MIND-USA

Published

Girard TD, et al. "Haloperidol and ziprasidone for treatment of delirium in critical illness". The New England Journal of Medicine. 2018. ePub 2018-10-22:1-12.
PubMed • Full text • PDF

Clinical Question

Among critically ill patients with hypoactive or hyperactive delirium, does administration of the antipsychotic medications haloperidol or ziprasidone improve the number of days alive without delirium or coma when compared to placebo?

Bottom Line

Among critically ill patients with hypoactive or hyperactive delirium, administration of the antipsychotic medications haloperidol or ziprasidone did not improve the number of days alive without delirium or coma when compared to placebo.

Major Points

The incidence of delirium within the ICU has been cited to be 30% however can reach much higher incidences in certain ICU populations particularly in conjunction with organ failure or mechanical ventilation. Delirium, similar to non-ICU patients, is associated with increased mortality, prolonged mechanical ventilation, and prolonged ICU stays. Many trials to date have looked at the use of antipsychotic therapy in both prevention of delirium as well as reducing the length of delirium.HOPE-ICU (2013) tested haloperidol, a typical antipsychotic, in preventing delirium without evidence of shortening duration or incidence of delirium. A pilot study by Devlin et al. (2010) identified the potential for the atypical antipsychotic, quetiapine, to reduce duration of delirium and increase rates of hospital discharge when added to haloperidol. Given these varied findings, the role of antipsychotic therapy in management of ICU-related delirium was unclear.

Published in 2018, the Modifying the Impact of ICU-Associated Neurological Dysfunction–USA (MIND-USA) trial compared haloperidol (brand name Haldol, typical antipsychotic) and ziprasidone (brand name Geodon, atypical antipsychotic) against placebo therapy for the treatment of delirium over a 14-day period. This study was the first to directly compare a typical antipsychotic medication with an atypical antipsychotic medication and a placebo-controlled group in delirium treatment. The study had a heavy predominance of hypoactive delirium patients (90%). There was no benefit seen in the primary outcome of days alive without delirium seen in either antipsychotic therapy group compared to placebo. Subgroup analysis did not reveal a decreased requirement of sedation administration despite antipsychotic therapy.

The major limitation in this study is the high proportion of participants with hypoactive delirium, who may have been less likely to receive a benefit from the study medications. The lack of consistent efficacy of antipsychotic medication in ICU-related delirium may be attributed to a poorly understood mechanism of delirium in critically ill patients who do not simply have an excess of neurological dopamine, which would respond to traditional dopamine-blockade therapy. Despite the negative result, it is feasible that antipsychotics may remain beneficial for particular groups of ICU patients who are either acutely agitated or suffering from delirium-related hallucinations or delusions.

A related 2022 AID-ICU trial compared haloperidol vs. placebo in 1,000 ICU patients.

Guidelines

As of November 2018, no guidelines have been published that reflect the results of this trial.

Design

• Multicenter, double-blind, randomized, placebo-controlled trial
• N=566
  • Haloperidol (n=192)
  • Ziprasidone (n=190)
  • Placebo (n=184)
• Setting: 16 hospitals in the US
• Enrollment: 2011-2017
• Follow-up: 14-day intervention period with 90 days of follow-up
• Analysis: Intention-to-treat
• Primary outcome: Days alive without delirium or coma

Population

Inclusion Criteria

• Age >18 years
• Medical/surgical ICU admission
• Requiring either NIPPV/Mechanical ventilation, vasopressors, or intra-aortic balloon pump
• Objectively confirmed delirium using the CAM-ICU

Exclusion Criteria

• Severe cognitive impairment at baseline
• High risk of medication side effects (Pregnant, breast-feeding, history of torsades, QTc >550ms at baseline, history of neuroleptic malignant syndrome, or allergy to trial drugs)
• Pre-existing treatment with antipsychotic medication
• Allergy to a study medication
• Moribund state
• Rapidly resolving organ failure
• Barriers to follow-up
• Incarcerated

Baseline Characteristics

Baseline characteristics from haloperidol group

• Demographics: Age 61 years, female sex 44%, white race 85%, black race 12%
• Delirium type: Hyperactive 10%, hypoactive 90%
• IQCODE score (to determine pre-existing dementia, out of 5, higher is worse): 3.0
• Admission diagnosis: ARDS 23%, sepsis 22%, 'airway protection' 24%, COPD/asthma/other pulmonary disorder 10%, surgery 7%, HF MI or arrhythmia 3%, cirrhosis or liver failure 2%, seizure or neurological disease 2%, other 7%
• Shock at admission: 30%
• Surgical ICU 27%
• Median APACHE II score: 28.5
• Median Charlson comorbidity score: 2
• Ventilation: Invasive 93%, noninvasive 4%
• Median (IQR) ICU days prior to admission: 2.4 (1.5-3.4)

Interventions

• Randomization to an arm:
  • Haloperidol: Haloperidol 1.25mg (>70 years old) or 2.5mg (<70 years old) IV q12h prn for 14 days or until discharged from ICU. Uptitrated if persistent delirium up to 20mg per day, downtitrated if no delirium.
  • Ziprasidone: Ziprasidone 2.5mg (>70 years old) or 5mg (<70 years old) IV q12h prn for 14 days or until discharged from ICU. Uptitrated if persistent delirium up to 40mg per day, downtitrated if no delirium.
  • Placebo
• Patients were assessed every 12 hours for delirium using CAM-ICU and Richmond Agitation-Sedation Scale.
  • If no delirium for 2 assessments, then half dose of antipsychotic.
  • If no delirium for 4 assessments, then hold antipsychotic therapy.
  • Stop if the patient develops torsades, NMS, DRESS, or new structural brain disease resulting in decreased level of consciousness

Outcomes

Placebo vs. haloperidol vs. ziprasidone

Primary Outcomes

Median days Alive without Delirium or Coma

Adjusted days: 8.5 (95% CI 5.6, 9.9) vs. 7.9 (4.4, 9.6) vs. 8.7 (5.9, 10.0)

The primary outcome was considered the adjusted analysis. Other day counts are unadjusted. All ORs are adjusted. These were adjusted for several covariates, detailed on page 8 of the supplementary appendix.

Unadjusted days: 7 (IQR 0-11) vs. 8 (0-11) vs. 8 (2-11)

Overall P=0.26

Haloperidol vs. placebo: OR 0.88 (95% CI 0.64-1.21)

Ziprasidone vs. placebo: OR 1.04 (95% CI 0.73-1.48)

Secondary Outcomes

For simplicity, IQRs for the remaining results are in Table 2 of the manuscript. Days are unadjusted and ORs are adjusted as described above.

Median days with delirium

4 vs. 4 vs. 4

Haloperidol vs. placebo: OR 1.12 (95% CI 0.86-1.46)

Ziprasidone vs. placebo: OR 1.02 (95% CI 0.69-1.51)

Median days with hyperactive delirium

0 vs. 0 vs. 0

Haloperidol vs. placebo: OR 1.18 (95% CI 0.86-1.61)

Ziprasidone vs. placebo: OR 1.09 (95% CI 0.70-1.70)

Median days to freedom from mechanical ventilation

3 vs. 2 vs. 3

Haloperidol vs. placebo: OR 0.92 (95% CI 0.71-1.19)

Ziprasidone vs. placebo: OR 0.96 (95% CI 0.74-1.15)

Median days to ICU discharge

5 vs. 5 vs. 6

Haloperidol vs. placebo: OR 0.95 (95% CI 0.81-1.12)

Ziprasidone vs. placebo: OR 1.02 (95% CI 0.88-1.17)

30-day mortality

27% vs. 26% vs. 28%

Haloperidol vs. placebo: OR 1.03 (95% CI 0.73-1.46)

Ziprasidone vs. placebo: OR 1.07 (95% CI 0.77-1.47)

90-day mortality

34% vs. 38% vs. 34%

Haloperidol vs. placebo: OR 1.17 (95% CI 0.99-1.40)

Ziprasidone vs. placebo: OR 1.02 (95% CI 0.79-1.30)

Adverse Events

Oversedation

11% vs. 11% vs. 12%

Prolonged QTc

1% vs. 1% vs. 2%

Torsades des pointes

0% vs. 0% vs. 0%

NMS

0% vs. 0% vs. 0%

Criticisms

• The high proportion with hypoactive delirium (90%) may have been less likely to benefit from antipsychotic medications compared to hyperactive delirium patients.
• The use of ziprasidone in the ICU population is not as widespread compared to the atypical antipsychotic, quetiapine.

Funding

National Institutes of Health and the VA Geriatric Research Education and Clinical Center