The purpose of the study is to determine whether dexmedetomidine is a more effective medication than haloperidol in the treatment of agitation and delirium in patients receiving mechanical ventilation in an intensive care unit. Haloperidol is a medication conventionally used for this purpose.

The investigators will study only patients who have recovered from their illness to the point that, were it not for agitation and delirium, they would no longer require mechanical ventilation.

The investigators hypothesize that patients receiving dexmedetomidine will be able to discontinue mechanical ventilation earlier than those receiving haloperidol.

Up to 80% of patients undergoing intensive care have delirium. Early in the ICU stay, delirium and agitation are usually prevented using analgesic and sedative drugs which essentially render the patient unconscious. This is appropriate in the context of aggressive treatment of pathophysiological instability, which often requires multiple painful procedures. However, after the underlying pathophysiological problem has resolved, patients sometimes remain delirious and agitated. This often requires ongoing heavy sedation, which in turn necessitates continued mechanical ventilation, and can worsen the (temporarily masked) delirium. Prolonged mechanical ventilation increases the risk of ventilator associated pneumonia and other life threatening complications.

The drug most commonly used to treat delirium is haloperidol, which reduces hallucinations and unstructured thought patterns, but also reduces the interaction with the environment. Haloperidol has significant side effects, including extrapyramidal reactions (in 1-10% of patients), neuroleptic malignant syndrome (in which it is the cause in 50% of cases), and prolonged QT syndrome (which can precipitate fatal arrhythmias).

An ideal sedative agent in this context would have fewer side effects, relieve agitation without causing excessive sedation, and be easily titrated. An analgesic action might allow less opioid use, also lessening delirium. Early studies in other contexts suggest dexmedetomidine has all these properties.

The investigators hypothesise that patients with ICU-associated delirium after the resolution of their underlying pathological process who receive dexmedetomidine will be able to be extubated earlier than those who receive haloperidol.

Inclusion Criteria:

• Patients will be eligible for the study if, in the opinion of the treating clinician, they continue to require mechanical ventilation only because their degree of agitation requires such a high dose of sedative medication (midazolam or propofol, the only commonly used specific sedatives in our unit) that extubation is not possible.

Exclusion Criteria:

• Patients who could not be extubated even if delirium or agitation were corrected. This will include:

  • Patients receiving high dose opioid analgesia (>20 m/morphine/day)
  • Patients shortly to return to the operating theatre
  • Patients undergoing repeated invasive procedures, in whom it is desirable to maintain deep sedation.
  • Patients likely to require ongoing airway protection or control, or ventilatory support (for example, spinal patients with an inadequate vital capacity)

• Known allergy to haloperidol or alpha2 agonists