This is a report of a systematic literature search and review of data from prospective trials of pharmacological interventions for treatment and prevention of delirium. The authors included randomized and non-randomized double- or single-blind and open-label clinical trials of any pharmacologic agent for the prevention or treatment of delirium. Outcomes included safety and efficacy of interventions for prevention and treatment of delirium, and benefits of interventions for reduction of duration and severity of established delirium episodes. This review also sought to evaluate evidence for effectiveness of treatment in reaching specific goals of reduction of intensive care unit and hospital lengths of stay and reduction of mortality. Finally, the authors focused on a limited number of studies of effectiveness of treatment for various delirium subtypes.

The search included English language publications from 1980 to late 2013. Case reports, retrospective studies, or treatments involving substance use disorders were excluded. 45 studies were included in the review. Results were categorized as: 1) antipsychotics for prevention, 2) antipsychotic treatment for ongoing delirium, 3) cholinesterase inhibitors, 4) anesthetic techniques perioperative sedation, and 5) miscellaneous.

Six of seven antipsychotic trials for prevention of delirium in patients undergoing surgical procedures were included. Three studies involved the use of haloperidol. Two studies (N=537) demonstrated lower rates of postoperative delirium utilizing intravenous haloperidol, while one negative trial (N=430) evaluated oral haloperidol. Use of both oral risperidione (N=227) and olanzapine (N=400) demonstrated reduced levels of postoperative delirium versus control.

Four studies of dexmedetomidine sedation in ICU settings showed reduced rates of delirium and adequate sedation. One study involving single dosages of ketamine, and another single study using melatonin, showed enhanced efficacy in preventing delirium in general medical patients.

Three randomized, double-blind, placebo-controlled trials of antipsychotic treatment in patients with ongoing delirium demonstrated that haloperidol, ziprasidone, and quetiapine were not associated differentially with higher rates of delirium resolution. Effects on episode duration and severity were inconsistent. Quetiapine use was associated with a shorter duration of delirium in one study (N=36) and more rapid resolution in another (N=32).

Decreases in ICU or total hospital length of stay could not be demonstrated with use of haloperidol, ziprasidone, or quetiapine in treatment of ongoing delirium.

Shorter time to extubation for mechanically ventilated patients was demonstrated in several reports (N=202).

While prolongation of QTc interval is a potential concern with antipsychotics, this was the least reported adverse effect reported in trials. Data from four delirium trials demonstrated that low-dose intravenous bolus haloperidol, oral haloperidol, oral risperidone, ziprasidone, and quetiapine were no more likely than placebo to produce a QTc change greater than 60 ms over baseline or above 500 ms. Continuous intravenous infusion of haloperidol was associated with a significantly longer mean QTc interval than dexmedetomidine.

Five studies using cholinesterase inhibitors (N=277) failed to show efficacy in prevention of delirium in surgical patients..

A variety of interventions for enhanced perioperative analgesia demonstrated effectiveness for use of nerve blocks where the surgical site was distal to the block. Three trials of gabapentin involving 380 patients showed modest reductions of postoperative delirium. One trial involving use of opioids to lessen delirium by treating postoperative pain was negative.

Three of the 24 antipsychotic trials analyzed treatment effects based on delirium subtypes. Delirium resolution occurred in a greater proportion of hypoactive than hyperactive subtypes with aripiperazole treatment (100% vs. 58.3%). Resolution rates in hypoactive and hyperactive subtypes were similar in one haloperidol trial (77.8% vs. 75%). Lower resolution rates in hypoactive than hyperactive subtypes were reported with use of quetiapine (48% vs. 83%).

This report of a meta-analysis of randomized controlled trials of three antipsychotic agents attempts to determine the effect of pretreatment or prophylaxis on the development, severity, and length of delirium following common surgical procedures.

Four authors conducted independent database searches of English language papers, published between January 1950 and April 2012, relating to prophylaxis of delirium using antipsychotic medication. Search terms included delirium, encephalopathy, ICU psychosis, prevention, antipsychotic, and prophylaxis. Inclusion criteria included use of antipsychotic treatment of delirium, randomized and placebo-controlled, prevention and use of a validated method of delirium diagnosis.

All returned the identical five articles (inter-rater reliability of 100%) that included a total of 1491 participants. Analysis of the pooled data suggested a 50% reduction in the relative risk for development of delirium among those receiving an antipsychotic. Other related observations were for a trend lacking statistical significance of decreased severity and length of delirium, decreased length of stay, and likelihood that the patient returned home rather than to institutional aftercare. For the primary outcome, prevention of delirium, the numbers needed to treat ranged between 4 and 12. Three antipsychotics, haloperidol, risperidone and olanzapine, were utilized. No differences in relative efficacy were detected. There were no adverse effects attributable to antipsychotic use in any of the 739 patients receiving them.

Conclusion: Brief, limited use of antipsychotic prophylaxis in the elderly who are at risk for delirium may reduce the incidence of delirium and reduce its associated mortality, disease burden, length of hospital stays, and costs.

The American College of Critical Care Medicine assembled a 20-person multidisciplinary task force with expertise in guideline development charged with updating evidence-based protocols for treatment of pain, agitation, and delirium in critically ill patients. Group consensus was reached using the Delphi Method after review of relevant literature using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) method through December 2010. The following recommendations were advanced relating to delirium.

1. Delirium is associated with increased mortality, increased ICU and hospital length of stay, and post-ICU cognitive impairment in adult ICU patients.
2. Routine monitoring of all ICU patients for delirium using the Confusional Assessment Method (CAM-ICU) is recommended and feasible in everyday practice.
3. In mechanically ventilated adult ICU patients at risk for developing delirium, dexmedetomidine infusions may be associated with a lower prevalence of delirium than with benzodiazepine sedation.
4. Evidence does not support a recommendation for the use of haloperidol, second-generation antipsychotics, or dexmedetomidine to prevent delirium in the ICU.
5. There is no evidence that use of haloperidol reduces length of delirium in ICU patients. Second-generation antipsychotics may reduce duration of delirium in critically ill patients.
6. Use of antipsychotics in patients with a history of arrhythmia or elevation of QTc interval over baseline is not recommended in the ICU.
7. Suggestion that adult ICU patients with delirium unrelated to alcohol or benzodiazepine withdrawal receive dexmedetomidine rather than other sedating agents to reduce duration of delirium.

Commentary:

Since publication of the 2012 Delirium monograph, several systematic reviews have added evidence informing practice for consultation-liaison psychiatrists. The monograph mentioned, but offered no recommendation for, use of pharmacological interventions to reduce frequency of onset of delirium in patients at high risk (elderly, cognitive impaired, prior history of delirium).

Additional data and analysis of previously available data suggests that use of haloperidol, olanzapine, or risperidone preoperatively may reduce the incidence and severity of delirium and reduce lengths of stay in patients undergoing common surgical procedures where delirium is a common complication. While there is no FDA-approved indication for use of antipsychotics in the prevention or treatment of delirium, accumulating evidence supports offering such treatment to reduce associated morbidities. Similar use in critically ill ICU patients, however, is not supported by existing evidence.

Use of oral haloperidol, quetiapine, and risperidone have not been associated with safety concerns in studies of treatment or prevention of delirium with the exception of intravenously infused haloperidol associated with increased QTc and arrhythmias.

Some light has been shed since the original monograph on use of second-generation antipsychotics used as interventions for delirium. Several such agents, risperidone, quetiapine, olanzaoine, ziprasidone, and aripiperazole, are equally useful in reducing symptoms of delirium. Efficacy has been demonstrated in a variety of delirium subtypes including hypoactive delirium.

The alpha 2 agonist dexmedetomidine was not mentioned in the 2012 APM Delirium Monograph. This agent may be useful in managing sedation in delirious, critically ill ICU patients having demonstrated advantages over benzodiazepine sedation. The potential for development of bradycardia or hypotension limits its use to patients with ongoing cardiac and blood pressure monitoring.

Structured screening for delirium using the Confusional Assessment Method (CAM-ICU) in critically ill ICU patients can improve early identification and indicated intervention for episodes of delirium.

Finally, further research is needed in identifying the advantages of various agents in the treatment of established delirium episodes. While data supports use of antipsychotics in prevention of delirium in patients undergoing common surgical procedures, more data is needed relating to actual reduction of morbidities, lengths of stay, postsurgical cognitive deficits, and costs of care.