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

Evidence-Based Reviews

Atypical antipsychotics for delirium: A reasonable alternative to haloperidol?

Newer agents may offer similar efficacy with fewer adverse effects

Vol. 10, No. 01 / January 2011

Discuss this article at

Ms. B, age 48, is admitted to our hospital after overdosing on unknown amounts of amitriptyline, diphenhydramine, and laxatives. Three days after admission, the psychiatry service is consulted to assess her for “bipolar disorder.” Although Ms. B does not have a psychiatric history, her internist believes her pressured speech and psychomotor agitation warrant investigation.

During the initial psychiatric interview, Ms. B is disoriented, with fluctuating alertness and cognition. The Confusion Assessment Method for the Intensive Care Unit (CAM-ICU)1 is positive for delirium. We perform a delirium workup while we start Ms. B on olanzapine, 5 mg/d orally and 5 mg intramuscular (IM) every 8 hours as needed.

Ms. B’s laboratory results (complete blood count, complete metabolic profile, urinalysis, chest roentgenogram, vitamin B12 level, blood alcohol level, urine drug screen, arterial blood gas, and head CT) are unremarkable except for her amitriptyline/nortriptyline level, which is in the toxic range. On physical examination, Ms. B’s heart rate and temperature are elevated, her pupils are dilated and sluggish, and her skin is hot and dry. Based on these findings, we determine that Ms. B’s delirium most likely is an anticholinergic syndrome from amitriptyline/diphenhydramine toxicity.2 We discontinue olanzapine after only 2 doses because of its potential anticholinergic effects.3

In hospitalized patients, delirium is one of the most frequently encountered mental disorders, but because of its variable presentation the condition often is underrecognized and undertreated, which leads to longer hospitalizations and increased mortality.4,5 Ms. B’s case illustrates the classical delirium presentation (Table 1),6 highlighting 2 hallmark features of the disorder: inattention and an acute fluctuating course.4 Unfortunately, delirium is a diverse disorder that may present with numerous nonclassical symptoms—including lethargy, emotionality, sleep irregularities, and neurologic abnormalities—in lieu of more commonly recognized symptoms.4,5

In addition to recommending identifying and addressing the underlying acute illness, American Psychiatric Association guidelines suggest prescribing psychotropic medications to treat delirium symptoms.5,7 Antipsychotics are considered first-line pharmacotherapy because they have been shown to lower hospital mortality rates8 and improve delirium symptoms even before underlying medical etiologies are treated.5 Haloperidol is the mainstay of delirium treatment.8 Compared with atypical antipsychotics in delirium treatment, haloperidol doses <3.5 mg/d have not been associated with an increase in extrapyramidal symptoms (EPS).9

Although not devoid of side effects, atypical antipsychotics are an alternative to haloperidol.8,10 This article briefly summarizes the current evidence on the use of atypicals for treating delirium.

Table 1

Delirium: Diagnostic criteria

Delirium describes a group of related disorders with variable clinical presentations and differing causation. Regardless of the etiology, all types of delirium share a set of common symptoms that include:

Disturbances of consciousness and attention

Changes in cognition such as memory deficit, language disturbance, or disorientation

Perceptual disturbances not better accounted for by dementia

Abrupt onset (usually hours to days)

Fluctuating symptoms throughout the course of the day

Source: Adapted from reference 6

CASE CONTINUED: IM ziprasidone

After reassessing our treatment options, we prescribe ziprasidone, 10 mg IM twice a day, and an additional 10 mg IM every 12 hours as needed. Ziprasidone’s minimal anticholinergic and sedative effects3 seem favorable for Ms. B’s delirium; however, this medication has several drawbacks, including IM administration, greater expense compared with intravenous haloperidol, and risk of adverse cardiac affects, specifically prolonged corrected QT (QTc) interval.11 Bioavailability of oral ziprasidone is markedly less than the IM preparation (~60% vs 100%, respectively), and oral bioavailability decreases to approximately 30% when taken without food.12 Given Ms. B’s her current mental state, we feel that IM ziprasidone is a more reliable means to achieve therapeutic efficacy.13

With respect to cardiac concerns, we evaluate Ms. B’s predisposing and precipitating risk factors.11 Family members confirm that she had no cardiac history. We obtain baseline ECGs and continually monitor her QTc interval, which remained at <500 msec during ziprasidone treatment.

Ms. B tolerates ziprasidone and we note modest improvement in her mental status after 2 days of treatment; her vigilant-A portion of the CAM-ICU improves, but she still screens positive for delirium. During the next week Ms. B develops several medical comorbidities, including ileus, urinary tract infection, and methicillin-resistant Staphylococcus aureus infection. Despite these complications her mental status continues to improve. Within 6 days, Ms. B’s attention and cognition improve dramatically. She is oriented and able to engage in medical decision-making, and she screens negative for delirium on the CAM-ICU. We begin to assess her for psychiatric disorders that may have contributed to her hospitalization.

Evidence for antipsychotics

Haloperidol has been the antipsychotic of choice for treating delirium symptoms. It is recommended by the Society of Critical Care Medicine7 and is regarded as safe, cost-effective, and efficacious for delirium5 despite a risk of dose-related EPS and potential cardiac conduction alterations.5,14

Risperidone is not indicated for treating delirium but is one of the most extensively studied atypical antipsychotic alternatives to haloperidol. Evidence consisting primarily of case reports has illustrated the potential efficacy of risperidone in treating delirium (Table 2).10,15-19

In 2004, Parellada et al17 observed significant mean improvements in all measures (Delirium Rating Scale [DRS], Mini-Mental State Exam [MMSE], positive subscale of the Positive and Negative Syndrome Scale [PANSS-P], and Clinical Global Impressions scale [CGI]) in 64 delirium patients treated with risperidone. In a 2004 double-blind trial of 28 delirium patients randomly assigned to risperidone or haloperidol, risperidone was effective but not significantly more efficacious than low-dose haloperidol for acute delirium treatment.18

Advantages of using risperidone include its lack of anticholinergic effects. Potential side effects include dose-related EPS and weight gain, which were observed in patients with schizophrenia and other psychotic disorders and dementia-related behavioral disorders.20,21

Olanzapine. Much like risperidone, olanzapine’s use in delirium is relatively well described in the literature (Table 3).22-24 In a randomized, placebo-controlled study comparing olanzapine with haloperidol, 175 patients were treated for 7 days with olanzapine, haloperidol, or placebo. Olanzapine and haloperidol showed significantly greater DRS score improvement than placebo.24 There was no difference between olanzapine and haloperidol outcomes; however, olanzapine showed significant improvement by days 2 and 3 compared with haloperidol. Haloperidol was associated with a significantly higher rate of dystonia compared with olanzapine.

Olanzapine carries a risk of anticholinergic effects. This can be a drawback, especially in patients such as Ms. B whose delirium has an anticholinergic component. Olanzapine is available in an IM formulation, which can be an advantage when addressing agitation and medical comorbidities of delirium.

Quetiapine. Case reports have suggested quetiapine is effective for delirium (Table 4).10,25-27 In a prospective, open-label trial, Sasaki et al26 treated 12 delirium patients with a single bedtime dose of quetiapine. All patients achieved remission within several days of beginning quetiapine, and the drug was well tolerated with no detected EPS or excessive sedation.

In 2010 Devlin et al27 reported on the efficacy and safety of quetiapine in a prospective double-blind, placebo-controlled study of 36 adult ICU patients. Compared with those receiving placebo, patients taking quetiapine had a statistically significant shorter time to first resolution of delirium, reduced duration of delirium, and less agitation as measured by the Sedation-Agitation Scale. Mortality, ICU length of stay, and incidence of QTc prolongation did not differ, but patients treated with quetiapine were more likely to be discharged home or to rehabilitation and to have more somnolence. Quetiapine’s side effect profile includes a low occurrence of EPS, sedation, and dose-related anticholinergic effects.25

Ziprasidone. The literature on ziprasidone for delirium so far is limited to a few anecdotal case reports (Table 5).28-31 In 2002, Leso and Schwartz28 successfully used ziprasidone to treat delirium in a patient with human immunodeficiency virus and cryptococcal meningitis. Ziprasidone was chosen for its lack of sedating effects and low EPS risk. The patient experienced significant clearing of his delirium and lowering of his DRS score. Ziprasidone eventually was discontinued because a fluctuating QTc interval associated with comorbid electrolyte imbalances—a potential drawback to ziprasidone.

In the case of Ms. B, ziprasidone appeared to be efficacious; however, improvement in her medical condition, rather than ziprasidone treatment, is the most likely explanation for the resolution of her delirium symptoms.

Aripiprazole. Alao et al30 reported on 2 delirium patients treated with 30 mg and 15 mg aripiprazole; improvement was monitored using the MMSE and DRS (Table 5).28-31 In both cases, confusion, disorientation, and agitation improved within 7 days of treatment. In the first case, the patient’s MMSE score improved from 5 to 28 and his DRS score decreased from 28 to 6. The second patient’s MMSE score improved from 7 to 27 and her DRS score went from 18 to 6.

Straker et al31 reported on 14 delirium patients treated with aripiprazole. Twelve patients had a ≥50% reduction in DRS, Revised-98 scores, and 13 showed improvement on CGI scores. The rate of adverse side effects was low. Three patients had prolonged QTc interval, but no patients developed arrhythmia or discontinued aripiprazole.

Table 2

Risperidone for delirium: What the evidence says



Peak clinical response

Results/adverse effects (AEs)

Sipahimalani et al, 199715

N=2 (age 14 and 60). Initial dose: 1 mg/d; maintenance dose: 2 mg/d

10 to 14 days

MMSE score increased. AEs: extrapyramidal symptoms (dystonia and cogwheeling)

Schwartz et al, 200210

N=11 (age range 14 to 74). Mean dose: 1.59 ± 0.8 mg/d

5.1 ± 4.3 days

CGI score decreased.
No reported AEs

Horikawa et al, 200316

N=10 (mean age: 56.8; range: 22 to 81). Mean dose: 1.7 mg/d

7.1 days

DRS score decreased significantly in 80% of patients (P=.03) AEs: sleepiness (30%), mild drug-induced parkinsonism (10%)

Parellada et al, 200417

N=64 (mean age: 67.3 ± 11.4 years). Mean dose: 2.6 ± 1.7 mg/d

3 to 7 days

Effective in 90.6% of patients with significant decreases in DRS, PANSS-P, and CGI and increase in MMSE (P < .001). AEs: drowsiness (3.1%), nausea (1.6%)

Hans et al, 200418

N=12 (mean age: 65.6). Mean dose: 1.02 mg/d

4 to 7 days

MDAS scores decreased significantly (P < .05).
No reported AEs

Bourgeois et al, 200519

N=1 (age 57). Initial dose: 8 mg/d; maintenance dose: 2 mg/d

9 days

MMSE score increased.
No reported AEs

CGI: Clinical Global Impressions scale; DRS: Delirium Rating Scale; MDAS: Memorial Delirium Assessment Scale; MMSE: Mini-Mental State Exam; PANSS-P: positive subscale of the Positive and Negative Syndrome Scale

Table 3

Olanzapine may have a role in treating delirium symptoms



Peak clinical response

Results/adverse effects (AEs)

Sipahimalani et al, 199822

N=11 (mean age: 63.5 ± 23.2 years). Mean dose: 8.2 ± 3.4 mg/d

6.8 ± 3.5 days

Marked decrease (>50%) in DRS score for 5 patients.
No reported AEs

Breitbart et al, 200223

N=79 (mean age: 60.6 ± 17.3 years; range: 19 to 89). Initial dose: 3 ± 0.14 mg/d; mean dose: 4.6 to 6.3 mg/d

2 to 7 days

MDAS decreased significantly (P < .001), with 76% of patients’ delirium reaching resolution (MDAS ≤10). AEs: sedation (30%)

Hu et al, 200424

N=74 (mean age: 74). Mean dose: 1.25 to 2 mg/d

2.78 ± 1.85 days

DRS score decreased significantly (P < .01) in 72.2% of patients. AEs: drowsiness (18.9%), dystonia (2.7%), dry mouth (2.7%)

DRS: Delirium Rating Scale; MDAS: Memorial Delirium Assessment Scale

Table 4

Evidence suggests quetiapine could reduce delirium symptoms



Peak clinical response

Results/adverse effects (AEs)

Schwartz et al, 200210

N=11 (age range: 19 to 91). Mean dose: 211.4 mg/d

6.5 days

Decrease in DRS score (>50% reduction in global delirium symptoms) for 10 patients. AEs: sedation

Al-Samarrai et al, 200325

N=2 (age 50 and 52). Mean dose: 200 to 400 mg/d

2 to 4 days

No specific rating scale used but clinical reduction in agitation and improvement in cognition were reported. AEs: drowsiness

Sasaki et al, 200326

N=12 (mean age: 67.3 ± 14.8 years). Mean dose: 44.9 ± 31.0 mg/d

4.8 ± 3.5 days

Decrease in DRS score and remission of delirium for all patients. Significant increase in MMSE (P=.0256).
No reported AEs

Devlin et al, 201027

N=18 (adult ICU patients). Initial dose: 100 mg/d

36 to 87 hours

Significantly shorter time to first resolution of delirium and duration of delirium compared with placebo. AEs: somnolence

DRS: Delirium Rating Scale; ICU: intensive care unit; MMSE: Mini-Mental State Exam

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