To Name :
To Email :
From Name :
From Email :
Comments :

Evidence-Based Reviews

Why off-label antipsychotics remain first-choice drugs for delirium

Short-term, low-dose therapy appears to be worth the risk despite black-box warning.

Vol. 6, No. 10 / October 2007

Delirium is a medical emergency that needs to be identified and treated vigorously. Antipsychotics—including haloperidol and atypical agents—effectively manage a wide spectrum of delirium symptoms and are an essential component in the standard multimodal approach.1 Even so, antipsychotics are not FDA-approved for treating delirium, and evidence on their safety in medically ill patients is limited—particularly in the elderly, in whom delirium occurs most often.

The FDA has warned of increased risk of death when atypical antipsychotics are used to treat behavioral disturbances in elderly patients with dementia.2 Similarly, a retrospective study of elderly patients taking antipsychotics found higher mortality rates associated with typical antipsychotics than with atypicals.3

This article discusses the risks and benefits of using antipsychotics to manage delirium. Based on the literature and clinical experience, we offer recommendations on choosing among the available agents and avoiding side effects.

A challenging diagnosis

Delirium is a neuropsychiatric syndrome precipitated by an underlying medical condition or a medication effect on the brain. Its characteristic symptoms—abrupt onset of disturbed consciousness, attention, cognition, and perception—tend to fluctuate during the day. Delirium most often occurs in elderly patients (Box)1,4-7—particularly with dementia—but also occurs in younger patients with serious illnesses such as cancer or HIV-AIDS.

Delirium is underdiagnosed and under-treated in medical settings,4,8 most likely because of its protean symptoms (Table 1)9 and fluctuating clinical findings. Neurologic abnormalities—including cortical and motor symptoms—also can occur.1

Mortality risk. Delirium is an independent risk factor for mortality.1,4,5 It is a marker for serious and potentially life-threatening medical problems, such as organ failure or sepsis. When antipsychotics fail to control delirium, the 3 most common reasons are:

  • delirium’s etiology has not been discovered or addressed
  • delirium’s etiology is resistant to treatment or potentially irreversible
  • antipsychotic dosage was inadequate.

Given the first 2 reasons, patients with uncontrolled delirium are likely to be more seriously ill and less likely to recover than those whose delirium more readily resolves. After prolonged episodes, patients also may have decreased cognitive function post-delirium.

3 subtypes. Delirium is classified as hyperactive, hypoactive, or mixed, depending on arousal disturbance and psychomotor behavior:

  • the hyperactive subtype includes hallucinations, delusions, agitation, and disorientation.
  • the hypoactive subtype includes confusion, sedation, and decreased alertness but rarely hallucinations or delusions.1

In two-thirds of delirium cases, patients show hypoactive or mixed symptoms.


Delirium: Harbinger of death in the elderly

Up to 1 in 4 patients (14% to 24%) have delirium at hospital admission, and the annual incidence of delirium is 6% to 56% among hospital populations.4 Elderly inpatients who develop delirium have an estimated mortality rate of 22% to 76% during that hospitalization.1 At the end of life, the prevalence of delirium may be as high as 85%.5

Serotonergic, noradrenergic, opiatergic, glutamatergic, and histaminergic neurotransmitter systems may contribute to delirium as a syndrome. Evidence implicates underactivity of the cholinergic system as the final common pathway.6,7

The acetylcholine-dopamine hypothesis explains the efficacy of dopamine antagonists in treating delirium by regulating the imbalance between cholinergic and dopaminergic activity.5,6 Cytokines—including interleukin-1, interleukin-2, and interleukin-6—and chronic hypercortisolism may also contribute to delirium.4

Antipsychotics: Limited evidence

The multimodal approach for managing delirium includes:

  • identifying and eliminating contributing factors
  • instituting nonpharmacologic interventions based on environmental strategies (Table 2)4
  • providing pharmacologic interventions—primarily antipsychotics—as needed.

Clinical trials. Most studies of antipsychotics for delirium have been open-label trials, case reports, and retrospective reviews. A review of 14 prospective studies10 showed that:

  • delirium severity improved with haloperidol, chlorpromazine, olanzapine, risperidone, or quetiapine
  • comparison trials did not identify any antipsychotic as more efficacious than another.

Serious adverse events attributable to antipsychotics were uncommon, although most trials did not systematically evaluate side effects. None included a placebo comparison to explain spontaneous improvements in delirium. The authors concluded that evidence is limited for using low-dose antipsychotics for short-term delirium treatment.

Michaud et al11 reviewed guidelines, systematic reviews, randomized controlled trials, and cohort studies on delirium management. They concluded that the experts agree on 3 points:

  • prevention should be emphasized
  • atypical antipsychotics are not first-choice drugs because of data on adverse events in the elderly
  • pharmacologic treatment is recommended when the patient’s condition prevents adequate care or puts the patient or staff at risk.

Conclusion. We believe these findings signify the lack of sufficient data on pharmacologic treatment of delirium. Further research is needed to assess the efficacy of antipsychotics in delirium treatment.

Conventional antipsychotics

Haloperidol, the most-studied antipsychotic in delirium treatment, often is the drug of choice because of its high potency, low sedative effect, few anticholinergic side effects, minimal cardiovascular side effects, no active metabolites, and multiple administration routes.1

An IV route can facilitate rapid onset of medication effects. Compared with oral haloperidol, IV administration is associated with a lower risk of extrapyramidal symptoms (EPS), which allows use of higher doses.

Any IV use of injectable haloperidol is off-label, however. If you choose the IV route, monitor patients carefully for cardiac arrhythmias. Haloperidol’s prescribing information carries a new warning of sudden death, QT prolongation, and torsades de pointes in patients given IV haloperidol.

Chlorpromazine. In a double-blind, randomized comparison trial of 30 hospitalized AIDS patients, our group12 found oral and IM haloperidol (n=11) or chlorpromazine (n=13) highly effective in controlling delirium. Delirium symptoms improved significantly in both hypoactive and hyperactive subtypes with low doses of either antipsychotic (approximately 2 mg of haloperidol equivalent/day).

No patients developed dystonic or dyskinetic symptoms. Lorazepam, given to 6 patients, worsened delirium and cognitive impairment.

Table 1

Recognizing delirium: Diagnostic clinical features*

Altered level of alertness and arousal

Rapidly fluctuating course

Attention disturbance

Increased or decreased psychomotor activity

Disturbance of sleep-wake cycle

Affective symptoms

Altered perceptions

Disorganized thinking and incoherent speech

Disorientation and memory impairment

* Not all symptoms are present in every case.

Source: Reference 9

Table 2

Nonpharmacologic approaches to managing delirium

Search for and correct all causes of delirium, including underlying disease or a medication effect

Create a calm, comfortable environment

Provide orienting objects such as calendars and clocks

Have family members present

Limit room and staff changes

Allow patients uninterrupted rest at night to improve the sleep-wake cycle

Consider 1-to-1 nursing observation, as necessary

Source: Reference 4

Atypicals in delirium: Trial data

Risperidone. Three open-label studies of risperidone in patients with delirium reported minimal risk of sedation and EPS.13-15

A 7-day, double-blind, flexible-dose trial of 24 patients with delirium16 found no significant difference between haloperidol (mean 1.71 mg/d) and risperidone (mean 1.02 mg/d) in clinical efficacy or response rate. The authors acknowledged, that they were unable to obtain identical-looking haloperidol and risperidone tablets for the trial.

Kim et al17 studied dopamine transporter gene polymorphism and use of haloperidol vs risperidone in 42 patients with delirium. Relatively low doses of both antipsychotics showed similar efficacy, and the authors concluded that dopamine transporter gene polymorphism did not influence delirium treatment.

Olanzapine. In an open trial of 79 inpatients with advanced cancer, olanzapine (mean 6.3 mg/d, range 2.5 to 20 mg/d) resolved delirium in 76% of patients, with no incidence of EPS.18 Age >70, history of dementia, hypoxia, cerebral metastasis, and hypoactive delirium were associated with poor response to olanzapine. This study is unique in assessing olanzapine’s efficacy in different delirium subtypes.

A prospective, randomized trial compared olanzapine (mean 4.5 mg/d, range 2.5 to 13.5 mg/d) with haloperidol (mean 6.5 mg/d, range 1 to 28 mg/d) in patients admitted with delirium to a critical care setting.19 Both treatment groups showed similar improvement over 5 days. No side effects were reported in the patients receiving olanzapine.

Quetiapine. A few authors have published their experience with quetiapine in treating delirium. An open-label, flexible-dose trial of 22 inpatients20 showed significant improvement in delirium severity with the use of quetiapine. No patients experienced EPS; sedation was the most common side effect.

Ziprasidone. In the first case report in which ziprasidone was used to treat delirium,21 an HIV/AIDS patient was given 100 mg/d. Delirium symptoms improved, but treatment was discontinued because of side effects (hypokalemia, hypomagnesemia, premature ventricular contractions, and QT interval prolongation).

Aripiprazole. Straker et al22 reported 14 cases delirium treated with aripiprazole, which showed few side effects. Twelve patients had a ≥50% decrease in Delirium Rating Scale scores, and 13 showed improvement in Clinical Global Impression scale scores.

Clinical options

When choosing an antipsychotic to treat delirium, consider the individual patient’s risks of EPS, sedation, anticholinergic side effects, cardiac arrhythmias, and drug-drug interactions.

Haloperidol. When medication is necessary for delirium, American Psychiatric Association (APA) guidelines consider low-dose haloperidol as first-line treatment (see Related Resources). Recommended dosage is 1 to 2 mg (0.25 to 0.5 mg for the elderly) every 4 hours as needed.

Adding oral or IV lorazepam (0.5 to 1 mg every 1 to 2 hours) to haloperidol may help rapidly sedate the agitated delirious patient and minimize the risk of EPS associated with haloperidol.1 Avoid benzodiazepine monotherapy unless delirium is related to alcohol or benzodiazepine withdrawal.

Chlorpromazine. We have successfully used oral or IV chlorpromazine (12.5 to 50 mg every 4 to 12 hours) instead of haloperidol plus lorazepam when increased sedation was required, especially:

  • in the ICU, where close blood pressure monitoring was feasible
  • for severe agitation in terminally ill patients to decrease distress for the patient, family and staff.

Monitor chlorpromazine’s anticholinergic and hypotensive side effects, particularly in elderly patients. Its anticholinergic effects could worsen delirium, but we are not aware of any studies or case reports supporting that clinical outcome.

Atypical antipsychotics also may be used to treat delirium, as supported by the literature. Recommended dosing, available routes administration routes, and clinical comments are summarized in Table 3.23

Table 3

Recommended antipsychotic dosing for delirium*





Typical agents


Initial: 0.5 to 1 mg Range: 0.5 to 2 mg every 2 to 12 hours

Oral, IV, SC, IM

‘First choice’ for delirium when antipsychotic treatment is needed (per APA guidelines)


Initial: 12.5 to 25 mg Range: 12.5 to 50 mg every 4 to 12 hours

Oral, IV, IM

Alternative to haloperidol plus lorazepam when increased sedation is needed

Atypical agents


Initial: 0.25 to 1 mg Range: 0.25 to 2 mg/d


Risk of sedation and orthostatic hypotension at higher doses


Initial: 2.5 to 5 mg nightly Range: 2.5 to 10 mg/d


Sedation (a potential limiting factor) may be beneficial for hyperactive delirium


Initial: 25 to 50 mg Range: 25 to 200 mg/d, usually divided into 2 daily doses


Sedation and orthostatic hypotension are dose-limiting factors


Initial: 20 mg bid Range: 20 to 160 mg/d, usually divided into 2 daily doses


Limited data in delirium because of concerns about QT interval prolongation in medically ill patients


Initial: 10 to 15 mg Range: 10 to 30 mg/d


‘Dopamine stabilizing’ effect might be preferable in hypoactive delirium

* For frail elderly patients, start with approximately one-half the suggested initial dose.

† Risperidone and aripiprazole are available in liquid formulations. Risperidone, olanzapine, and aripiprazole are available in orally disintegrating tablets.

APA: American Psychiatric Association; IM: intramuscular; IV: intravenous; SC: subcutaneous

Source: Reference 23

Managing adverse effects

Reassess patients frequently during a delirium episode to adjust the antipsychotic dose, search for underlying causes, and monitor for side effects (Table 4). In frail elderly patients, start with approximately one-half the recommended initial dose to reduce the side effect risk.

Antipsychotics may not be appropriate in certain populations with delirium, particularly in patients with:

  • dementia of Lewy body type or Parkinson’s disease dementia
  • stroke
  • history of adverse reactions to antipsychotics.

Mortality risk. All atypicals carry a “black-box” warning of increased risk of death when treating behavioral disturbances in elderly patients with dementia-related psychosis. The FDA advisory is based on a meta-analysis by Schneider et al2 of 17 placebo-controlled trials totaling 3,353 patients with Alzheimer’s disease or dementia. Risk of death in the drug-treated patients was 1.6 to 1.7 times greater than in those who received placebo. Most deaths were associated with cardiovascular disease or infection (including pneumonia).

Did you miss this content?
Botulinum toxin for depression? An idea that’s raising some eyebrows