Delirium in the Intensive Care Unit: Is Dexmedetomidine Effective?

Delirium presents as acute onset of confusion characterized by a change or fluctuation in baseline mental status or attention, disorganized thinking, or an altered level of consciousness. It is difficult to obtain a precise incidence rate for delirium in the ICU because of variance in screening tools used, types of ICU, and patients sampled. Two tools have been used to assess patients at risk for developing delirium. The Intensive Care Delirium Screening Checklist (ICDSC), based on Diagnostic and Statistical Manual of Mental Disorders criteria,¹³  is a simple 8-item checklist. An ICDSC score greater than 4 indicates the patient is at risk for developing delirium. The ICDSC tool and scoring is further described in Figure 1.¹⁴  The Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) is based on the original Confusion Assessment Method and was adapted by Ely et al¹⁵  for use in nonverbal patients receiving mechanical ventilation. The CAM-ICU tool is summarized in Figure 2.¹⁴ 

Roberts et al¹⁶  completed a 6-month multisite, prospective epidemiological study in Australia. They used the ICDSC to determine the incidence of delirium in 185 ICU patients, 142 of whom (76.7%) received mechanical ventilation. Eighty-four patients (45%) were at risk for developing delirium, on the basis of ICDSC scores of greater than 4.¹⁶  The Delirium Epidemiology in Critical Care study, a 1-day point-prevalence multicenter assessment, was completed in 104 ICUs in 11 countries in North and South America and Spain.⁴  A total of 975 patients were screened with the CAM-ICU tool. Patients who were deeply sedated, as defined by the Richmond Agitation-Sedation Scale, were not included in the study because their sedation level did not allow for accurate assessment of delirium. A final sample of 232 patients who did not require mechanical ventilation were enrolled, and 75 (32.3%) were identified with delirium on the basis of their CAM-ICU scores.⁴  As illustrated by these 2 studies, the incidence of delirium varies. However, according the literature reviewed, an estimated 30% of ICU patients experience delirium.^3,4,7 

The reticular formation is a complex network of interconnected nuclei situated in the brainstem. Part of the function of the reticular formation is to maintain wakefulness and regulate vital reflexes, such as cardiovascular function and respiration. The reticular formation projects into the thalamus, basal ganglia, and specific areas of the cerebral cortex and limbic system. The projections of the reticular formation into the cerebral cortex and limbic areas are referred to as the reticular activating system. The precise cause of delirium is unknown, but neurotransmitter imbalance, increased levels of inflammatory mediators, and impaired oxidative metabolism may disrupt the reticular activating system.^1,17 

Imbalances in neurotransmitters result in neuronal instability and unpredictable neurotransmissions, which are manifested as acute confusion.^17,18  Inflammatory cytokines are thought to contribute to the development of delirium by altering blood-brain barrier permeability and further affecting neurotransmission.¹  Thus, delirium is caused by alterations in neurotransmitter production and activity.^18,19 

A recent study of neuroimaging with computed tomography has provided a better understanding of the effects of delirium on the brain.²⁰  The investigators compared computed tomography scans of patients during and after episodes of delirium and found a reduction in overall cerebral blood flow during delirium. The blood flow was less in subcortical and occipital regions, suggesting that delirium is a widespread, rather than a localized, dysfunction.²⁰ 

Several factors contribute to the increased risk of developing delirium among all hospitalized patients. In the ICU, sedative medications or poorly managed pain can predispose patients to developing delirium. In a prospective cohort study, Inouye and Charpentier²¹  examined risk factors for delirium in 196 patients aged 70 years and older who were admitted to general medicine wards in Yale New Haven Hospital, Connecticut. The investigators found an increased risk of delirium due to physical restraints, addition of more than 3 medications in a 24-hour period, presence of a urinary catheter, and any iatrogenic event.²¹  The study did not include critically ill patients, but many of the same risk factors are frequently encountered in the ICU environment (see Table 1).^21,22 

Delirium has 3 presentations: hyperactive, hypoactive, and mixed delirium. Hyperactive delirium is associated with general disruptive behaviors such as shouting, hitting, and self-removal of catheters in association with extreme levels of agitation, emotional lability, and anxiety.^18,23  Peterson et al²³  found that 6 of 614 patients (1%) had hyperactive delirium and all cases occurred in patients younger than 65 years.

Hypoactive delirium is characterized by withdrawal, lethargy, apathy, and lack of responsiveness. Peterson et al²³  observed that the hypoactive type occurred in 267 of 614 ICU patients (43.5%) and more commonly presented in the older patients (51.8%; odds ratio, 3.0; 95% CI, 1.7-5.3). Hypoactive delirium is associated with a worse prognosis because it is less easily recognized and is often undertreated.^7,18 

In mixed delirium, patients fluctuate between hyperactive and hypoactive symptoms.¹⁹  Peterson et al²³  found that mixed delirium was the most common type of presentation in patients aged 65 years or older.

Short- and long-term consequences of delirium include prolonged mechanical ventilation, self-extubation, use of physical restraints, longer ICU and hospital stays, long-term memory and physical impairments, and increased mortality at 6 months.^18,22,24  Long-term memory impairment caused by delirium is a type of long-term cognitive impairment, a syndrome of significant and persistent cognitive deficits following a critical illness. The syndrome involves impairment of executive functioning, which includes planning, problem solving, inhibition, and control of behavior.²² 

The management of delirium begins with identifying patients at risk by using the CAM-ICU and ICDSC screening tools. Using the tools does not prevent the development of delirium, but ICU RNs can use these tools as part of routine nursing care to identify patients at risk.^22,24 

Ideally, patients at risk of delirium would be identified early and strategies to prevent the onset of delirium would be instituted. The ABCDEFS of delirium, summarized in Table 2, provide clinicians with simple preventive strategies to help reduce the likelihood that a patient will develop delirium.^1,6,19,22,25 

According to the Society of Critical Care Medicine (SCCM) clinical practice guidelines, using drug protocols to prevent delirium is not recommended because compelling evidence indicates that such protocols do not reduce the development or duration of delirium.¹⁹  Initial management of delirium is aimed at treating the underlying cause, reducing the duration of delirium, and preventing complications.²²  Management includes nonpharmacological and pharmacological approaches. Nonpharmacological management strategies for delirium are outlined in Table 3.^14,22,24 

No pharmacological interventions are currently approved by the FDA for the treatment of delirium. Medications used to treat delirium are haloperidol and atypical antipsychotics such as olanzapine, quetiapine, and risperidone.^2,22,26 

Antipsychotics are used to treat acute and chronic psychotic disorders, acute agitation, and bipolar disorder. First-generation antipsychotics are also referred to as neuroleptics, conventional antipsychotics, or typical antipsychotics. All first-generation antipsychotics are dopamine receptor antagonists that block the D₂ receptor. However the drugs have varied effects on a₁, histamine, and muscarinic receptors, resulting in different adverse effect profiles.^26,27  First-generation antipsychotics are subject to extensive metabolism through the cytochrome P450 enzyme system.

In the setting of liver impairment or drug interactions, the clearance of antipsychotics is reduced. The most common adverse effects associated with first-generation anti-psychotics include extrapyramidal symptoms, tardive dyskinesia, neuroleptic malignant syndrome, QT prolongation and torsades de pointes, sudden death, and an increased risk of mortality when used to treat psychiatric symptoms associated with dementia in older patients.^22,26 

Haloperidol is a first-generation antipsychotic that gained popularity in the 1990s for treatment of agitated patients in the ICU.²⁷  Since then, haloperidol has been used to manage delirium, but to date no evidence has shown that it reduces the duration of delirium.^2,19  Additionally, haloperidol is not FDA approved for the management of delirium.²⁷ 

Despite the lack of evidence and FDA approval, haloperidol continues to be used to treat delirium in patients in the ICU.^19,28  In a survey completed through Vanderbilt University in 2006, physicians, nurse practitioners, and physician assistants were asked about their behaviors and attitudes regarding sedation practices for patients in the ICU with delirium. A total of 1384 health care providers completed the survey, with 1190 respondents (86%) identifying haloperidol as the medication used for the management of delirium.²⁸ 

Haloperidol continues to be used to treat delirium in ICUs despite the lack of evidence for its use. The SCCM guidelines for managing pain, agitation, and delirium do not support the use of haloperidol to prevent or treat delirium.¹⁹  Wang and colleagues²⁹  suggested that haloperidol is the preferred agent for treating delirium because of its desirable characteristics for delirium management. For example, haloperidol has a rapid onset of action, has no active metabolites, does not require dosing adjustments for organ dysfunction, and can be given orally and parenterally.²⁹ 

Because of the lack of evidence, the efficacy of haloperidol and the optimal haloperidol regimen for treating delirium are unknown.³⁰  For patients with hyperactive delirium who do not have cardiac or electrolyte abnormalities, Maldonado¹⁷  recommended low-dose haloperidol (<20 mg/24 h). To help monitor therapy and allow for dose reductions in case of adverse effects, Maldonado¹⁷  suggested obtaining a baseline electrocardiogram and electrolyte levels and reviewing current medications before administering haloperidol. If the QTc interval increases to greater than 25% of baseline or greater than 500 milliseconds, the therapy should be stopped and an atypical antipsychotic, such as risperidone or quetiapine, should be administered.¹⁷  For patients with hypoactive delirium, Maldonado²²  recommended very low-dose haloperidol (0.25-1 mg/24 h).

Practitioners may have apprehension and concerns about using antipsychotic agents, given their established adverse effect profile.²²  Therefore, the rate of atypical anti-psychotic use in the symptomatic management of delirium has been increasing.^17,22  Compared with antipsychotics such as haloperidol, atypical antipsychotics have reduced affinity for dopamine receptors but have a range of affinities for serotonin, acetylcholine, and norepinephrine receptors in the central nervous system.²⁶ 

There is weak evidence that atypical antipsychotics reduce the duration of delirium, as demonstrated by a small prospective, randomized, double-blind, placebo-controlled study.³¹  In the study, 36 patients with delirium in 3 adult ICUs in Boston, Massachusetts; Portland, Maine; and Montreal, Quebec, Canada, were randomized to receive quetiapine 50 mg (n = 18) or placebo (n = 18) every 12 hours. Any patient could also receive haloperidol as required. The median duration for resolution of delirium in the quetiapine group was 1 day, which was significantly less than the 4.5 days in the placebo group (P = .001).³¹  However, the study had a small sample size, and further research is needed before the definitive use of quetiapine can be advocated.

Lonergan et al³²  completed a Cochrane review comparing haloperidol with the second-generation antipsychotics risperidone, olanzapine, and quetiapine for the treatment of delirium. Three studies met the selection criteria. The review found no significant differences between low-dose haloperidol (<3 mg/d) and olanzapine and risperidone for the management of delirium (odds ratio, 0.63; 95% CI, 0.29-1.38; P = .25). The review did not find a higher incidence of adverse effects with low-dose haloperidol. Compared with placebo, haloperidol did not reduce the incidence of delirium but did decrease the severity and duration of delirium in postoperative patients.³²  The conclusions should be considered within the context of the small sample sizes and limited scope of the included studies.

Although weak evidence supports the use of atypical antipsychotic agents in the management of delirium, one hindrance to their use is that no intravenous formulations are available.²⁶  Additionally, the adverse effects of atypical antipsychotics, specifically sedation, tachycardia, and orthostatic hypotension, may be detrimental to critically ill patients and may limit their use in this population.²⁶ 

Benzodiazepines contribute to the development of delirium through the mechanisms outlined in Table 4. In instances of delirium tremens and withdrawal syndromes, benzodiazepines are the preferred treatment of choice.¹  More recently, delirium development has been associated with the use of benzodiazepines in patients with alcohol withdrawal. Maldonado³³  completed a literature review to develop a benzodiazepine-sparing treatment protocol for patients with alcohol withdrawal and suggested that the use of benzodiazepines is an independent risk factor for the development of delirium. At this time, the SCCM guidelines still recommend the use of benzodiazepines for alcohol withdrawal; however, the results of the Maldonado review³³  may change clinical practice in the future.

The 2013 SCCM guidelines recommend using nonbenzodiazepine sedation strategies including agents such as propofol over sedation with benzodiazepines. Propofol is an intravenous sedative that binds to multiple central nervous system receptors, including γ-aminobutyric acid, muscarinic, and nicotinic receptors.¹⁹  Therefore, propofol has a similar mechanism of action as benzodiazepines, but there are certainly differences between the 2 agents.

Propofol is a highly lipid-soluble drug that is distributed into peripheral tissues and can rapidly cross the blood-brain barrier. It has a rapid onset (40 seconds to 3 minutes) and has anxiolytic and amnestic effects.³⁴  Propofol has high hepatic clearance and is rapidly cleared with short-term infusions. With increased duration of propofol infusions, steady-state volume of distribution increases, which leads to a longer half-life. Propofol commonly causes respiratory depression and hypotension. Rarely, hypertriglyceridemia, acute pancreatitis, myoclonus, and propofol infusion syndrome can occur.^19,34 

Propofol has been associated with fewer deliriogenic effects than benzodiazepines.²²  However, in 2 of 3 studies comparing propofol with dexmedetomidine, propofol was associated with a higher incidence of delirium.^5,35,36  Propofol may be a more appropriate agent than benzodiazepines, but dexmedetomidine may be preferable to propofol.²² 

Pharmacological management of delirium also involves reviewing medications that may be causing or potentiating the delirium. Maldonado²²  recommended conducting a medication review. Medications known to cause delirium or have a high anticholinergic potential should be discontinued, and a suitable alternative should be used.

In summary, research findings are equivocal with respect to the use of medications to prevent delirium, partly because of heterogeneous and small sample sizes. The ideal clinical situation is the prevention or at least the early detection of delirium and the use of nonpharmacological management strategies.²²  When pharmacological therapy is necessary, the ideal drug would relieve symptoms without causing excessive sedation, would have fewer adverse effects than haloperidol, and would not interact with other drugs. In addition, the ideal drug would have an analgesic effect, thereby sparing the need for opioids.³⁷  Dexmedetomidine has a medication profile with these characteristics.

Dexmedetomidine is an α₂-adrenoreceptor agonist that has sedative, analgesic, and anxiolytic properties.^11,38  α₂-Adrenoreceptors are found throughout the body in the central nervous system, peripheral nervous system, liver, pancreas, kidney, and eye.^2,38  In the brainstem, the locus coeruleus contains many α₂-adrenoreceptors and plays a key role in wakefulness and regulation of nociceptive neurotransmission.³⁸  Stimulation of α₂-adrenoreceptors in the brain and spinal cord has inhibitory effects leading to sedation, analgesia, hypotension, and bradycardia.³⁸  Dexmedetomidine acts as an agonist in the locus coeruleus and inhibits norepinephrine release, resulting in depression of alertness and sympathetic activity that manifests as sedation, hypotension, and bradycardia.¹¹ 

Dexmedetomidine is administered intravenously, has an onset of action of approximately 15 minutes, and reaches peak concentration after 1 hour of continuous infusion.¹¹  Dexmedetomidine has a half-life of 2 hours and follows first-order linear kinetics. It is metabolized almost completely in the liver to inactive metabolites. Patients with hepatic dysfunction may require lower doses, but renal dosing is not required.³⁹ 

The recommended loading dose of dexmedetomidine is 1 μg/kg over 10 minutes, followed by a continuous infusion of 0.2 to 0.7 μg/kg per hour, titrated to the desired level of sedation.¹²  A higher dose of dexmedetomidine (1.5 μg/kg per hour) has been administered without clinically significant adverse effects.¹¹  When dexmedetomidine is given to replace another sedative, a loading dose is generally not required.⁸  Bolus therapy and rapid dosage adjustment to achieve the desired level of sedation have been associated with more adverse drug events.^39,40  It is recommended that dexmedetomidine be infused for 24 hours.¹⁶  However, infusions of greater than 24 hours have not shown any adverse effects.³⁹ 

Dexmedetomidine can be administered safely in conjunction with anesthetics, sedatives, hypnotics, neuromuscular blockade agents, and opioids.¹²  Most patients treated with dexmedetomidine require additional sedative or analgesic agents.^38,41  Concurrent use of dexmedetomidine and other sedatives may enhance the effects of dexmedetomidine and may require dose reductions to reduce the risk of pharmacodynamic interactions.¹²  Interactions have been seen specifically with sevoflurane, isoflurane, propofol, alfentanil, and midazolam.

Dexmedetomidine can be infused before, during, and after extubation of mechanically intubated patients.⁸ 

Because of inhibition of norepinephrine release, most of the significant adverse effects of dexmedetomidine are related to sympatholytic activity and are often dose related. Martin et al⁴²  compared dexmedetomidine with placebo in 401 postsurgical patients in an ICU in a double-blind, randomized, multicenter, controlled trial. The most commonly reported adverse effects with dexmedetomidine were hypotension (61 of 203 patients [30%], P < .001), hypertension (24 of 203 patients [12%], P = .005), and bradycardia (18 of 203 patients [18%], P = .003).⁴²  Riker et al⁴³  found a greater incidence of bradycardia in patients receiving dexmedetomidine (103 of 244 patients [42.2%]) than in those receiving midazolam (23 of 122 patients [18.9%], P < .001).

Increased episodes of hypotension and bradycardia are associated with higher or rapid dose increases of dexmedetomidine.⁴¹  The loading dose of dexmedetomidine may cause temporary hypertension because of the initial activation of vascular α₂-adrenoreceptors. Persistent hypertension may be corrected by decreasing the rate of infusion.¹² 

Dexmedetomidine should be administered with caution to patients with advanced heart block, severe ventricular dysfunction, hypovolemia, preexisting diabetes mellitus, or chronic hypertension because the drug may cause more pronounced hypotension and bradycardia in this patient population.⁸  Caution should be used when administering dexmedetomidine along with medications that may also cause bradycardia or have sympatholytic effects.⁸  Medications that have negative chronotropic effects should be used with care because they may increase the risk of hypotension by causing an additive pharmacodynamic effect.¹²  Nurses in cardiovascular ICUs need to be particularly vigilant because patients are at increased risk of cardiovascular adverse effects with the concomitant use of chronotropic drugs.

The dexmedetomidine drug monograph indicates an increased risk of adverse effects with abrupt discontinuation. These adverse effects include tachycardia, hypertension, anxiety, agitation, headache, tremor, nausea, and vomiting.⁸  The adverse effects of withdrawal are due to a rebound effect. However, in a prospective, double-blind, randomized, multicenter trial of 244 patients, Riker et al⁴³  did not report adverse events or drug-related withdrawal effects with abrupt discontinuation of dexmedetomidine infusions.

We reviewed studies that examined the role of dexmedetomidine in the management of delirium (summarized in Table 5). We did not include studies in which dexmedetomidine was part of a preventive strategy. Overall, dexmedetomidine seems to have a favorable effect on reducing the incidence and duration of delirium.^{5,35,43,44,47,49} 

The Dexmedetomidine Compared to Morphine study and the Dexmedetomidine for Sepsis in ICU Randomized Evaluation trial did not show statistically significant results that dexmedetomidine reduced delirium.^36,48  However, the end point of delirium was not a primary outcome of the latter study, and the result should be considered within this context.³⁶  The results are supported by a recent meta-analysis of 8 studies by Lui et al⁵⁰  comparing dexmedetomidine with propofol in patients who underwent cardiac surgery. Of the 8 studies reviewed, 4 reported the incidence of delirium, with sample sizes ranging from 55 to 295 patients. Among these 4 studies, 18 of 193 patients (9.3%) in the dexmedetomidine group and 47 of 200 patients (23.5%) in the propofol group had delirium.⁵⁰ 

Lui et al⁵⁰  demonstrated that dexmedetomidine sedation significantly reduced postoperative delirium (risk ratio, 0.40; 95% CI, 0.24-0.64; P = .0002). They recommended that the results be viewed with caution because the number of patients was limited and the effect may have been overestimated. The primary outcomes, selection of patients, study periods, sedation levels, and outcome definitions varied. The methods of assessments differed across trials, which may have affected the precision and reliability of outcome comparisons.⁵⁰ 

At face value, the results of the review demonstrated a reduction in delirium with dexmedetomidine administration. However, generalizing the results should be done with caution because the studies used different inclusion and exclusion criteria, sampling methods, and medication protocols. In some studies, delirium was a secondary outcome. Given the heterogeneity of the studies, it is difficult for us to make a consensus statement on the use of dexmedetomidine for delirium.

As with antipsychotic and atypical antipsychotic agents in patients with delirium, the use of dexmedetomidine requires further investigation in adequately powered, double-blind, randomized, controlled trials. However, in comparison with current pharmacological management strategies, dexmedetomidine does appear to have some advantages. A comparison of the clinical effects of dexmedetomidine and other pharmacological agents is presented in Table 6.⁵¹ 

In comparison with the most common adverse effects of antipsychotics (QTc prolongation, extrapyramidal effects, and sedation), the adverse effect profile of dexmedetomidine may be less detrimental to critically ill patients.⁵²  Additionally, haloperidol is metabolized in the liver by the cytochrome P450 enzymes, resulting in more drug interactions and potentiating the drug’s effect.⁸  Clinicians must be aware of interactions and be diligent in ensuring that patients do not receive medications that have known interactions.

Dexmedetomidine has advantages over current pharmacological management strategies, considering the pharmacokinetic profile of the drug and the pathophysiology of delirium. Maldonado et al⁵  proposed that dexmedetomidine has antidelirium properties for the following reasons:

• Dexmedetomidine has a high specificity and selectivity for α₂-adrenoreceptors and blocks norepinephrine, disrupting 1 of the neurotransmitter pathways theorized to play a role in delirium.

• Dexmedetomidine produces sedative effects without respiratory depression. Sedatives and analgesics can cause respiratory depression, which can lead to hypoxemia. In the critically ill patient, decreased oxygen supply and increased oxygen demand can lead to decreased oxygen availability to cerebral tissue and the development of delirium.

• Dexmedetomidine has no anticholinergic effects.

• Dexmedetomidine theoretically promotes a more physiologic sleep-wake cycle, reducing the incidence of delirium.^2,5 

• Dexmedetomidine reduces the need for deliriogenic agents such as benzodiazepines.

Hipp and Ely²⁷  suggested that lower delirium rates associated with dexmedetomidine use resulted from avoidance of deliriogenic agents, such as benzodiazepines and opioids, rather than from a protective effect of dexmedetomidine against delirium.

Dexmedetomidine appears to have superior qualities over the antipsychotic medications in the management of delirium. However, because evidence related to the successful use of the drug is limited, we would strongly caution clinicians to be vigilant when administering dexmedetomidine to ICU patients with delirium.

The critical care nurse plays a fundamental role in implementing strategies to prevent delirium and in assessing the development of delirium. We cannot emphasize enough the importance of using strategies as outlined in Table 3 to prevent delirium and using screening tools to identify delirium. Maldonado²²  recommends identifying the underlying cause of delirium and tailoring the treatment plan according to each patient’s situation. However, critical care RNs must assess and manage delirium to resolve the situation effectively and prevent complications.

Pharmacological intervention is likely to be required to treat delirium. For patients who need continuous intravenous infusion of a pharmacologic agent, the SCCM 2013 clinical practice guidelines indicate a “moderate quality of evidence” recommending the use of dexmedetomidine over benzodiazepines.¹⁹  The guidelines do not provide a recommendation for the use of other nonbenzodiazepine drugs or atypical antipsychotics to manage delirium. According to the SCCM guidelines, haloperidol is not recommended for reducing the duration of delirium.¹⁹  However, no other statement is provided regarding the use of haloperidol or atypical antipsychotics to manage delirium.

We support the SCCM clinical practice guideline recommendations and suggest using dexmedetomidine instead of benzodiazepines to manage delirium.¹⁹  We also recommend using dexmedetomidine for patients who have delirium and no contraindications to dexmedetomidine. When dexmedetomidine is the pharmacological agent used to treat delirium, critical care nurses must have a comprehensive understanding of the pharmacodynamics of the drug.

The best approach to delirium is to use screening tools for early identification of the condition. However, when delirium does occur and pharmacological management is warranted, considering the patient’s characteristics and underlying cause of delirium is crucial for effective management and reduction of adverse effects. Pharmacological management may include dexmedetomidine. CCN