Sedation during mechanical ventilation in infants and children: dexmedetomidine versus midazolam.

Background: We sought to compare the efficacy of midazolam versus dexmedetomidine for sedation during mechanical ventilation in infants and children.

Methods: We performed a prospective, randomized trial in a pediatric intensive care unit in a tertiary care center. Infants and children requiring mechanical ventilation underwent a continuous infusion of either midazolam (starting dose of 0.1 mg/kg/h) or dexmedetomidine (starting dose of either 0.25 or 0.5 [micro]g/kg/h) with intermittent morphine, as needed. The efficacy of sedation was assessed using the Ramsay sedation scale, pediatric intensive care unit sedation score, and the tracheal suctioning score as well as bispectral monitoring.

Results: There were 10 patients in each group. Sedation as assessed by the clinical sedation scores and the bispectral index was equivalent in the 3 groups. There were 36 morphine boluses administered to the midazolam group versus 29 and 20 morphine boluses administered respectively to the 0.25 and 0.5 [micro]g/kg/h dexmedetomidine groups (P = 0.02 for midazolam versus 0.5 [micro]g/kg/h dexmedetomidine). Total morphine use (mg/kg/24 h) was 0.74 [+ or -] 0.5, 0.55 [+ or -] 0.38, and 0.28 [+ or -] 0.12 in the midazolam and the two dexmedetomidine groups respectively (P = not significant for midazolam versus 0.25 dexmedetomidine, P = 0.01 for midazolam versus 0.5 dexmedetomidine). In the two dexmedetomidine groups, 5 of 6 patients who at some point had a Ramsay score of 1 were less than 12 months of age while only 1 was more than 12 months of age (P < 0.05).

Conclusions: At a dose of 0.25 [micro]g/kg/h, dexmedetomidine was approximately equivalent to midazolam at 0.22 mg/kg/h. At 0.5 [micro]g/kg/h, dexmedetomidine provided more effective sedation as demonstrated by the need for fewer bolus doses of morphine, a decrease in the 24-hour requirements for supplemental morphine, as well as a decrease in the total number of assessment points with a Ramsay score of 1 (inadequate sedation) and the number of patients who had a Ramsay score of 1.

Key Words: children, dexmedetomidine, mechanical ventilation, midazolam, morphine sulfate, sedation

**********

Sedative and analgesic agents are commonly administered to alleviate the discomfort associated with mechanical ventilation. In addition to humanitarian concerns, appropriate levels of sedation are needed to prevent patient-ventilator asynchrony, thereby allowing for effective mechanical ventilation, as well as preventing inadvertent tracheal extubation or dislodgement of invasive vascular devices. In the pediatric-aged patient, benzodiazepines and opioids are the agents used most commonly to achieve these goals.

Dexmedetomidine (Precedex; Abbott Laboratories, Abbott Park, IL) is a centrally-acting, [[alpha].sub.2] agonist that is currently approved by the U.S. Food and Drug Administration for short-term use ([less than or equal to]24 h) to provide sedation in adults. Preliminary experience in the adult population has demonstrated its efficacy in alleviating the discomfort associated with mechanical ventilation. (1,2) To date, reports regarding the use of dexmedetomidine in pediatric patients have been anecdotal. (3,4) The current study prospectively compares the efficacy of midazolam with two doses of dexmedetomidine (0.25 and 0.5 [micro]g/kg/h) to provide sedation during mechanical ventilation in infants and children.

Methods

The study was approved by the Institutional Review Board of the University of Missouri and written, informed consent was obtained from a parent. Patients admitted to the pediatric intensive care unit who required mechanical ventilation with an endotracheal tube were considered eligible for inclusion. Patients with preexisting central nervous system dysfunction or acquired conditions resulting in ongoing dysfunction were excluded. After endotracheal intubation and before obtaining informed consent, sedation was provided by intermittent doses of midazolam (0.1 mg/kg/h) as needed. The patients were randomized to receive either midazolam (starting dose of 0.1 mg/kg/h) or dexmedetomidine (starting dose of either 0.25 or 0.5 [micro]g/kg/h). After randomization, if the patient was judged to be inadequately sedated, a bolus of either midazolam (0.1 mg/kg) or dexmedetomidine (0.25 or 0.5 [micro]g/kg) was given before the initiation of the infusion. The bolus dose was added to the infusion when calculating the mean daily infusion rate of the medication (midazolam or dexmedetomidine). After the initiation of the infusions, supplemental sedation was provided by intermittent doses of morphine (0.08-0.1 mg/kg/h) as needed. If multiple additional doses of morphine were required (3-4 over an 8 hour period), the midazolam infusion was increased by 0.05 to 0.1 mg/kg/h and the dexmedetomidine infusion was increased by 0.15 to 0.25 [micro]g/kg/h. Before increasing the infusion, if needed, the patient was bolused with the current hourly rate over a 5-minute period. This bolus dose was included in the total daily dose of the medication. After 24 hours on either the midazolam or the dexmedetomidine infusions, if ongoing mechanical ventilation was still necessary, the patient was switched to the alternative agent and the study continued.

The quality of sedation was assessed every 2 hours using 3 clinical sedation scales and the Bispectral Index Monitor (BIS monitor; Aspect Medical, Newton, MA). The 3 clinical sedation scores included the Ramsay sedation scale, the pediatric intensive care unit (PICU) sedation score, and the tracheal suctioning score (Table 1). Heart rate and blood pressure were recorded every 2 hours. Statistical analysis included a nonpaired t test with a Bonferroni correction, when multiple comparisons were used, to compare demographic data (age, weight), daily morphine use, and hemodynamic parameters (heart rate, systolic/diastolic blood pressure). A Wilcoxon ranked-sum test was used to analyze nonparametric data including BIS numbers and sedation scores (Ramsay, PICU sedation score, tracheal suction score). A two-tailed Fisher exact test was used to compare demographic data (sex distribution), number of infusion changes, requirement for morphine boluses, number of patients with a Ramsay score of 1 and the total number of assessment points with a Ramsay score of 1 between the 3 groups. All data are presented as the mean [+ or -] SD with P < 0.05 considered significant.

Results

There were 10 patients in each of the 3 groups (midazolam, dexmedetomidine 0.25 [micro]g/kg/h, and dexmedetomidine 0.5 [micro]g/kg/h). There were no statistically significant differences among patient demographics, study duration, and data collection for the 3 groups (Table 2). The mean infusion rates were 0.22 [+ or -] 0.05 mg/kg/h for the midazolam group, 0.28 [+ or -] 0.07 [micro]g/kg/h in the 0.25 dexmedetomidine group, and 0.68 [+ or -] 0.15 [micro]g/kg/h in the 0.5 dexmedetomidine group. Using the 3 sedation scores and the BIS number, there were no differences noted between the 3 groups (Table 3).

Thirty-six morphine boluses were administered to the midazolam group versus 29 and 20 morphine boluses administered to the 0.25 and 0.5 [micro]g/kg/h dexmedetomidine groups respectively. (P = not significant (NS) for midazolam versus 0.25 dexmedetomidine and P = 0.02 for midazolam versus 0.5 dexmedetomidine). Total morphine use (mg/kg/24 h) was 0.74 [+ or -] 0.5, 0.55 [+ or -] 0.38, and 0.28 [+ or -] 0.12 in the midazolam group and the two dexmedetomidine groups (0.25 and 0.5 [micro]g/kg/h) respectively (P = NS for midazolam versus 0.25 dexmedetomidine, P = 0.01 for midazolam versus 0.5 dexmedetomidine, and P < 0.05 for 0.25 dexmedetomidine versus 0.5 dexmedetomidine). During the course of the study, there were 15 infusion changes in the midazolam group versus 6 and 4 infusion changes in the 0.25 and 0.5 [micro]g/kg/h dexmedetomidine groups respectively (P = 0.06 for midazolam versus 0.25 dexmedetomidine, and P = 0.01 for midazolam versus 0.5 dexmedetomidine).

When considering those patients that were inadequately sedated as judged by a Ramsay score of 1, this occurred at 14 assessment points in 6 of the 10 patients sedated with midazolam, at 11 assessment points in 4 of the 10 patients sedated with 0.25 [micro]g/kg/h of dexmedetomidine, and at 5 assessment points in 2 of the 10 patients sedated with 0.5 [micro]g/kg/h of dexmedetomidine (P = NS for midazolam versus 0.25 dexmedetomidine and P = 0.052 for midazolam versus 0.5 dexmedetomidine). In the two dexmedetomidine groups, 5 of 6 patients who at some point had a Ramsay score of 1 were less than 12 months of age while only 1 was more than 12 months of age (P < 0.05).

When evaluating hemodynamic parameters including systolic/diastolic blood pressure and heart rate, there was no difference noted between the 3 groups in regards to systolic and diastolic blood pressure (Table 4). Heart rates were significantly lower in the two dexmedetomidine groups than in the midazolam group (Table 4). One patient was removed from the study because of bradycardia (heart rate of 40-50 beats/min). Data from this patient is not included in the current cohort of patients. The patient was a 5 week old, 3.6 kg infant with trisomy 21 who was concurrently receiving digoxin. No hypotension was noted and the bradycardia resolved within 1 hour of discontinuing the dexmedetomidine infusion. This patient has been reported previously as an isolated case report of bradycardia during dexmedetomidine infusion. (5) No other adverse hemodynamic effects (bradycardia or hypotension) were noted in the study population.

Discussion

The current study provides prospective data regarding the efficacy of dexmedetomidine for sedation during mechanical ventilation in infants and children. We found the efficacy of 0.25 [micro]g/kg/h of dexmedetomidine to be approximately equivalent to that of midazolam at 0.22 mg/kg/h. Although the level of sedation as assessed by the three clinical sedation scales and the BIS number was equivalent in the 3 groups, patients receiving the higher dose of dexmedetomidine (0.5 [micro]g/kg/h) required fewer bolus doses of morphine, used less total morphine per 24 hours, and had fewer points when a Ramsay score of 1 was assigned, thereby demonstrating the higher dose of dexmedetomidine (0.5 [micro]g/kg/h) to be somewhat superior to midazolam.

Dexmedetomidine is the pharmacologically active dextro-isomer of medetomidine. The [[alpha].sub.2] adrenergic agonist class of drugs can be divided into 3 groups: imidazolines, phenylethylamines, and oxalozepines. Both dexmedetomidine and clonidine are imidazole compounds that exhibit a high ratio of specificity for the [[alpha].sub.2] versus the [[alpha].sub.1] receptor. Clonidine exhibits an [[alpha].sub.2]:[[alpha].sub.1] specificity ratio of 200:1 while that of dexmedetomidine is 1,600:1. (6) Therefore, dexmedetomidine is considered a full agonist at the [[alpha].sub.2] adrenergic receptor. By activation of specific transmembrane [[alpha].sub.2] adrenergic receptors at various locations throughout the central nervous system, dexmedetomidine produces physiologic effects including sedation, anxiolysis, and analgesia. These physiologic effects are mediated via stimulation of postsynaptic [[alpha].sub.2] adrenergic receptors that activate a pertussis toxin-sensitive guanine nucleotide regulatory protein (G protein) resulting in inhibitory feedback and decreased activity of adenylyl cyclase. (7,8) The subsequent reduction in cyclic adenosine monophosphate (cAMP) and cAMP-dependent protein kinase activity causes a predominant dephosphorylation of various species of ion channels, (9) which modifies ion translocation and membrane conductance, resulting in decreased neuronal activation. (10) Dexmedetomidine also activates receptors in the medullary vasomotor center, thereby reducing norepinephrine turnover and decreasing central sympathetic outflow resulting in alterations in sympathetic function including a decrease in heart rate and blood pressure. Additional effects result from the central stimulation of parasympathetic outflow and inhibition of sympathetic outflow from the locus caeruleus in the brainstem. The latter effect plays a prominent role in the sedation and anxiolysis produced by these agents as decreased noradrenergic output from the locus caeruleus allows for increased firing of inhibitory neurons including the [gamma]-amino butyric acid (GABA) system resulting in sedation and anxiolysis. (11-13) Dexmedetomidine, through activation of [[alpha].sub.2] adrenergic receptors of the dorsal horn of the spinal cord, regulates the release of Substance P resulting in primary analgesic effects as well as potentiation of opioid-induced analgesia.

Previous reports in adults have demonstrated dexmedetomidine to be an effective agent for sedation and anxiolysis during mechanical ventilation. Venn et al. evaluated the efficacy of dexmedetomidine in 119 postoperative cardiac and general surgical patients requiring postoperative ventilation. (1) The patients were randomized to dexmedetomidine administered as a loading dose of 1 [micro]g/kg over 10 minutes followed by an infusion of 0.7 [micro]g/kg/h or placebo. Despite having equivalent Ramsay scores, patients receiving dexmedetomidine had a statistically significant decrease in supplemental midazolam and morphine requirements. Similar findings were reported by Martin et al. in a prospective, randomized trial of 401 postsurgical patients. (2)

Aside from the current study, experience with dexmedetomidine in pediatric-aged patients has been anecdotal. We have previously reported our experience with dexmedetomidine in two separate cohorts of patients. (3,4) The first series included four patients in whom dexmedetomidine was used to provide sedation during mechanical ventilation, as an adjunctive agent intraoperatively for controlled hypotension during posterior spinal fusion, and for procedural sedation during upper gastrointestinal endoscopy. The second series included five patients in additional clinical scenarios including sedation during spontaneous ventilation in a toddler with status asthmaticus, in the management of substance withdrawal (tobacco, cannabinoids, and ethanol) in an adolescent after complex cardiothoracic surgery, and in the postanesthesia arena for the control of shivering and emergence delirium. Although anecdotal, these two series demonstrated that dexmedetomidine was effective in all of these clinical scenarios, except for procedural sedation during upper gastrointestinal endoscopy, and was free of significant adverse effects.

As with many of the agents currently used for sedation in the intensive care unit population, there is a potential for adverse cardiorespiratory effects with dexmedetomidine. In the study of Venn et al, (1) 18 of 66 patients that received dexmedetomidine developed adverse hemodynamic effects that included either hypotension (mean arterial pressure less than 60 mm Hg or a greater than 30% decrease from baseline) or bradycardia (heart rate less than 50 beats/min). In 11 of these 18 patients, the hemodynamic effects occurred during the bolus infusion. Talke et al (14) noted similar hemodynamic effects in a cohort of adult patients after vascular surgery. Although only 1 of 22 patients experienced hypotension, they also noted one patient who developed a 5 to 10 sinus pause after anesthetic induction with thiopental and fentanyl suggesting that there may be a synergistic vagotonic effect of dexmedetomidine with other agents. In their perioperative study, Peden et al (15) noted a similar episode of sinus pause as well as episodes of bradycardia when dexmedetomidine was combined with propofol, and altered their protocol to include the administration of an anti-cholinergic agent. Given their experience and that reported in the literature, they recommended the administration of anticholinergic agent to patients less than 40 years of age.

In our study population, the baseline heart rate was significantly lower in both of the dexmedetomidine groups when compared with midazolam; as noted previously, one patient who was concurrently receiving digoxin developed clinically significant bradycardia (heart rate 40-50 beats/min), which necessitated discontinuation of the dexmedetomidine. Although no episodes of hypotension were noted, given the potential adverse effects of dexmedetomidine on cardiac output as demonstrated by the study of Bloor et al. in healthy, adult volunteers, (16) close monitoring of cardiovascular function is suggested until a larger experience with this agent in pediatric-aged patients is accumulated.

Similarly, there is the potential for adverse effects from dexmedetomidine on respiratory function. As all of the patients in the current study were receiving mechanical ventilation, no attempt was made to discern the potential effects of dexmedetomidine on ventilatory function in the current cohort of patients. Although several of the previous studies in adults have demonstrated the feasibility of weaning of ventilatory support with tracheal extubation while receiving a dexmedetomidine infusion, other studies have clearly demonstrated potential deleterious ventilatory effects from dexmedetomidine. (17,18) After the administration of 2 [micro]g/kg of dexmedetomidine, Belleville et al. demonstrated a rightward shift and down-sloping of the C[O.sub.2] response curve in healthy male volunteers. (17) They also noted a decreased ventilatory response to an end-tidal C[O.sub.2] of 55 mm Hg during carbon dioxide challenge.

Conclusion

In summary, we present preliminary experience with dexmedetomidine in pediatric-aged patients. At a dose of 0.25 [micro]g/kg/h, dexmedetomidine was approximately equivalent to midazolam at 0.22 mg/kg/h. A higher infusion rate of 0.5 [micro]g/kg/h provided more effective sedation as demonstrated by the need for fewer bolus doses of morphine, a decrease in the 24-hour requirements for supplemental morphine as well as a decrease in the number of Ramsay scores of 1 and the number of patients who had a Ramsay score of 1. Overall, dexmedetomidine was less effective in patients less than 12 months of age as 5 of the 6 patients who received a Ramsay score of 1 during dexmedetomidine administration (either 0.25 or 0.5 [micro]g/kg/h) were less than 12 months of age.

Table 1. Clinical sedation scales used for the study

Ramsay score
  1 = Anxious, agitated, restless
  2 = Eyes open, cooperative, oriented, tranquil
  3 = Responds (opens eyes) only to command, light touch, normal tone of
      voice
  4 = Brisk response to light glabellar tap or loud noise/voice
  5 = Sluggish response to light glabellar tap or loud noise/voice
  6 = No response to light glabellar tap or loud noise/voice
PICU sedation score
  1 = Awake, alert
  2 = Occasionally drowsy, easy to arouse
  3 = Frequently drowsy, easy to arouse
  4 = Somnolent
Tracheal suctioning score
  1 = Patient is restless or distressed when not disturbed
  2 = Patient is awake and moving, but not distressed if left alone
  3 = Movement only with nursing care, major limb movement/distress with
      tracheal suctioning
  4 = Cough, grimace or minor limb movement with suctioning
  5 = No response to tracheal suctioning

Table 2. Demographic data of the three groups

              midazolam       dexmedetomidine 0.25  dexmedetomidine 0.5

Age (months)  36 [+ or -] 34  44 [+ or -] 54        39 [+ or -] 44
Weight        19 [+ or -] 20  22 [+ or -] 27        21 [+ or -] 24
Gender (M/F)   6/4             7/3                   7/3
Infusion
  duration
  (h)         22 [+ or -] 8   21 [+ or -] 10        22 [+ or -] 9

Table 3. BIS number and clinical sedation scores of the three groups (a)

                 midazolam          dexmedetomidine    dexmedetomidine
                                    0.25               0.5

BIS number       57 [+ or -] 8      51 [+ or -] 12     60 [+ or -] 10
Ramsay score      3.6 [+ or -] 0.9   3.7 [+ or -] 1.1   3.5 [+ or -] 0.8
PICU sedation
  score           2.9 [+ or -] 1.2   3.1 [+ or -] 0.9   2.9 [+ or -] 1.1
Tracheal
  suction score   3.5 [+ or -] 1.2   3.2 [+ or -] 1.4   3.5 [+ or -] 1.3

(a) BIS, bispectral index score.

Table 4. Hemodynamic parameters of the three groups (a)

                      midazolam        dexmedetomidine  dexmedetomidine
                                       0.25             0.5

Systolic BP (mm Hg)    94 [+ or -] 22   88 [+ or -] 31   96 [+ or -] 26
Diastolic BP (mm Hg)   46 [+ or -] 16   52 [+ or -] 19   49 [+ or -] 21
Heart rate (beats/
  min)                142 [+ or -] 36  122 [+ or -] 31  112 [+ or -] 26

(a) BP, blood pressure.

Accepted August 29, 2003.

Copyright [c] 2004 by The Southern Medical Association

0038-4348/04/9705-0451

References

1. Venn RM, Bradshaw CJ, Spencer R, et al. Preliminary UK experience of dexmedetomidine, a novel agent for postoperative sedation in the intensive care unit. Anaesthesia 1999;54:1136-1142.

2. Martin E, Ramsay G, Mantz J, et al. The role of the [[alpha].sub.2] adrenoceptor agonist dexmedetomidine in post-surgical sedation. J Intensive Care Med 2003;18:29-41.

3. Tobias JD, Berkenbosch JW. Initial experience with dexmedetomidine in paediatric-aged patients. Paediatr Anaesth 2002;12:171-175.

4. Tobias JD, Berkenbosch JW, Russo P. Additional experience with dexmedetomidine in pediatric patients. South Med J 2003;96:871-875.

5. Berkenbosch JW, Tobias JD. Development of bradycardia during sedation with dexmedetomidine in an infant concurrently receiving digoxin. Pediatr Crit Care Med 2003;4:203-205.

6. Virtanen R, Savola JM, Saano V, et al. Characterization of the selectivity, specificity and potency of medetomidine as an [[alpha].sub.2] adrenoceptor agonist. Eur J Pharmacol 1988;150:9-14.

7. Correa-Sales C, Reid K, Maze M. Pertussis toxin-mediated ribosylation of G proteins blocks the hypnotic response to an [[alpha].sub.2]-agonist in the locus coeruleus of the rat. Pharmacol Biochem Behav 1992;43:723-727.

8. Correa-Sales C, Nacif-Coelho C, Reid K, et al. Inhibition of adenylate cyclase in the locus coeruleus mediates the hypnotic response to an [[alpha].sub.2] agonist in the rat. J Pharmacol Exp Ther 1992;263:1046-1049.

9. Nacif-Coelho C, Correa-Sales C, Chang LL, et al. Perturbation of ion channel conductance alters the hypnotic response to the [[alpha].sub.2]-adrenergic agonist dexmedetomidine in the locus coeruleus of the rat. Anesthesiologv 1994;81:1527-1534.

10. Sculptoreanu A, Scheuer T, Catterall WA. Voltage-dependent potentiation of L-type [Ca.sup.2+] channels due to phosphorylation by cAMP-dependent protein kinase. Nature 1993;364:240-243 (letter).

11. Correa-Sales C, Rabin BC, Maze M. A hypnotic response to dexmedetomidine, an [[alpha].sub.2] agonist, is mediated in the locus coeruleus in rats. Anesthesiology 1992;76:948-952.

12. Doze VA, Chen BX, Maze M. Dexmedetomidine produces a hypnoticanesthetic action in rats via activation of central [[alpha].sub.2] adrenoceptors. Anesthesiology 1989;71:75-79.

13. Nelson LE, Lu J, Guo T, et al. The [[alpha].sub.2]-adrenoceptor agonist dexmedetomidine converges on an endogenous sleep-promoting pathway to exert its sedative effects. Anesthesiology 2003;98:428-436.

14. Talke P, Chen R, Thomas B, et al. The hemodynamic and adrenergic effects of perioperative dexmedetomidine infusion after vascular surgery. Anesth Analg 2000;90:834-839.

15. Peden CJ, Cloote AH, Stratford N, et al. The effect of intravenous dexmedetomidine premedication on the dose requirement of propofol to induce loss of consciousness in patients receiving alfentanil. Anaesthesia 2001;56:408-413.

16. Bloor BC, Ward DS, Belleville JP, et al. Effects of intravenous dexmedetomidine in humans: Part II--Hemodynamic changes. Anesthesiology 1992;77:1134-1142.

17. Belleville JP, Ward DS, Bloor BC, et al. Effects of intravenous dexmedetomidine in humans: Part I--Sedation, ventilation, and metabolic rate. Anesthesiology 1992;77:1125-1133.

18. Hall JE, Uhrich TD, Barney JA, et al. Sedative, amnestic, and analgesic properties of small-dose dexmedetomidine infusions. Anesth Analg 2000;90:699-705.

RELATED ARTICLE: Key Points

* Dexmedetomidine is a centrally acting, [[alpha].sub.2] agonist which is currently FDA approved for short-term use ([less than or equal to] 24 hours) to provide sedation in adults.

* When administered at a dose of 0.25 [micro]g/kg/h, dexmedetomidine provided a level of sedation equivalent to midazolam of 0.22 mg/kg/h.

* Dexmedetomidine at 0.5 [micro]g/kg/h provided superior sedation to midazolam as evidenced by a decrease in the 24 hour requirement for supplemental morphine and a decrease in the number of Ramsay scores of 1 (inadequate sedation).

* Dexmedetomidine was less effective in patients less than 12 months of age, as 5 of the 6 patients who received a Ramsay score of 1 during the infusion of either dexmedetomidine dose (0.25 or 0.5 [micro]g/kg/h) were less than 12 months of age.

Joseph D. Tobias, MD, and John W. Berkenbosch, MD

From the Division of Pediatric Critical Care/Pediatric Anesthesiology, Departments of Child Health and Anesthesiology, University of Missouri, Columbia, MO.

Abbott Pharmaceuticals provided dexmedetomidine free of charge for the purpose of this study. This study was performed at the University of Missouri, Columbia, MO.

Reprint requests to Joseph D. Tobias, MD, Department of Anesthesiology, University of Missouri, 3W40H, One Hospital Drive, Columbia, MO 65212. Email: [e-mail]