Etomidate and midazolam for procedural sedation: prospective, randomized trial

Gregory S. Hunt MD^a,b, Matthew T. Spencer MD^b,*, Daniel P. Hays PharmD^b

^aDepartment of Emergency Medicine, St. Luke’s Hospital, New Bedford, MA 02740, USA

^bDepartment of Emergency Medicine, University of Rochester Medical Center, Rochester, NY 14642, USA

Received 17 November 2004; accepted 2 February 2005

Abstract

Objective: To evaluate whether there is a difference in the time of sedation and time to patient disposition in patients undergoing procedural sedation with etomidate and midazolam.

Methods: Prospective, randomized, double-blind trial comparing etomidate (0.10 mg/kg) and midazolam (0.035 mg/kg) for patients requiring procedural sedation for reduction of Joint dislocations or long bone fractures.

Results: Forty-five patients were enrolled (24 randomized to etomidate, 21 to midazolam). Groups were similar in demographics and analgesic dosing. Mean time of sedation for etomidate was 15 minutes (SD, 10.97) and for midazolam was 32 minutes (SD, 16.13) ( P b .001). Mean time to disposition for etomidate was 121 minutes (SD, 73.28) and for midazolam was 111 minutes (SD, 96.36) ( P = .700). The mean quality of sedation for etomidate was 7.91 (SD, 1.53) and for midazolam was 7.48 (SD, 2.89) ( P = .570).

Conclusions: The use of etomidate compared with midazolam for procedural sedation provides a significant reduction in recovery time, without a reduction in time to patient disposition, while providing equal sedation quality.

D 2005

Introduction

Procedural sedation, formerly referred to as conscious sedation, is a Clinical process that provides alteration of a patient’s consciousness to perform painful procedures such as reduction of a fracture or dislocation [1,2]. It requires close patient monitoring that is time consuming and ties up

Oral Presentation: July 2003 NYACEP, Lake George, NY; Poster Presentation: October 2003 ACEP Research Forum, Boston, Mass.

T Corresponding author. Tel.: +1 585 275 2148; fax: +1 585 473 3516.

E-mail address: [email protected] (M.T. Spencer).

precious emergency department (ED) resources [3]. A sedation agent with a duration adequate for completion of the procedure but with a rapid recovery rate is ideal.

Midazolam, a short-acting benzodiazepine that acts on c-aminobutyric acid receptors, is one of the preferred drugs used for procedural sedation [4]. Recently, shorter-acting agents with faster recovery times such as etomidate have been suggested as alternatives [5-7].

Etomidate is a fast-acting nonbarbiturate, nonbenzodia- zepine, imidazole derivative [8]. It causes sedation and hypnosis by acting on c-aminobutyric acid receptors in the central nervous system and was first used in the United

0735-6757/$ - see front matter D 2005 doi:10.1016/j.ajem.2005.02.042

States in 1982 (Europe in 1972) [9-12]. It is most commonly used as an Induction agent for rapid sequence intubation but has recently been studied as a Sedative agent for use in procedural sedation because of its fast onset, Short duration of action, and limited side effects [4,6,7,11,12].

We evaluate, in a prospective double-blind trial, whether there is a difference in the time of sedation (T1) and time to patient disposition (T2) in patients undergoing procedural sedation with etomidate or midazolam.

Materials and methods

Study design

This is a prospective, double-blind, randomized trial comparing etomidate with midazolam for ED patients older than 10 years requiring procedural sedation. Full Research Subjects Review Board approval was granted for this study.

Study setting and population

The study was conducted at an academic tertiary care medical center with both adult and pediatric EDs, an emergency medicine residency program, and a pediatric emergency medicine fellowship. The facility functions as the regional trauma and Burn center and manages a volume of approximately 63000 adult and 25000 pediatric ED visits per year.

Study protocol

All patients presenting to the ED requiring procedural sedation were eligible for enrollment. Specific inclusion criteria were age above 10 years; dislocation of the shoulder, hip, knee, elbow, or wrist; and any displaced/angulated Long bone fractures requiring procedural sedation for reduction. Exclusion criteria were age below 10 years, previous adverse reaction to midazolam or etomidate, and hemodynamic instability [4,13].

Patient enrollment took place over 1 year between July 2002 and June 2003. Patients were identified and informed consent was obtained by trained patient enrollers, emergen- cy medicine resident physicians, and emergency medicine attending physicians.

During sedation, vital signs including oxygen saturation, heart rate, and blood pressure were recorded every 5 minutes by nursing staff as per hospital protocol. Patients were also given supplemental oxygen, with Emergency airway management equipment, suction, and reversal agents (naloxone and flumazenil) immediately available [2,3]. Analgesia was provided as per protocol with patients receiving morphine sulfate (0.1-0.15 mg/kg unless contra- indicated) preprocedure and fentanyl (50 lg increments) during the procedure [14]. Etomidate and midazolam were administered at doses of 0.10 and 0.035 mg/kg, respectively, with a maximum of 3 doses [3,6,9,15]. If reduction was not achieved after 3 doses of the study drug, the study was

terminated and arrangements were made through the orthopedic department to have reduction completed under general anesthesia in the operating room.

To reduce selection and pretest biases, our department pharmacist prepared and randomized the study drugs ahead of time in single-use multidose vials with midazolam diluted in normal saline so both medications could be given at an equal volume per kilogram dose. Drugs were randomized in ordered blocks of 10, each with 5 etomidate and 5 midazolam vials randomly distributed throughout the block. These vials were then placed in the ED drug vending system labeled as procedural sedation study drug and identified with a study code number that corresponded to the patient enrollment number. Once a patient was enrolled, nursing staff obtained the medication from the drug vending system by removing the lowest number of vials remaining. Both medications are clear and colorless, so the patients and ED staff administering the medication were blinded to their identity.

All data were recorded on a standardized data collection

sheet by the physician performing the sedation. Data collected included date, name, age, medical record number, weight, sex, emergency medicine attending, procedure type, preprocedure analgesia, procedural analgesia, sedation study drug number, number of doses of study drug given, time of initial study drug administration, time when patient’s recovery score was 8 or higher, time when patient disposition was determined, quality of sedation, side effects of sedation medication, whether any reversal agent was administered, whether the procedure was successful, and whether the patient recalled the procedure [3,6,7]. All data sheets were then given to the emergency medicine patient enrollers who filed them in a locked file cabinet for retrieval by the study investigators. Any data missing from the data sheet were obtained by the study investigators through chart review. All data were then entered into an anonymous database for statistical analysis and the data sheets were destroyed.

Our null hypothesis was that etomidate would not decrease the total T1 and the total T2 when compared with midazolam. Our secondary hypotheses were that etomidate would provide the same quality of sedation and procedure success rate, fewer or comparable side effects, an equal number of medication doses for procedure success, and similar amnestic properties to midazolam.

Measurements

Primary outcome measures were T1 and T2. T1 was defined as the time from the administration of the first dose of study drug until the patient’s modified Aldrete Recovery Score was 8 or higher [3,6,7]. T2 was defined as the time from the administration of the first dose of study drug until the time the patient was either admitted to the hospital or discharged home. Secondary outcome measures were the number of doses of study drug required, quality of sedation (scale, 1-10), complications, procedure success, and patient recall.

Table 2 Primary outcome data
	Etomidate	Midazolam	P
T1	15 F 11	32 F 16	b.001
T2	121 F 73	111 F 96	.70
Mean in minutes F SD.

Fig. 1 Patient flow diagram.

Data analysis

A prestudy power analysis assuming a 10% reduction in sedation time to be clinically important and a power of 80% indicated that the study would require 40 patients in each study arm. An off-site independent statistician performed the final statistical analysis using Systat 8.0 statistical software package (Systat, Point Richmond, CA). Quantita- tive variables were analyzed using independent t tests, whereas qualitative variables were analyzed using 2-way v²

Table 1 Demographics and secondary outcome measures

Etomidate Midazolam P

analysis. P values less than .05 were considered significant. The study was unblinded at the midpoint to review data and ensure proper compliance.

Results

Patient enrollment was discontinued after midpoint data analysis revealed statistical significance for our primary hypothesis. A total of 45 patients was enrolled, 24 in the etomidate group and 21 in the midazolam group (Fig. 1). One patient in the etomidate group was excluded because of protocol violations (a fourth dose of study drug was administered). There was no statistical difference in demo- graphic data or analgesia dose between the 2 groups (Table 1). The mean T1 for patients in the etomidate group was 15 minutes (SD, 10.97) whereas that for the midazolam group was 32 minutes (SD, 16.13) ( P b .001). The mean T2 in the etomidate group was 121 minutes (SD, 73.28) whereas that for the midazolam group was 111 minutes

(SD, 96.36) ( P = .700) (Table 2).

The mean number of doses of etomidate was 2.04, with a median of 2, whereas that of midazolam was 2.14, also with a median of 2 ( P = .94). The mean quality of sedation for patients receiving etomidate was 7.91 (SD, 1.53), with a median of 8, whereas that for patients receiving midazolam was 7.48 (SD, 2.89), also with a median of 8 ( P = .57). Successful reduction of either a fracture or dislocation occurred in 42 of 45 patients (93%). Midazolam was the sedation medication used in all unsuccessful Reduction attempts (86%). Recollection of the reduction procedure was expressed by 22% of patients in the etomidate group and by no patient in the midazolam group ( P = .04) (Table 1).

Age	43.3 F 23	53.4 F 26	.17
Men	16 (67%)	16 (76%)	.48
Shoulder dislocation	10 (42%)	12 (57%)
Hip dislocation	7 (30%)	5 (24%)
Elbow dislocation	2 (8%)	1 (5%)
Knee dislocation	0	1 (5%)
Other dislocations	1 (4%)	0
Fracture fentanyl dose (lg)	4 (17%) 87.5 F 71	2 (10%) 102.4 F 54	.43
No. of doses

Sedation complications were seen in 26% of the patients receiving etomidate and 24% of patients receiving mid- azolam ( P = .690) (Table 3). Etomidate complications included myoclonus, desaturation, and agitation. Midazolam complications included desaturation, apnea, and hypoten-

Table 3 Complications

mean	2.04	2.14	.94	Symptom	Etomidate (n)	Midazolam (n)
median	2.0	2.0		Myoclonus	4	0
Quality of sedation				Desaturation	2	5
mean	7.91 F 1.5	7.48 F 2.9	.57	Apnea	0	1
median	8.0	8.0		Hypotension	0	1
Reduction success	24 (100%)	18 (86%)	.06	Agitation	1	0
Patient recall	4 (22%)	0	.04	Total complications	7	7
Side effects	7 (29%)	5 (24%)	.69	No. of patients	6 (26%)	5 (24%)

sion. Most complications resolved on their own; however, a few patients required minor physical stimulation, increased oxygen flow, or an increase in the rate of intravenous fluids. Two patients required Bag valve mask ventilation because of Oxygen desaturation (one from each study group). The patient in the etomidate group had an oxygen saturation of 90%, whereas that in the midazolam group had an oxygen saturation of 68%. No patient was hospitalized as a result of any complication and no reversal agent was required (etomidate does not have a reversal agent).

Discussion

We found that etomidate compared with midazolam for procedural sedation provides a significant reduction in T1 while providing equal sedation quality without an increase in necessary dosage or side effects. These results correlate with those described previously by Burton et al as well as others who have found etomidate to be a safe and effective agent for use in procedural sedation [5-7,9,12]. The reduction in T1 was so significant with etomidate ( P b.001) that our trial was stopped after midpoint data analysis because significance for our primary hypothesis had already been achieved. Reduced T1 decreases the amount of direct observation required by the nursing staff. This, in turn, frees them to care for other patients, leading to increasED patient flow and ED efficiency. We also chose to examine T2 because the benefit of a sedative agent with a fast recovery rate is both decreased use of ED resources as well as a theoretical reduction in total length of ED stay. Only one other author has addressed this issue with etomidate. Dickinson et al [16] retrospectively reviewed 53 pediatric ED patients who received etomidate as a sedative agent for fracture reduction. They found that the mean time from administration of the first dose of etomidate to patient discharge was 94 minutes but did not compare this to patients receiving other sedative agents. Interestingly, although we found a significant reduction in T1, we did not find a significant difference in T2. Although this may seem contradictory, there are a number of possible explanations for these findings. First, these results may reflect the fact that we ended the trial early. If we had continued data collection through all 80 patients, a difference in T2 between the 2 groups may have been achieved. Second, there are many other biases that can occur at a busy medical center that affect T2 other than T1. Patients, postreduction, have a lower acuity now that they are stable and presumably more comfortable. There is often a queue of patients waiting for radiographs and those waiting for postreduction films may be assigned a lower priority. Lastly, interactions with consulting services can often delay patient disposition because there are many layers of approval and coordination required, especially

when the patient is being admitted to the hospital.

Among the secondary end points of this study, we found that etomidate may also have a lower failure rate when used for the reduction of joint dislocations or long bone fractures.

We obtained a 100% success rate using etomidate compared with only 86% using midazolam. These results are better than those described by most authors who report success rates of 89% to 93% for etomidate but are similar to those by Dursteler and Wightman [8] in a case report of their facility’s use of etomidate for a patient with a difficult hip dislocation that was reduced successfully 100% of the time when etomidate was used [6,7,12,16]. Our results should be viewed with caution, however, because this study was not designed or powered to specifically address this hypothesis and there are other factors such as physician experience or fracture complexity that also affect reduction success.

Our data also indicate that etomidate has less amnestic properties than midazolam because 22% of our patients reported recollection of the procedure. Prior trials have described rates of amnesia of approximately 93% to 96% using the same dosage of etomidate; however, in one trial, up to 27% of these patients experienced partial recall of the procedure with only 4% experiencing complete recall [6,12]. Burton et al [7] did not examine amnesia in their comparison of etomidate and midazolam, instead they used a visual analog scale for pain assessment in all patients after they recovered fully from their sedation and found no statistical difference between the 2 medications. Taken as a whole, it appears that etomidate may produce less amnesia but is as sedating and does not result in increased pain during the procedure when compared with midazolam.

Limitations

The major limitation of this study was the small number of pediatric patients enrolled. Patients younger than 10 years were not included because the packaging insert for etomidate does not recommend its use in patients younger than 10 years and there are limited data as to safety, dosing, or reliability in this age group [12,13,17]. Of those older than 10 years, there were only 5 patients younger than 16 years enrolled. With this limited sample size, we cannot come to any conclusion about etomidate’s use for proce- dural sedation in children.

Also, a data sheet that details each patient’s disposition (admission vs discharge), whether social work was involved, whether consulting services were involved, and what type of radiographic studies were still required before disposition would be ideal. Future directions would include trials specifically directed at variables affecting disposition, pedi- atric patients of all ages, etomidate’s amnestic properties at procedural sedation doses, or, more importantly, a patient’s pain assessment during the procedure.

Conclusions

In summary, etomidate compared with midazolam for procedural sedation provides a significantly reduced T1 without a reduction in T2. Etomidate is also as safe as and

possibly more effective than midazolam while providing equal sedation quality and less amnesia.

Acknowledgments

We thank Susan Kraus for statistical analysis and Mohani Ramphal for administrative support.

References

1. Reeves JG, Glass PS, Lubarsky DA. Nonbarbiturate intravenous anesthetics. In: Miller RD, editor. Anesthesia. 5th ed. Philadelphia (Pa)7 Churchill Livingstone; 2000. p. 245 - 9.
2. Gross JB, Bailey PL, Caplan RA, et al. Practice guidelines for sedation and analgesia by non-anesthesiologists: a report by the American Society of Anesthesiologists Task Force on Sedation and Analgesia by Non-Anesthesiologists. Anesthesiology 1996;84(2):459 - 71.
3. Ducharme J. Procedural sedation in the ED: how, when, and which agents to choose. Emerg Med Pract 2000;2(6):1 - 20.
4. Sokolove PE, Price DD, Okada P. The safety of etomidate for emergency rapid sequence intubation of pediatric patients. Pediatr Emerg Care 2000;16(1):18 - 21.
5. Levine BJ, Neil BJ, Vickers SM, Rosenbaum RA, et al. The use of etomidate for procedural sedation in the emergency department (abstract). Acad Emerg Med 2001;8(5):243.
6. Ruth WJ, Burton JH, Bock AJ. Intravenous etomidate for procedural sedation in emergency department patients. Acad Emerg Med 2001; 8(1):13 - 8.
7. Burton JH, Bock AJ, Strout TD, et al. Etomidate and midazolam for reduction of Anterior shoulder dislocation: a randomized, controlled trial. Ann Emerg Med 2002;40(5):496 - 504.
8. Dursteler BB, Wightman JM. Etomidate-facilitated hip reduction in the emergency department. Am J Emerg Med 2000;18(2):204 - 8.
9. Doenicke A. Etomidate, a new intravenous hypnotic. Acta Anaes- thesiol Belg 1974;25(3):307 - 15.
10. Famewo CE, Odugbesan CO. Clinical trial of etomidate. Can Anaesth Soc J 1977;24(1):35 - 8.
11. Giese JL, Stanley TH. Etomidate: a new intravenous anesthetic induction agent. Pharmacotherapy 1983;3:251 - 8.
12. Vinson DR, Bradbury DR. Etomidate for procedural sedation in emergency medicine. Ann Emerg Med 2002;39(6):592 - 8.
13. McDowall RH, Scher CS, Barst SM. Total intravenous anesthesia for children undergoing brief diagnostic or Therapeutic procedures. J Clin Anesth 1995;7:273 - 80.
14. Stockham RJ, Stanley TH, Pace NL, et al. Fentanyl pretreatment modifies anaesthetic induction with etomidate. Anaesth Intensive Care 1988;16(2):171 - 6.
15. Micromedix Healthcare Series. 2001;110:12.
16. Dickinson R, Singer AJ, Carrion W. Etomidate for pediatric sedation prior to fracture reduction. Acad Emerg Med 2001;8(1):74 - 7.
17. Holdcroft A, Morgan M, Whitwam JG, Lumley J. Effect of dose and premedication on induction complications with etomidate. Br J Anaesth 1976;48:199 - 205.