Subscribe to eTOC

Remodeling ICU Care: Trials Point Away from Antipsychotics for Delirium, Suggesting an Alternative Approach for Care



EFFECTS OF HALOPERIDOL, ziprasidone, and placebo on 90-day survival

A new study finding that two antipsychotics were no more effective than placebo for delirium in critical illness spurs discussion about alternative methods for managing that care.

Neither haloperidol nor ziprasidone — both antipsychotics — significantly alters the duration of delirium in critical illness, according to a new clinical trial published online October 22 in the New England Journal of Medicine.

The trial randomized 566 patients with acute respiratory failure or shock and hypoactive or hyperactive delirium to receive intravenous boluses of haloperidol (maximum dose, 20 mg daily), ziprasidone (maximum dose, 40 mg daily), or placebo. The primary endpoint was days alive without delirium or coma (during the 14-day intervention period).

The study found that the adjusted median number of days alive without delirium or coma was 8.5 in the placebo group, compared with 7.9 in the haloperidol group and 8.7 in the in the haloperidol group and 8.7 in the ziprasidone group.

The frequency of excessive sedation — a safety endpoint — also did not differ significantly between the trial groups.

This was the first randomized phase 3 trial to compare these two agents against placebo, said lead author Wes Ely, MD, the Grant W. Liddle endowed chair of medicine at Vanderbilt University School of Medicine, who with colleagues developed the Confusion Assessment Method for the intensive care unit (ICU) — the primary tool by which delirium and health-related quality-of-life outcomes are measured in ICU-based trials and clinically at the bedside in ICUs worldwide.

“We've been using these drugs in the ICU for the treatment of delirium for 40 years, and this is the first time there has been trial like this,” said Dr. Ely, who established Vanderbilt's Center for Critical Illness, Brain dysfunction and Survivorship.

Previous studies had also pointed toward the ineffectiveness of these agents in managing post-operative delirium, but those trials had either involved smaller pilot groups or lower doses of the drugs, Dr. Ely pointed out.

“The way these people use these drugs in the ICU is IV high dose,” said Dr. Ely. “So the response to these findings in the past has been, ‘That's not the way I use it.’ Our trial now shows that even at these much higher doses, they're still not achieving the desired outcome of shortening delirium duration.”

“One possible reason that we found no evidence that the use of haloperidol or ziprasidone resulted in fewer days with delirium or coma than placebo is that the mechanism of brain dysfunction that is considered to be targeted by antipsychotic medications — increased dopamine signaling — may not play a major role in the pathogenesis of delirium during critical illness,” the study authors wrote. “Another possible reason is that heterogeneous mechanisms may be responsible for delirium in critical illness.”


S. Andrew Josephson, MD, FAAN, chair of neurology at the University of California, San Francisco (UCSF), who was not involved with the study, said the findings offer “yet another call to neurologists and neurointensivists that we need to stop the practice” of employing antipsychotics for the control of delirium in critical illness.

Dr. Josephson, director and founder of UCSF's neurohospitalist program, who has served as medical director of inpatient neurology for more than a decade, said: “These [antipsychotic] medications are used widely based on clinical impressions, yet the evidence has not supported this practice to date. These medications are not without risk. They are associated with an increased risk of death and cardiovascular events in the elderly, and without any evidence for their efficacy, they should be avoided.”


DR. WES ELY: “Weve been using these drugs in the ICU for the treatment of delirium for 40 years, and this is the first time there has been trial like this.”

Thomas P. Bleck, MD, FAAN, a neurointensivist at Rush University Medical Center, wrote, however, in an accompanying editorial that the finding that the placebo was just as effective as active rescue medications may be related to the fact that the majority of patients in the trial had hypoactive delirium, for which the drugs may not have an effect. “It would be interesting to know whether hyperactive patients are less likely to injure themselves (e.g., by unplanned endotracheal extubation) when given active drug as a rescue agent,” he wrote.

Dr. Bleck, professor of neurological sciences, neurosurgery, internal medicine (in the division of pulmonary and critical care medicine, and the section of infectious diseases) and anesthesiology, added: “I would still consider using dopamine antagonists in patients at imminent risk of these types of injurious behaviors, but I would have less confidence in their benefits than I had in the past.”

Drs. Josephson and Ely both acknowledged that there are some circumstances when antipsychotics might still appropriately be used in critical care patients with delirium. “For example, I would use an antipsychotic in the ICU if it were a single scenario where a patient was posing a direct threat of harm to self or staff, but generally these should be rare interventions,” Dr. Josephson said.

Another scenario in which antipsychotics may be considered, Dr. Ely said, is for patients with heart failure, emphysema, or chronic obstructive pulmonary disease, who need a bilevel positive airway pressure mask. “We usually have to give sedation for them to tolerate it, and most sedation will slow the breathing rate down to the point of needing mechanical ventilation in a patient with such comorbid conditions,” he said. “Haloperidol doesn't suppress the respiratory drive, so it might be used in such a situation. But [in general], the days of writing a prescription for an antipsychotic to treat delirium should end.”

“Delirium is a very common condition, and one for which we do not have great pharmacologic treatments,” said Dr. Josephson. “Our most effective therapies focus on modifying the environment and care plan. We need to concentrate on those well-proven interventions, which have been documented to help this large population of patients who experience delirium.”


Dr. Ely's group suggests embracing a care rubric they refer to as the ABCDEF bundle to improve pain managements and reduce delirium and the long-term consequences for adult patients in the ICU. [See “The ABCDEF Bundle for ICU Management” and for more information.]

Among strategies, they suggest that clinicians assess, prevent, and manage pain; conduct both spontaneous awakening trials and spontaneous breathing trials, setting a time each day to stop sedative medications and orient the patient to time and day in an effort the free the patient from a ventilator; choose the safest analgesic and sedative agents, based on the evidence, to use or avoid in a given patient; encourage early mobility and exercise; and establish open communication with the family of the patient to engage them in the patient's recovery and provide them with clear expectations about the trajectory of recovery and support resources.

Dr. Ely's team recently published findings from their “ICU Liberation Collaborative” in Critical Care Medicine. In this prospective, multicenter cohort study involving 15,226 adults who were in the ICU at least one day, these ABCDEF strategies were associated with a lower likelihood of seven outcomes: hospital death within seven days, next-day mechanical ventilation, coma, delirium, physical restraint use, ICU readmission, and discharge to a facility other than home.

“We need to focus on interventions that actually matter to patients: waking them up, stopping medicine, involving family,” Dr. Ely said.

The ABCDEF bundle is remodeling ICU care, Dr. Ely said. “The old way involved keeping the patient in bed, tied down, and sedated, which is so patronizing and undignified. What this bundle does is to treat the patient with dignity and love, get them woken up and out of bed. That's where we need to go with critical care.”

Dr. Josephson contends that neurologists should play a key role in this work. “This is another opportunity for neurologists and neurointensivists to really embrace delirium as a disorder we should be treating. While neurologists are sometimes called in, the vast majority of people with delirium are not cared for by neurologists, and many physicians practice in institutions where neurologists would not be involved in care of these patients,” he said.

“Delirium in the ICU should be conceptualized as ‘brain failure’ (similar to the way kidney or heart failure is thought of as organ failure),” Dr. Josephson said. “We as neurologic experts should try to own this disease and concentrate our efforts in research and clinical care to better treat these patients. We need to embrace this disorder as ours and start working with our hospital systems to be sure we're delivering care that is commensurate with the current evidence.”


A: Assess, prevent, and manage Pain, using validated tools that can be used in every patient, every day.

B: Both spontaneous awakening trials and spontaneous breathing trials, providing powerful medications when needed but stopping them when unnecessary to avoid over-use and unwanted side effects.

C: Choice of analgesia and sedation, using published literature to assess the safest agents to use and ones to avoid in a given patient.

D: Assess, prevent, and manage delirium, using validated tools.

E: Early mobility and exercise. Early mobility decreases days of delirium, days on mechanical ventilation, and ICU and hospital length of stay, while improving functional outcomes.

F: Family engagement and empowerment. Providers should establish open communication with the patient's support system, engage them in the patient's recovery, and provide them with clear expectations about the trajectory for recovery and support resources.


• Girard TD, Exline MC, Carson SS, et al. Haloperidol and ziprasidone for treatment of delirium in critical illness N Engl J Med 2018; Epub 2018 Oct 22.
    • Bleck TP. Editorial: Dopamine antagonists in ICU delirium N Engl J Med 2018; Epub 2018 Oct 22.
      • Pun BT, Balas MC, Barnes-Daly MA, et al. Caring for critically ill patients with ABCDEF bundle: Results of the ICU Liberation Collaborative in over 15,000 adults Crit Care Med 2018; Epub 2018 Oct 18.