By James E. McFeely, MD

Medical Director, Critical Care Units, Alta Bates Summit Medical Center, Berkeley, CA

Dr. McFeely reports no financial relationships relevant to this field of study.

SYNOPSIS: In this randomized, placebo-controlled trial, neither haloperidol nor ziprasidone altered the duration of delirium when compared to placebo.

SOURCE: Girard TD, Exline MC, Carson SS, et al. Haloperidol and ziprasidone for treatment of delirium in critical illness. N Engl J Med 2018; Oct 22. doi: 10.1056/NEJMoa1808217. [Epub ahead of print].

This randomized, placebo-controlled trial was designed to test whether haloperidol or ziprasidone affected delirium in ICU patients with respiratory failure or shock. The study was carried out between 2011 and 2017 in 16 U.S. centers. In total, researchers screened 20,914 patients, resulting in 1,183 consenting patients, of whom 566 developed delirium and were studied. Delirium was identified using the Confusion Assessment Method for the ICU (CAM-ICU) assessment tool.1 The Richmond Agitation-Sedation scale (RASS) was used to define hypoactive (RASS < 0) and hyperactive (RASS > 0) delirium.2 Of the 566 patients randomized, 89% demonstrated hypoactive delirium. Approximately 190 patients were randomized to each treatment group (haloperidol, ziprasidone, or placebo). Once delirium was identified, treatments were initiated at small doses (e.g., 2.5 mg IV haloperidol every 12 hours for patients < 70 years of age) and were increased gradually until delirium was no longer present or maximum doses were achieved. The mean doses administered were modest (e.g., haloperidol 11 mg ± 5 mg), and investigators temporarily withheld the study drug at least once during the trial for approximately half the patients. The mean length of exposure to the drug was four days. The primary endpoint was days alive without delirium or coma during the 14-day intervention. Secondary endpoints included 30- and 90-day survival, days to liberation from ventilator, and length of stay in the ICU and the hospital. Patient characteristics were similar across the three treatment groups. No treatment effect was seen for any primary or secondary endpoint with all odds ratios (including unity). There was no difference in side effects between the treatment groups. The investigators concluded that haloperidol and ziprasidone did not affect duration of delirium among patients with shock or respiratory failure.


Girard et al are to be commended for tackling such an important project. The difficulty in completing this trial is evidenced by the screening of more than 20,000 individuals to consent 1,183 — only to enroll just 566 patients. This low percentage enrollment affects the generalizability of this result, particularly for the subgroup of patients (all 57 of them) defined as having hyperactive delirium. The trial design was good, with appropriate entry and exclusion criteria, safety screening, duration of treatment, and appropriateness of endpoints. There was surprisingly good adherence (88%) to the Awakening and Breathing Coordination, Delirium Monitoring and Management, and Early Mobility (ABCDE) bundle. The use of the CAM-ICU screening tool is standard of care for identification of delirium. While the RASS scale is used widely for assessment of agitation levels, it was not developed to differentiate subtypes of delirium.

The study results do not support using antipsychotics in nonagitated delirious patients routinely. Study limitations include the low percentage enrollment and the relatively low average dose of antipsychotics administered. Would a more rapid increase in dose have produced an effect? Intensivists are more likely to treat delirium with antipsychotics when the patient is agitated. The data from the 57 patients described as having hyperactive delirium are not robust enough to support any conclusions regarding the role of drugs in this subpopulation. Girard et al did not address the use of these drugs in the many subgroups excluded from the trial (e.g., preexisting cognitive impairment or high risk of medical complications from the drugs).

The results support minimizing the use of haloperidol and ziprasidone in delirious patients who are not agitated. Current best practice is to adhere to the ABCDE bundle,3 remove causative agents when possible, and continue antipsychotics (only if they appear effective and for the minimum time necessary). The jury is still out regarding how to manage agitated delirium. Hopefully, investigators can proceed to a trial of patients specifically with agitated delirium perhaps using a different study drug, such as quetiapine, that may be more effective.4


  1. Ely EW, Inouye SK, Bernard GR, et al. Delirium in mechanically ventilated patients: Validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA 2001;286:2703-2710.
  2. Sessler CN, Gosnell MS, Grap MJ, et al. The Richmond Agitation-Sedation Scale: Validity and reliability in adult intensive care unit patients. Am J Respir Crit Care Med 2002;166:1338-1344.
  3. Devlin JW, Skrobik Y, Gélinas C, et al. Executive summary: Clinical practice guidelines for the prevention and management of pain, agitation/sedation, delirium, immobility, and sleep disruption in adult patients in the ICU. Crit Care Med 2018;46:1532-1548.
  4. Devlin JW, Robert RJ, Fong JJ, et al. Efficacy and safety of quetiapine in critically ill patients with delirium: A prospective, multicenter, randomized, placebo-controlled pilot study. Crit Care Med 2010;38:419-427.