuring cor-
tients require continued monitor- rection of electrolyte distur-
ing of medical, cognitive and bances, heart failure, and
functional status until they re- infections. Cognitive function
turn to baseline. The frequency can be monitored by using meas-
of monitoring depends on the ures similar to those for delirium
setting and ongoing instability. diagnosis. Assessment of activi-
At a minimum, it should be done ties of daily living (ADL) is par-
daily in the hospital; weekly in ticularly useful for monitoring
recently discharged patients, in- functional recovery from deliri-
cluding those admitted to reha- um. Recovering patients will
bilitation facilities; and monthly have an increased need for assis-
upon the patients’ return to the tance, which can be tapered as
community. Those who are delirium and function improves.
acutely delirious and are being Patients whose cognitive or ADL
managed as outpatients require function does not return to base-
frequent monitoring, perhaps line 1 to 2 months after an
daily on an initial basis, with re- episode of delirium should be
duced frequency as their status considered for comprehensive
improves. Persistent or worsen- geriatrics assessment and/or neu-
ing symptoms may require ropsychological testing.

Table 3. Pharmacologic Management of Agitated Delirium

Agent* Drug Class Dosage Benefits Adverse Effects Comments
Haloperidol Typical 0.25–1 mg PO or Relatively nonsedating; EPS, especially if Usually, agent of choice†
antipsychotic IV q 4 h prn few hemodynamic > 3 mg per day
agitation effects
Olanzapine Atypical 2.5–10 mg PO daily; Fewer EPS than More sedating than Small trials‡PO route less
antipsychotic dissolving tablet haloperidol haloperidol effective for acute
IV: 2.5–10 mg PO management
qd OR
IM: 2.5–10 mg daily
Quetiapine Atypical 25–50 mg PO bid Fewer EPS than Most sedating of Small trials‡
antipsychotic haloperidol atypical antipsychotics
Hypotension
Risperidone Atypical 0.25–1 mg Relatively nonsedating; Might have slightly Small trials‡
antipsychotic PO or IV q 4 h prn few hemodynamic effects fewer EPS than
agitation haloperidol
Lorazepam Benzodiazepine 0.25–1 mg PO or Use in sedative and More paradoxic Second-line agent, except
IV tid prn for alcohol withdrawal, excitation, respiratory in specific cases noted
agitation and history of the depression than
neuroleptic malignant haloperidol
syndrome

bid = twice a day; EPS = extrapyramidal symptoms; h = hour; IV = intravenously; prn = as needed; PO = by mouth; q = each; tid = three times a day.
*Use for delirium is an off-label indication. Due to the small number and size of trials investigating the use of these agents for agitation in delirium, the rec-
ommendations above are Class B.
†In a randomized trial comparing haloperidol, chlorpromazine, and lorazepam in the treatment of agitated delirium in young patients with AIDS, all were
found to be equally effective; however, haloperidol had the fewest side effects or adverse sequelae (51).
‡All atypical antipsychotics have been tested only in small equivalency trials with haloperidol. The Food and Drug Administration has attached warnings to
these agents because of the increased risk for stroke and mortality that has been associated with long-term use, primarily for agitation in dementia.

© 2011 American College of Physicians ITC6-12 In the Clinic Annals of Internal Medicine 7 June 2011
As discussed, minimizing the dura- poor long-term outcomes (7). Pa-
tion of delirium is an important tients who have recovered com-
treatment goal. Patients with deliri- pletely from delirium remain vul-
um that persists for less than 2 nerable to repeated episodes,
weeks will probably fully recover, cognitive decline, functional de-
although it may take weeks to cline, and death. Interventions to
months. Patients with delirium try to improve these long-term out-
lasting longer than 2 weeks are comes have not been well-devel-
much less likely to return to base- oped or studied. At the very least,
line function (54). these patients should be considered
high risk for delirium when rehos-
There is a growing body of evi- pitalized or having surgery, and ap-
dence that delirium, even if re- propriate preventive measures
solved, confers an increased risk for should be implemented.

Treatment... The key elements of delirium treatment are identifying causative and
contributing factors (e.g., medications, infections, electrolyte disturbances); ad-
dressing or reversing these factors to the extent possible; and providing excellent
supportive care to reduce risk for superimposed complications. Minimizing the du-
ration of delirium is an important treatment goal.

CLINICAL BOTTOM LINE

Practice
What measures do stakeholders potentially precipitating factors
Improvement
use to evaluate the quality of care should be evaluated for and identi-
54. Kiely DK, Jones RN,
for patients with delirium? fied causes treated. Bergmann MA, Mur-
No formal quality measures for phy KM, Orav EJ,

delirium have been adopted by U.S. What do professional Marcantonio ER. As-
sociation between
stakeholders. Incident delirium in organizations recommend with delirium resolution
and functional re-
the hospitalized patient was con- regard to prevention, screening, covery among new-

sidered by the Center for Medicare diagnosis, and treatment? ly admitted posta-
cute facility patients.
Several clinical practice guidelines J Gerontol A Biol Sci
& Medicaid Services as a “no-pay” Med Sci.
for delirium have been published
condition, but was ultimately not 2006;61:204-8.
by diverse groups, such as the [PMID: 16510867]
included because of insufficient ev- 55. Shekelle PG,
American Psychiatric Association, MacLean CH, Mor-
idence supporting preventability of
the American Medical Directors ton SC, Wenger NS.
most cases by using current state- Association (for nursing home pa-
Acove quality indica-
tors. Ann Intern
of-the art methods (the most effec- tients), and international organiza- Med. 2001;135:653-
67. [PMID: 11601948]
tive preventive strategies reduce tions. The most recent, comprehen- 56. Young J, Murthy L,
delirium by 30% to 40%). In addi- sive guideline was published in
Westby M, Akunne
A, O’Mahony R;
tion, such “no-pay” status might 2010 by the National Institute for Guideline Develop-
ment Group. Diag-
lead to a potential disincentive to Health and Clinical Excellence nosis, prevention,
diagnose delirium, which could ex- (NICE) of the United Kingdom
and management of
delirium: summary
acerbate current poor recognition Health Service, based on “systemat- of NICE guidance.
BMJ. 2010;341:c3704.
and documentation (see above). ic reviews of the best available evi- [PMID: 20667955]
The ACOVE guidelines (55) in- dence and explicit considerations of 57. O’Mahony R, Murthy
L, Akunne A, Young
clude a single delirium quality indi- cost-effectiveness” (56, 57). Key J; on behalf of the
Guideline Develop-
cator within its indicators for hos- contents of this guideline are sum- ment Group. Synop-
pital care: If a hospitalized sis of NICE Guideline
marized in the Box: Key Recom- for prevention of
vulnerable elder has a suspected or mendations of the NICE Guide- delirium. Ann Intern
Med. 2011;154:746-
definite diagnosis of delirium, then line for Delirium. 51.

7 June 2011 Annals of Internal Medicine In the Clinic ITC6-13 © 2011 American College of Physicians
 Key Recommendations of the NICE Guideline for Delirium
Assess delirium risk factors when patients are admitted to the hospital.
Prevent delirium by addressing risk factors using a multicomponent intervention.
Screen for incident delirium by assessing recent changes or fluctuations in cognitive function,
perception, physical function, and social behavior on admission and at least daily thereafter.
Diagnose delirium by carrying out a clinical assessment based on formal criteria conducted by a
trained health care professional; document in medical record.
Manage delirium by:
• Identifying and managing possible underlying causes
• Ensuring effective communication, reorientation, and providing reassurance
• Considering the involvement of family, friends, and caregivers
• Providing care in a suitable environment.
If a person with delirium is distressed or a risk to themselves or others:
• Use verbal and nonverbal deescalation techniques, such as quietly sitting at the bedside and en-
gaging the patient in conversation, playing relaxing music
• If these are not effective or are inappropriate, consider short-term antipsychotics at the lowest
clinically appropriate dose and titrate cautiously according to symptoms.

In the Clinic
PIER Modules
In the Clinic http://pier.acponline.org

Tool Kit
PIER module on delirium and postoperative delirium. PIER modules provide
evidence-based, updated information on current diagnosis and treatment in an
electronic format designed for rapid access at the point of care.

Patient Information
www.annals.org/intheclinic/toolkit-delirium.html
Patient Information material that appears on the following page for
Delirium duplication and distribution to patients.
www.nlm.nih.gov/medlineplus/ency/article/000740.htm
www.nlm.nih.gov/medlineplus/spanish/ency/article/000740.htm
Information on delirium in English and Spanish from the National
Institutes of Health’s MedlinePlus.
www.uptodate.com/patients/content/topic.do?topicKey=~V0VnMxZiE1
Patient information on delirium from UpToDate, an online, evidence-
based, peer-reviewed information resource.
www.hpna.org/pdf/teachingsheet_managingdelirium.pdf
Handout for patients and their families on managing delirium from the
Hospice and Palliative Nurses Association.

Clinical Guidelines
www.health.vic.gov.au/acute-agedcare/delirium-cpg.pdf
Recommendations to guide the clinical assessment and management of
delirium in older Australians in hospital and across health care settings,
prepared in 2006 for the Australian Health Ministers’ Advisory Council.
www.bmj.com/content/341/bmj.c3704.full
Guidelines on the diagnosis, prevention, and management of delirium
from the British National Institute for Health and Clinical Excellence
(NICE) in 2010.
www.psychiatryonline.com/pracGuide/pracGuideTopic_2.aspx
American Psychiatric Association Practice Guidelines on treatment of
patients with delirum.

Diagnostic Tests and Criteria

www.hospitalelderlifeprogram.org/private/camdisclaimer.php?pageid=01.08.00
The CAM tool can be used to screen for overall cognitive impairment.
www.annals.org
Inouye SK, van Dyck CH, Alessi CA, et al. Clarifying confusion: the
confusion assessment method. A new method for detection of delirium.
Ann Intern Med. 1990;113:941-8.
www.ncbi.nlm.nih.gov/pubmed/18293243
Nursing Delirium Screening Scale
www.medscape.com/viewarticle/581322_appendix3
Delirium Detection Score

© 2011 American College of Physicians ITC6-14 In the Clinic Annals of Internal Medicine 7 June 2011
THINGS YOU SHOULD In the Clinic
Annals of Internal Medicine
KNOW ABOUT DELIRIUM

What is delirium?
• A state of severe confusion that may come and go.
• The confusion may include difficulty staying focused
and fully alert and conscious.
• It may include disorientation and inability to re-
member recent events.

What are risk factors for delirium?

• Being older than 65.
• Having severe, chronic, or terminal illness.
• Having previous brain disease or damage, such as
dementia, Parkinson disease, or stroke.
• Having vision or hearing impairment.
• Being malnourished or dehydrated.

What are the common triggers for

delirium?
• Having multiple medical procedures.
• Sudden, severe illness, particularly infection.
• Admission to the ICU.
• Certain medications, particularly narcotics, antide-
pressants, anticonvulsants, or sleep medications, or
taking multiple medications.
• Drug or alcohol abuse.
• Emotional stress.

How does delirium differ from

dementia? What should you do if a family

Patient Information
• Delirium develops suddenly, over a few hours or
days. members has delirium?
• Dementia develops over many months or years. • Seek immediate medical attention.
• Delirium is usually temporary and goes away once • Try to provide information to the doctor about the
the cause is treated patient’s medications and medical conditions.
• Dementia is usually long-lasting and usually does • The doctor will ask when the symptoms of delirium
not get better. started and how behavior has changed specifically.

For More Information

www.mayoclinic.com/health/delirium/DS01064
Information on delirium from Mayo Clinic.

www.nlm.nih.gov/medlineplus/delirium.html
Information resources on delirium from the National Institutes of
Health’s MedlinePlus.

www.healthinaging.org/agingintheknow/chapters_ch_trial.asp?ch=57
Information on delirium, including how it differs from dementia,
from the Foundation for Health in Aging.
CME Questions

1. A 79-year-old woman was hospitalized 4 fever, dysuria, chest discomfort, or abdominal He is given increasing doses of lorazepam to
days ago after sustaining a right hip fracture discomfort. She rates her back pain as 1 on a cause sedation, and assist-control ventilation is
in a fall. She underwent surgical repair with scale of 1 to 10. She continues to require resumed. The following day he is calm but is
right hip replacement 3 days ago and did not immediate-release morphine. not focused and fails to follow commands
awaken from general anesthesia until 12 consistently.
Vital signs are normal. The patient is alert and
hours after extubation. As her alertness has oriented to name and place but cannot Which of the following is the best test to
increased, she has become increasingly remember the year or date. The patient is still assess the patient’s mental status?
agitated, yelling at the nurses and flailing her agitated and confused, picking at her clothes A. Beck Depression Inventory
arms; mechanical 4-limb restraints were during the examination.
B. Confusion Assessment Method for the
placed 2 days ago. The patient has a 4-year Which of the following is the most
history of progressive cognitive decline Intensive Care Unit (CAM-ICU)
appropriate management for this patient?
diagnosed as Alzheimer dementia. She also C. CT scan of the head
has chronic atrial fibrillation treated with A. Discontinue fentanyl patch D. Metabolic profile
chronic warfarin therapy. She has no other B. Initiate haloperidol E. Mini-Mental State Examination
pertinent personal or family medical history. C. Initiate lorazepam
Current medications are donepezil, D. Measure serum electrolytes, calcium, and 5. A 78-year-old woman is evaluated in the ICU
memantine, atenolol, warfarin, and low- renal and hepatic function for disorientation. The patient recently
molecular-weight heparin. developed the acute respiratory distress
E. Schedule an MRI
syndrome secondary to community acquired
On physical examination today, temperature
3. A 75-year-old woman with a history of chronic pneumonia, and mechanical ventilation was
is 37.2 0C (99.0 0F), blood pressure is 100/68
mm Hg, pulse rate is 100/min and irregular, obstructive pulmonary disease is evaluated in the started 2 days ago. She lives alone and
respiration rate is 18/min, and BMI is 21. ICU for delirium. She had a median sternotomy functions well independently.
The patient can move all 4 limbs with and repair of an aortic dissection and was The patient is on a ventilator; she has received
guarding of the right lower limb. She is extubated uneventfully on postoperative day 4. small doses of lorazepam over the past 48
inattentive and disoriented to time and Two days later she developed fluctuations in her hours and appears comfortable. She has
place and exhibits combativeness mental status and inattention. While still in the recently become disoriented, is not interacting
alternating with hypersomnolence. The ICU, she became agitated, pulling at her lines, as well with her family as she had before, and
remainder of the neurologic examination is attempting to climb out of bed, and asking to has had fluctuations in mental status over the
unremarkable, without evidence of focal leave the hospital. Her arterial blood gas values past 24 hours.
findings or meningismus. are normal. The patient has no history of alcohol On physical examination, pulse rate is 92/min,
Which of the following is the most likely abuse. Frequent orientation cues, calm but vital signs are otherwise normal.
diagnosis? reassurance, and presence of family members Neurologic examination shows no focal
have done little to reduce the patient’s agitated abnormalities, and cranial nerve examination is
A. Acute cerebral infarction
behavior. normal. She is calm and awake but cannot
B. Acute worsening of Alzheimer dementia follow directions to do the “random letter A
C. Meningitis Which of the following is the most appropriate
test‚Äù by squeezing the examiner’s hand only
D. Postoperative delirium therapy for this patient’s delirium?
on hearing the letter “A”; she also cannot
A. Diphenhydramine organize her thinking to answer simple
2. An 85-year-old terminally ill woman is B. Haloperidol questions. When asked whether she is seeing
evaluated in a home hospice setting. She has C. Lorazepam things or hearing things that are not there, she
metastatic breast cancer to the spine, lungs, shakes her head “No.” Laboratory studies
D. Propofol
and liver. She has had progressive anorexia and show hemoglobin of 9.9 g/dL (99 g/L) and a
weight loss and is dependent on family for all leukocyte count of 11,000/µL (11 × 109/L)
4. A 68-year-old man with chronic obstructive
activities of daily living. She has an advance with a normal differential. Metabolic panel
pulmonary disease, hypertension, and
directive stating she does not want reveals plasma glucose of 180 mg/dL (10.0
hyperlipidemia is being weaned from
cardiopulmonary resuscitation or artificial mmol/L); serum total thyroxine and thyroid-
mechanical ventilation after an exacerbation. stimulating hormone levels are normal.
nutrition. Her pain has been well controlled on The patient’s current medications are
a fentanyl transdermal patch and immediate- ipratropium bromide and albuterol (both by Which of the following is the most likely
release morphine as needed for pain. These metered-dose inhaler through the ventilator), diagnosis?
medications have been stable over the past prednisone, lisinopril, and atorvastatin. A. Delirium
month. Last night, the patient became
He is started on a spontaneous breathing trial, B. Dementia
confused and agitated, trying to get out of bed
which he initially tolerates well but later shows C. Psychosis
and repeatedly stating she needed to look for
evidence of oxygen desaturation and agitation. D. Stroke
her deceased husband. There is no dyspnea,

Questions are largely from the ACP’s Medical Knowledge Self-Assessment Program (MKSAP, accessed at
http://www.acponline.org/products_services/mksap/15/?pr31). Go to www.annals.org/intheclinic/
to complete the quiz and earn up to 1.5 CME credits, or to purchase the complete MKSAP program.

© 2011 American College of Physicians ITC6-16 In the Clinic Annals of Internal Medicine 7 June 2011