Delirium in Elderly Patients:How You Can Help

April 1, 2006

Delirium in older adults needs to berecognized early and managed as amedical emergency. Prompt detectionand treatment improve both shortandlong-term outcomes.1,2 Becausedelirium represents one of the nonspecificpresentations of illness in elderlypatients, the disorder can be easilyoverlooked or misdiagnosed. Misdiagnosismay occur in up to 80% of cases,but it is less likely with an interdisciplinaryapproach that includes inputfrom physicians, nurses, and familymembers.3

Delirium in older adults needs to berecognized early and managed as amedical emergency. Prompt detectionand treatment improve both shortandlong-term outcomes.1,2 Becausedelirium represents one of the nonspecificpresentations of illness in elderlypatients, the disorder can be easilyoverlooked or misdiagnosed. Misdiagnosismay occur in up to 80% of cases,but it is less likely with an interdisciplinaryapproach that includes inputfrom physicians, nurses, and familymembers.3

Successful prevention of deliriumrequires systematic evaluation of hospitalizedelderly patients. In this article,I identify key risk factors to be alert forin the evaluation. I also outline a multidisciplinaryapproach to diagnosisand management.

Delirium is characterized by aglobal disorder of attention and cognition.It is present in 10% to 40% of elderlypersons at the time of hospitaladmission; the incidence rises to 25%to 60% during the stay.4 Delirium goesunrecognized in 32% to 66% of this population.4 It affects 2.2 million hospitalizedelderly persons at an estimatedcost to Medicare of over $4 billion(1994 dollars) per year.5

The prevalence of delirium in elderlypatients differs according to thetype of hospital admission; it variesfrom 10% to 15% on general medicaland surgical wards to more than 50%in persons with hip fractures. The riskof delirium is related to the underlyinghealth of the patient and the magnitudeand impact of the medical or surgicalintervention. The risk increaseswith prolonged length of hospital stay.

Patients may present with the hyperactive,hypoactive, or mixed form ofdelirium.

  • The hyperactive state is often recognized,but a misdiagnosis of dementiaor the administration of antipsychoticor sedative drugs alone willdelay the proper diagnosis and increasethe complications of delirium.
  • The hypoactive form of deliriumoften goes unrecognized, which resultsin similar delays and complications.It is no surprise, therefore, thatdelirium has been associated withhospital mortality rates of 10% to 65%in addition to an increased risk ofinstitutionalization.5-7
  • The mixed form includes the hyperactiveand hypoactive states. Thepatient may cycle from one to anotherover a period of hours.

Delirium is a predictor of longtermfunctional decline and permanentloss of functional independence.6-8 Onestudy showed loss of function in an averageof 1 activity of daily living duringhospitalization; delirium was the solepredictor of this loss of independence.9

PREVENTIONIdentifying risk factors. Moststudies have focused on incident delirium--that is, the onset of delirium isdocumented during the hospital stay--and have identified risk factors andprecipitating factors. (For patients whopresent with symptoms of delirium athospital admission, an interdisciplinaryapproach to improve overall functionand prevent further decline in mentalstatus is most appropriate.)

One study of incident deliriumidentified 4 independent baseline riskfactors present on admission in elderlypatients in whom delirium subsequentlydeveloped during hospitalization10:

  • Vision impairment (adjusted relativerisk [RR], 3.5).
  • Severe illness (RR, 3.5).
  • Cognitive impairment (eg, dementia)(RR, 2.8).
  • A high blood urea nitrogen-creatinineratio (RR, 2.0).

This predictive model can beused to identify patients at highest riskfor delirium during the hospital stay.Functional impairment, comorbid conditions,and related pharmacotherapymay be associated with a decline incognitive reserve and an increasedrisk of delirium. Table 1 shows therates of delirium and death or nursinghome placement related to these riskfactors.

 Table 1 - Risk factors for delirium and risk of death or nursing home placement*
Risk factors present Risk of delirium, %Risk of death or nursing home placement, %

0 = Low risk 33

1 or 2 = Intermediate risk 1614

3 or 4 = High risk 3226

*Risk factors: vision impairment, severe illness, cognitive impairment, high blood urea nitrogen–creatinine ratio. Adapted from Inouye SK et al. Ann Intern Med. 1993.

Prevention strategies must address,at a minimum, the following 5factors known to precipitate incidentdelirium8:

  • Use of physical restraints (RR, 4.4).
  • Malnutrition (RR, 4.0).
  • More than 3 medications added duringthe hospital stay (RR, 2.9).
  • Use of a bladder catheter (RR, 2.4).
  • Any iatrogenic event (RR, 1.9).

The presence of 1 or 2 of thesefactors was associated with a 20% increasein the incidence of delirium.The presence of 3 or more was associatedwith a 35% increase.

Preventive strategies for nonsurgicalpatients. An intervention toprevent delirium in patients on medicalwards has addressed the needs ofthose with cognitive, visual, or hearingimpairment and minimized sleep deprivation,immobility, and dehydration.11 This approach significantly reducedthe incidence of delirium (from15% to 9.9%). A decrease in the useof sleep medications and improvedcognition were also demonstrated,although there was no difference inlength of hospital stay between the in-tervention and "usual-care" groups.Thus, justification for implementingthe protocol as a standard of care requiresfurther analysis of its long-termimpact on health outcomes, includingrates of readmission, functional decline,institutionalization, and death, aswell as cost. The authors emphasizethat primary prevention is the most effectivestrategy, because once deliriumhas occurred, intervention is lesseffective. Cost-effectiveness considerationsmay dictate that these multidisciplinaryprotocols be targeted to highriskpatient groups.

Preventive strategies for surgicalpatients. These interventions are individualizedaccording to the type ofsurgery--whether urgent or elective--and underlying health status. Not surprisingly,the highest rates of deliriumoccur in the subset of patients admittedfor hip fracture repair. These patientsare likely to be undergoing surgery onan urgent basis; they are often frail andhave risk factors for incident delirium atthe time of injury. In addition, they oftenhave comorbid conditions and functionaldependence. Despite the increasedrisk, delirium is significantly underdiagnosedin this population.12 Standardpostoperative care procedures, includingprevention of hypoxemia and hypotension,result in decreased incidenceand severity of delirium.13

Delirium rates are lower in patientsundergoing elective joint replacementthan in those undergoingemergent hip fracture repairs. Risk factorsin the former group can be addressedpreoperatively, and preventivestrategies in those with cognitive, visual,or hearing impairment can be implementedto minimize sleep deprivation,immobility, and dehydration. Earlydetection and management of postoperativedelirium helps reduce the severityof delirium in both groups.13,14

Risk factors for delirium in patientsundergoing elective noncardiacsurgery include age greater than 70years; a history of alcohol abuse; cognitiveor functional impairment; metabolicdisturbances (elevated serum levelsof sodium, potassium, or glucose);or a history of noncardiac thoracicsurgery or abdominal aortic aneurysmsurgery.15 Significant intraoperativeblood loss and hemodynamic instabilityare additional risk factors.16

A number of studies have shownthat patients in whom delirium developshave higher complication rates, longerhospital stays, and increased rates oftransfer to rehabilitation or long-termcare facilities.5 The causes of these pooroutcomes--which have been observedin both medical and surgical settings--may vary according to the patient'spreadmission status, illness severity,and management of the condition forwhich the patient was admitted. Specificprotocols are being developed that intenselymonitor for and prevent or aggressivelymanage delirium and itscomplications; they may improve outcomesand reduce the cost of care.

In spite of the recognized neuropsychologicalconsequences of cor-onary artery bypass graft (CABG)surgery, there are few prospective studiesof delirium in elderly patients undergoingthis procedure. Cognitive changesfollowing CABG surgery were recentlyshown to persist in 42% of patients for atleast 5 years after the procedure.17 Deliriumwould have developed as the initialpresentation of acute cognitive change inmany of these patients. Postoperativedelirium develops in an estimated 30% ofpersons 65 years and older who undergoCABG surgery.18,19 If this condition isnot recognized and treated, it may resultin long-term cognitive impairment. Protocolsto prevent delirium in this veryhigh-risk population have yet to be published.Developing such strategies mayrequire a better understanding of thepathophysiology of delirium and morespecific targeting to the mechanism ofischemic and reperfusion injury. Interventionsfor early detection and managementof delirium that have reducedhospital length of stay and rates of longterminstitutionalization and death inother subsets of patients may protectagainst the long-term consequences ofdelirium in patients undergoing CABGsurgery.

MAKING THEDIAGNOSISHistory and examination. Table2 lists the common causes and precipitantsof delirium that require earlyrecognition and management. Sincethe cause of delirium is often multifactorial,it is prudent to screen for all ofthese conditions. Detection of deliriumand related conditions requires acareful history taking, physical examination,and patient and family interviewsthat establish an accurate pictureof the patient's pre-illness status.Selected laboratory studies as indicatedfrom the clinical evaluation, includingthose shown in Table 3, should beconsidered in the evaluation.

 Table 2 – Common medical causes of delirium in hospitalized elderly patients
Medications: prescription or over-the-counter 


Review the patient's current medications(including over-the-counterpreparations and alcohol). Ascertainwhether withdrawal from benzodiazepinesor alcohol may have precipitatedthe delirium. Discontinue agents onthe "Beers list" of medications that arebest avoided by elderly persons (Table4).20 These include opioids, such as hydrocodone,codeine, meperidine, andpropoxyphene; amitriptyline; trimethobenzamide;anticholinergic drugs; benzodiazepines(especially long-actingformulations, such as alprazolam); andmuscle relaxants.

 Table - Delirium management
Action Example

Assess cause and order appropriate laboratory and diagnostic tests Pharmacologic: determine whether the patient is taking 3 or more medications, including prescription and OTC agents; ascertain whether the patient is undergoing withdrawal from benzodiazepines or alcohol Infection: order CBC count (and blood cultures if temperature > 37.7°C [100°F] or < 35°C [95°F]) Metabolic: assess electrolyte, BUN, creatinine, calcium, glucose, albumin levels Cardiopulmonary: order ECG, chest radiograph, oxygen saturation measurement Urinary/fecal retention: perform rectal examination; assess PVRV Neurologic: assess new focal signs

Manage precipitating factors Pain: acetaminophen, 650 mg PO/PR qid scheduled dose, and prn morphine, 2.5 - 5 mg SC q3h or 5 - 10 mg PO q4h, or oxycodone, 5 - 10 mg q4h. Avoid long-acting preparations of narcotics or patient-controlled analgesia in patients with delirium Oxygenation: oxygen for saturation > 92% Fluid/electrolytes: intravenous administration if patient takes < 1200 mL PO; normal saline if Na level < 145 mmol/L Nutrition: liquid supplement between meals Bowel/bladder: prune juice/milk of magnesia; intermittent catheterization q8h if PVRV > 150 mL Agitation: loxapine, starting dosage 2.5 - 5 mg PO bid prn to a maximum of 10 mg tid prn. For severe agitation, give loxapine as scheduled dose for first 2 d, then reassess

Involve multidisciplinary team (physicians, nurses, social workers, physical therapists, speech therapists, occupational therapists, and pharmacists) Cognition: evaluate with MMSE, CAM, BCRS Immobility: early mobilization, remove tubes, no restraints Dysphasia: assess swallowing, modify diet Pressure ulcers: skin precautions, mobilization ADL function: provide appropriate assistance Social support: involve family/caregivers

OTC, over-the-counter; CBC, complete blood cell; BUN, blood urea nitrogen; PVRV, urine post-void residual volume; Na, sodium; MMSE, Mini-Mental State Examination; CAM, Confusion Assessment Method; BCRS, Brief Cognitive Rating Scale; ADL, activities of daily living.

Use metoclopramide with caution,because it is associated with extrapyramidalside effects that maymanifest with delirium-like features.Cimetidine and, to a lesser extent, ranitidine,may also cause delirium.

 Table 4 –Drugs to avoid in elderly persons
Antiarrhythmics    Digoxin > 0.125 mg    Disopyramide Anticholinergics Antidepressants    Amitriptyline Antiemetics    Trimethobenzamide Antihistamines    Cyproheptadine    Diphenhydramine    Hydroxyzine    Promethazine Antihypertensives    Methyldopa Antiplatelet agents    Dipyridamole    Ticlopidine Benzodiazepines Iron supplements > 325 mg Muscle relaxants NSAIDs    Indomethacin Opioids    Meperidine    Pentazocine    Propoxyphene 

Adapted from Beers MH. Arch Intern Med. 1997.

Diagnostic tools. Delirium is aclinical diagnosis. The changing levelsof confusion and consciousness associatedwith delirium may lead to amissed diagnosis and, hence, a lostopportunity for improving outcomes.The Confusion Assessment Method(CAM) is a useful tool that enablesnonpsychiatric clinicians to quicklydistinguish delirium from other causesof altered mental status in the hospitalsetting (Table 5).21 The criteriainclude acute onset and fluctuatingcourse, inattention, and either disorganizedthinking or altered level ofconsciousness (agitation or somnolence).The CAM is sensitive and specificfor the diagnosis of delirium, althoughthe sensitivity may vary from0.13 to 1.0,2,5,14,22,23 depending on theexperience and training of the personusing it.

It may often be difficult to distinguishdelirium from dementia. Theconditions may also coexist, since dementiais a risk factor for delirium. TheCAM was developed from the standardcriteria for delirium. Table 6 offersadditional criteria to help differentiatebetween delirium and dementia.

The approach outlined in Table 3reflects my experience in providinggeriatrics consultation at a tertiarycare teaching hospital in Edmonton,Canada. My observations have beensupported by the emerging literature.My colleagues and I have learned thata consistent interdisciplinary approachto the detection and management ofpotential causes is critical to improvingoutcomes in patients with delirium.The use of this protocol for 5 monthsto manage 143 medical and surgicalpatients with delirium resulted in a 20%reduction in length of hospital stay andimproved discharge disposition.

Antipsychotic drugs and sedativesare best avoided; if required(such as when a patient's agitationmay compromise treatment or harmthe patient or others), use them forspecific therapeutic indications at thelowest effective dose. To minimize theseverity of delirium, avoid physical restraintsand the use of drugs as chemicalrestraints.

Inform families that delirium is atemporary and treatable condition andinvolve them in the care plan. Familyinvolvement is often beneficial both tothe patient and the family. Familymembers are often not aware of thetransient nature of delirium and believe it to be a permanent change inmental status. This misunderstandingcan lead them to prematurely advocatefor placement in long-term care facilities.When fully informed, families aremore likely to adopt positive attitudesand become an important part of thetreatment team.

Loxapine, an intermediate-actingantipsychotic agent, is our first choicefor managing agitating delirium. Werecommend a low oral dosage of 2.5 to10 mg bid or tid. Although no studieshave compared this drug to the morecommonly used antipsychotic agentshaloperidol and risperidone, loxapineis associated with a much lower risk ofextrapyramidal side effects than these2 agents. Because loxapine has bothsedative and antipsychotic properties,other sedatives do not have to beadded. Postural hypotension occurs infrequentlyat low doses of loxapine.

 Table 5 - The Confusion Assessment Method*
Criterion Evidence

1. Acute onset and fluctuating course Acute change in mental status from the patient's baseline; abnormal behavior comes and goes or increases and decreases in severity

2. Inattention Patient is easily distracted or has difficulty keeping track of what was said

3. Disorganized thinking Rambling or irrelevant conversation, unclear or illogical flow of ideas, or unpredictable switching from subject to subject

4. Altered level of consciousness Patient is hyperalert, lethargic, stuporous, or comatose

If haloperidol is used, the recommendeddosage range is 0.5 to 1 mgbid (up to 2 mg tid). It may be combinedwith lorazepam, 0.5 to 1 mg bid(up to 2 mg tid), when additional sedationis required. Haloperidol may bemore useful for patients whose deliriumis precipitated or exacerbated byanticholinergic agents.

It is critical to aggressively withdrawunnecessary medications andavoid medication changes that may contributeto delirium. Cognitive-enhancingdrugs and antidepressants are bestwithheld until the delirium clears. It isparticularly important to avoid trying tomanage agitation with complicated regimensof antipsychotics, benzodiazepines,narcotics, and anticholinergics,since the side effects of these combinationsmay increase agitation.

 Table 6 - Comparison of delirium and dementia
Delirium Dementia

Sudden onset Insidious onset

Precise time of onset Uncertain time of onset

Usually reversible Slowly progressive

Short duration (usually days to weeks) Long duration (years)

Fluctuations (usually over minutes to hours) Good days and bad days

Abnormal levels of consciousness Normal level of consciousness

Typically, an association with drug use or withdrawal or with acute illness Typically, no association with drug use or acute illness

Almost always worse at night (sundowning) Often worse at night

Inattention Attention not sustained

Variable disorientation Disorientation to time and place

Typically slow, incoherent, and inappropriate language Possible difficulty finding the right word

Impaired but variable recall Memory loss, especially for recent events


REFERENCES:1. Inouye SK, Schlesinger MJ, Lydon TJ. Delirium:a symptom of how hospital care is failing older personsand a window to improve quality of hospitalcare. Am J Med. 1999;106:565-573.
2. Rockwood K, Cosway S, Stolee P, et al. Increasingthe recognition of delirium in elderly patients.J Am Geriatr Soc. 1994;42:252-256.
3. Beers MH, Berkow R, eds. The Merck Manual of Geriatrics. 3rd ed. Whitehouse Station, NJ: MerckResearch Laboratories; 2000.
4. Inouye SK. Delirium in hospitalized older patients:recognition and risk factors. J Geriatr PsychiatryNeurol. 1998;11:118-125, 157-158.
5. Inouye SK. The dilemma of delirium: clinical andresearch controversies regarding diagnosis andevaluation of delirium in hospitalized elderly medicalpatients. Am J Med. 1994;97:278-288.
6. Francis J, Kapoor WN. Prognosis after hospitaldischarge of older medical patients with delirium.J Am Geriatr Soc. 1992;40:601-606.
7. Williams-Russo P, Urquhart BL, Sharrock NE,Charlson ME. Post-operative delirium: predictorsand prognosis in elderly orthopedic patients. J AmGeriatr Soc. 1992;40:759-767.
8. Inouye SK, Charpentier PA. Precipitating factorsfor delirium in hospitalized elderly persons. Predictivemodel and interrelationship with baseline vulnerability.JAMA. 1996;275:852-857.
9. Murray AM, Levkoff SE, Wetle TT, et al. Acutedelirium and functional decline in the hospitalizedelderly patient. J Gerontol. 1993;48:M181-M186.
10. Inouye SK, Viscoli CM, Horwitz RI, et al. A predictivemodel for delirium in hospitalized elderlymedical patients based on admission characteristics.Ann Intern Med. 1993;119:474-481.
11. Inouye SK, Bogardus ST Jr, Charpentier PA, etal. A multicomponent intervention to prevent deliriumin hospitalized older patients. N Engl J Med.1999;340:669-676.
12. Gustafson Y, Brannstrom B, Norberg A, et al.Underdiagnosis and poor documentation of acuteconfusional states in elderly hip fracture patients.J Am Geriatr Soc. 1991;39:760-765.
13. Gustafson Y, Brannstrom B, Berggren D, et al.A geriatric-anesthesiologic program to reduce acuteconfusion states in elderly patients treated forfemoral neck fractures. J Am Geriatr Soc. 1991;39:655-662.
14. Fisher BW, Flowerdew G. A simple model for predicting postoperative delirium in older patientsundergoing elective orthopedic surgery. J Am GeriatrSoc. 1995;43:175-178.
15. Marcantonio ER, Goldman L, Mangione CM, etal. A clinical prediction rule for delirium after electivenoncardiac surgery. JAMA. 1994;271:134-139.
16. Marcantonio ER, Goldman L, Orav EJ, et al.The association of intraoperative factors with thedevelopment of postoperative delirium. Am J Med.1998;105:380-384.
17. Newman MF, Kirchner JL, Phillips-Bute B, et al.Longitudinal assessment of neurocognitive functionafter coronary-artery bypass surgery. N Engl J Med.2001;344:395-402.
18. Rolfson DB, McElhaney JE, Rockwood K, et al.Incidence and risk factors for delirium and other adverseoutcomes in older adults after coronary arterybypass graft surgery. Can J Cardiol. 1999;15:771-776.
19. MacDonald P, Stadnyk K, Cossett J, et al. Outcomesof coronary bypass surgery in elderly people.Can J Cardiol. 1998;14:1215-1222.
20. Beers MH. Explicit criteria for determining potentiallyinappropriate medication use by the elderly:an update. Arch Intern Med. 1997;157:1531-1536.
21. Inouye SK, van Dyck CH, Alessi CA, et al. Clarifyingconfusion: the confusion assessment method.A new method for detection of delirium. Ann InternMed. 1990;113:941-948.
22. Pompei P, Foreman M, Cassel CK, et al. Detectingdelirium among hospitalized older patients. ArchIntern Med. 1995;155:301-307.
23. Rolfson DB, McElhaney JE, Jhangri GS,Rockwood K. Validity of the confusion assessmentmethod in detecting postoperative delirium in theelderly. Int Psychogeriatr. 1999;11:431-438.