Pro-Con Perspective

Dexmedetomidine vs Midazolam for Sedating Mechanically Ventilated ICU Patients

By James E. McFeely, MD, Medical Director Critical Care Units, Alta Bates Summit Medical Center, Berkeley, CA, is Associate Editor for Critical Care Alert.

By Andrew Luks, MD, Pulmonary and Critical Care Medicine, University of Washington, Seattle, is Associate Editor for Critical Care Alert.

They report no financial relationship to this field of medicine.

Editor's Note: The alpha2-receptor agonist dexmedetomidine (Precedex®), introduced for perioperative and procedural sedation and other short-term applications, is approved by the FDA for use in the ICU, although the package insert emphasizes that this approved use is for not more than 24 hours. This study by Riker et al investigated whether dexmedetomidine, which has several potential advantages over benzodiazepines and opioids for more prolonged ICU sedation, was superior to midazolam for this use in mechanically ventilated patients. Its results appear quite positive, yet the article has provoked controversy, with experienced, critical care clinicians interpreting it quite differently. Editorial Board members James McFeely, MD, and Andrew Luks, MD, independently evaluated the Riker paper, and provide divergent perspectives on its findings and implications for the use of dexmedetomidine for sedating mechanically ventilated ICU patients.

Synopsis: This prospective, double-blind, multicenter, randomized trial found that dexmedetomidine sedated patients as well as midazolam and resulted in a shorter duration of mechanical ventilation, with fewer patients experiencing delirium.

Source: Riker RR, et al. Dexmedetomidine vs midazolam for sedation of critically ill patients: A randomized trial. JAMA 2009;301:489-499.

Midazolam is commonly used for sedation in critically ill patients but concerns persist as to whether it induces delirium, and also prolongs mechanical ventilation and ICU stay, due to its accumulation in lipid stores with prolonged administration. Riker and colleagues sought to determine whether dexmedetomidine could serve as a suitable alternative to midazolam, achieving similar levels of sedation with fewer adverse consequences. To investigate this question, dexmedetomidine was frequently used at a dose higher than the limits approved by the FDA and well beyond the 24-hour time limit listed on the drug label.

The authors conducted a prospective, double-blind, randomized trial at 68 centers in 5 countries over a 2.5-year period. Patients were included if they were ≥ 18 years of age, had been on mechanical ventilation for ≤ 4 days, and had an anticipated duration of mechanical ventilation of ≥ 3 days. Exclusion criteria included hypotension requiring the use of 2 or more vasopressors; admission diagnoses of trauma, burns, acute myocardial infarction, or severe neurologic injury; dialysis; acute hepatitis; and severe chronic liver disease. Included patients were taken off their pre-existing sedative medications and randomized to receive either dexmedetomidine (0.8 µg/kg/hr) or midazolam (0.06 mg/kg/hr), intravenously, with adjustments as needed to achieve sedation as assessed by the Richmond Agitation and Sedation Scale (RASS) with a score between -2 and +1. Patients who were not adequately sedated by the study drug could receive open-label midazolam boluses of 0.01-0.05 mg/kg every 10-15 minutes until adequate sedation was achieved. Fentanyl and haloperidol boluses could also be administered for pain and delirium, respectively, and every patient had a daily sedation vacation. The primary endpoint was the percentage of time within the target sedation range (RASS score, -2 to +1). Secondary endpoints included prevalence and duration of delirium, as assessed by the Confusion Assessment Method for the ICU (CAM-ICU) protocol, use of fentanyl and open-label midazolam, the duration of mechanical ventilation, and length of ICU stay.

A total of 366 patients received the study drugs (244 dexmedetomidine and 122 midazolam) and baseline characteristics were evenly matched in the two groups. Despite the multinational study design, the overwhelming majority of study patients were from the United States. The mean maintenance infusion doses were 0.056 mg/kg/hr for midazolam (4.4 mg for an 80 kg patient) and 0.83 µg/kg/hr for dexmedetomidine.

There was no difference in time within the target RASS range between dexmedetomidine (77.3%) and midazolam (75.1%). Fifty-four percent of the dexmedetomidine patients experienced delirium compared to 76.6% of the midazolam-treated patients, a result that held up regardless of whether the patients had delirium at the time of study enrollment. Open-label midazolam use was more common in the dexmedetomidine group (63% vs. 49%; P = 0.02) while fentanyl and haloperidol use were the same in the two groups. The median time to extubation was shorter by 1.9 days in the dexmedetomidine group (3.7 vs 5.6 days; P = 0.01), but there were no significant differences in the length of ICU stay (5.9 vs 7.6 days; P = 0.24). Forty-two percent of the dexmedetomidine patients developed bradycardia, compared to only 20% of the midazolam patients, and 11% of patients experiencing bradycardia required treatment such as decreasing or stopping the study drug or administering atropine.

Commentary by James E. Mcfeely, MD

This article provides compelling evidence that dexmedetomidine is safe for prolonged use and at higher doses than currently recommended by the package insert. The study was well grounded, using treatment methodologies common in clinical practice and a patient group of appropriate interest for prolonged mechanical ventilation. The daily arousal assessment and use of bolus sedatives and narcotics are also consistent with current clinical practice. The trial was of a robust size and the outcome measures were important.

Dexmedetomidine appears to be at least as good as midazolam in maintaining sedation within a goal range. It appears superior, however, in a number of other areas. In particular, it is associated with a significant decrease in frequency of delirium. Midazolam is noted to be associated with delirium, which has negative short-term as well as long-term effects on patient outcome. Dexmedetomidine was also associated with an almost 2-day reduction in ventilator days, at least partially due to its lack of respiratory-depressant effect.

Complications of dexmedetomidine included a modest increase in frequency of hyperglycemia and an increased frequency of bradycardia. The reduction in heart rate was probably related to sympatholytic blunting. Of note, only 12 of 244 patients required intervention for bradycardia.

This paper should provoke a reassessment of the role of dexmedetomidine in all our ICUs. Those unfamiliar with the drug should become familiar with it and get it on formulary. Those who have it but are not using it should reassess its role, particularly in patients belonging to the same cohort as the subjects of this clinical trial.

Commentary by Andrew M. Luks, MD

On the surface, the results of this study would seem to suggest that dexmedetomidine is an ideal sedative for the ICU. It achieved levels of sedation comparable to one of our standard agents while causing less delirium and decreasing the duration of mechanical ventilation. Aside from the concerns about bradycardia, a problem we have seen at our institution when using drug levels high enough to achieve adequate sedation, there are some important limitations in this study that warrant attention and should dampen the enthusiasm for this medication that is sure to arise from this study.

The most important limitation is the protean list of study exclusion criteria that was actually much longer than the abbreviated list described above. The study population was also not very ill, with mean APACHE II scores of 18-19 in the two study groups. Given these issues, a sizeable majority of the patients at many of the institutions where we practice, particularly tertiary medical centers, would never have been included in this study and we are left to wonder whether these results would hold up in a broader population of critically ill patients.

There is also reason to question the data regarding the duration of mechanical ventilation, one of the important secondary study endpoints that will be used to justify the use of the medication. The study did not include explicit extubation criteria, nor did the investigators appear to use protocols for initiating spontaneous breathing trials, a tactic previously demonstrated to shorten the duration of mechanical ventilation compared to physician-driven decision making. To the extent that extubation practices differed across institutions and countries in the study, this might have had an impact on the study results in this regard.

Finally, the study did not address another important issue with this medication — its high cost. At our institution, dexmedetomidine administered at a rate of 0.7 µg/kg/hr to a 70 kg patient costs $370/day, more than 10 times the daily cost of midazolam and 90 times more than that of fentanyl. It is true that the higher cost may be offset by the decreased costs associated with a shorter duration of mechanical ventilation (e.g., less ventilator-associated pneumonia), but until we have data suggesting that target levels of sedation, less delirium, and shorter duration of mechanical ventilation can actually be achieved in a range of critically ill patients substantially broader than those examined in this study and more reflective of what we see in our daily practice, we should not be committing to this expensive alternative.