a b s t r a c t

Objective: Emergency physicians often need to control agitated patients who present a danger to themselves and hospital personnel. Commonly used medications have limitations. Our primary objective was to compare the time to a defined reduction in agitation scores for ketamine versus benzodiazepines and haloperidol, alone or in combination. Our secondary objectives were to compare rates of medication redosing, vital sign changes, and adverse events in the different treatment groups.

Methods: We conducted a single-center, prospective, observational study examining agitation levels in acutely agitated emergency department patients between the ages of 18 and 65 who required sedation medication for acute agitation. Providers measured agitation levels on a previously validated 6-point sedation scale at 0-, 5-, 10-, and 15-min after receiving sedation. We also assessed the incidence of adverse events, repeat or rescue med- ication dosing, and changes in vital signs.

Results: 106 patients were enrolled and 98 met eligibility criteria. There was no significant difference between groups in initial agitation scores. Based on agitation scores, more patients in the ketamine group were no longer agitated than the other medication groups at 5-, 10-, and 15-min after receiving medication. Patients receiving ketamine had similar rates of redosing, changes in vital signs, and adverse events to the other groups.

Conclusion: In highly agitated and violent emergency department patients, significantly fewer patients receiving ketamine as a first line sedating agent were agitated at 5-, 10-, and 15-min. Ketamine appears to be faster at con- trolling agitation than standard emergency department medications.

(C) 2017

Introduction

Background

Emergency physicians often need to control violent, psychotic, or in- toxicated patients who present a danger to themselves and hospital personnel. Rational reasoning, bargaining, shows of force with security guards, and even physical restraint are sometimes ineffective in control- ling the acutely agitated patient. Chemical sedation is sometimes neces- sary to prevent injuries to patients and staff, and to allow safe medical evaluation and treatment. Benzodiazepines and typical antipsychotics such as haloperidol, the most commonly used Sedative agents, have lim- itations including slow onset, respiratory depression, and variability in Clinical response [1,2].

? This work was presented at the following meetings:-social media and Critical Care Conference, Chicago, IL, June 2014-Society for Academic Emergency Medicine Annual Meeting, San Diego, CA May 2014.

* Corresponding author at: Box 359702, 325 Ninth Ave, Seattle, WA 98104-2499, USA.

E-mail address: [email protected] (J. Riddell).

Importance

While several recent studies have shown the efficacy of ketamine for sedation in the prehospital setting [3,4] and as a Rescue medication in emergency department (ED) patients who failed previous sedation at- tempts [5], there is limited data evaluating the effectiveness of ketamine as a first line agent for sedating agitated patients in the ED.

Goals of this investigation

The goal of our study was to compare the effectiveness and safety of ketamine to standard sedatives in agitated ED patients. Specifically, our primary objective was to compare the time to a defined reduction in ag- itation scores for ketamine versus benzodiazepines and haloperidol, alone or in combination. Our secondary objective was to assess the inci- dence of adverse events, repeat or rescue medication dosing, and chang- es in vital signs. We hypothesized that ketamine would produce the desired clinical effect in a shorter time, with similar side effects, stable hemodynamics, and less repeat dosing.

http://dx.doi.org/10.1016/j.ajem.2017.02.026

0735-6757/(C) 2017

Table 1 Study group demographics.
	Ketamine	Midazolam	Lorazepam	Haloperidol	Combo	p
	(n=24)	(n=17)	(n=33)	(n=14)	(n=10)
Median age (range)	29 (19-58)	43 (18-51)	43 (20-63)	44 (21-58)	40.5 (21-58)	0.033
Male sex, no. (%)	19 (79.2%)	18 (94.7%)	19 (57.6%)	11 (78.6%)	9 (90.0%)	0.026
Race, no. (%)
African American	3 (12.5%)	1 (5.3%)	5 (15.2%)	1 (7.1%)	1 (10.0%)	0.931
Asian	1 (4.3%)	0 (0.0%)	1 (3.1%)	1 (7.1%)	0 (0.0%)
Hispanic	10 (41.7%)	10 (52.6%)	13 (39.4%)	8 (57.1%)	7 (70.0%)
White	10 (43.5%)	7 (36.8%)	13 (40.6%)	4 (28.6%)	2 (20.0%)
Drug use, no. (%) Yes	13 (54.2%)	12 (63.2%)	26 (78.8%)	12 (85.7%)	6 (60.0%)	0.168
No	11 (47.8%)	7 (36.8%)	7 (21.9%)	2 (14.3%)	4 (40.0%)
Unknown
Alcohol use, no. (%) Yes	8 (33.3%)	8 (42.1%)	8 (24.2%)	5 (35.7%)	2 (20.0%)	0.365
No	12 (52.2%)	7 (36.8%)	14 (21.9%)	5 (35.7%)	2 (20.0%)
Unknown	4 (17.4%)	4 (21.1%)	11 (34.4%)	4 (28.6%)	6 (60.0%)
Prior psychiatric visits, no. (%)
Yes	7 (30.4%)	7 (36.8%)	17 (53.1%)	7 (50.0%)	5 (50.0%)	0.459

Methods

Study design

This was a single-center, prospective, observational study examining agitated ED patients requiring medication for sedation. IRB approval was obtained from our local institution.

Setting and participants

The study took place at a single urban level one trauma center ED with an annual census of 115,000 visits. A convenience sample was en- rolled from May 2013 to January 2015.

acutely agitated patients between the ages of 18 and 65 who re- quired chemical sedation for acute agitation according to an emergency medicine resident or attending physician were eligible for enrollment. Pregnant women, prisoners and persons in police custody were exclud- ed. Also excluded were those triaged to a low acuity zone of the ED that did not have continuous cardiorespiratory monitoring. Our study popu- lation included only those patients so severely agitated that they re- quired care in a high acuity treatment area with available cardiorespiratory monitoring.

Variables

We initially defined four groups for comparison: ketamine, benzodi- azepines, haloperidol, and a benzodiazepine plus haloperidol. Because the benzodiazepine group was comprised entirely of midazolam and

lorazepam, we considered them separately, and compared a total of five groups. Due to the observational nature of this study, medication dosages were not uniform. Current published dosage recommendations for these medications include: ketamine 4-6 mg/kg intramuscular or 1-2 mg/kg intravenous (IV) [6-7], haloperidol 5-10 mg IM [8], mid- azolam 5-10 mg IM [9] or 5 mg IV¹⁰, lorazepam 1-2 mg IV or IM [11].

Outcomes

Providers measured the patients’ agitation on a previously validated 6-point sedation scale that was developed to monitor changes in agita- tion in ED patients [10,12]. In keeping with previous studies, adequate sedation was defined as an agitation score of <= 2 (see Appendix). Agita- tion was recorded by the treating physician prior to medication admin- istration and at 5-, 10-, and 15-min after medication administration. Providers also recorded the time at which they thought adequate seda- tion had been achieved.

The incidence of adverse events, repeat or rescue medication dosing, and changes in vital signs were retrospectively abstracted from the elec- tronic health record.

Measurement

ED physicians completed a structured data collection form (see Ap- pendix). One trained author (AT) reviewed the relevant portion of each patient’s medical record for the index visit and any return visits to our ED within 7 days. The abstractor was blinded to the study hypoth- esis until abstraction was complete. The abstractor received two 90-

Table 2 Study group dispositions.
	Ketamine	Midazolam	Lorazepam	Haloperidol	Combo	p
Disposition, no. (%)
Hospital admission	14 (58.3%)	10 (52.6%)	15 (46.9%)	7 (50.0%)	3 (30.0%)	0.944
Psychiatric admission	8 (33.3%)	1 (5.3%)	2 (6.3%)	1 (7.1%)	1 (10.0%)
Discharge home	2 (8.3%)	8 (42.1%)	15 (46.9%)	6 (42.9%)	6 (60.0%)
Admission location, no. (%) Intensive care unit	7 (29.2%)	5 (26.3%)	2 (6.3%)	3 (21.4%)	1 (10.0%)	0.412
Stepdown unit	1 (4.2%)	3 (15.8%)	4 (12.5%)	2 (14.3%)	0 (0.0%)
Telemetry floor	1 (4.2%)	0 (0.0%)	3 (9.4%)	1 (7.1%)	0 (0.0%)
Unmonitored floor	5 (20.8%)	2 (10.5%)	6 (18.8%)	1 (7.1%)	1 (10.0%)
LOS ED, minutes (SD)	332.7 (242.8)	402.5 (304.9)	356.2 (298.4)	301.1 (193.3)	388.6(394.4)	0.865
LOS ICU, hours (SD)	157.8 (131.3)	182.0 (91.2)	123.0 (-)	236 (137.2)	0.0	0.889
LOS Hospa	105.6 (97.7)	140.3 (99.4)	96.5 (106.7)	110.8 (99.0)	95.0 (44.6)	0.899
48 h bouncebackb	2 (8.3%)	2 (10.5%)	2 (6.1%)	1 (7.1%)	0 (0.0%)	0.062

^a Hospital length of stay only if admitted.

^b return ED visit within 48 h only if discharged home from the ED.

Table 3

Doses of medications administered during study.

	n	Mean dose (Mg)a				Range
		IV	IM	IN		IV	IM	IN
Ketamine (Mg/kg)	24	0.87 (n = 18)	2.97 (n = 6)	-		0.31-2.20	0.88-3.94	-
Midazolam (mg)	19	3.08 (n = 12)	2.25 (n = 4)	2.00 (n = 3)		1.00-4.00	1.00-4.00	2.00
Lorazepam (mg)	33	1.90 (n = 28)	2.40 (n = 5)	-		1.00-4.00	2.00-4.00	-
Haloperidol (mg)	14	-	5.71 (n = 14)	-		-	5.00-10.00	-
Combo (mg)	10	L: 2.00 (n = 5)	H: 5.00 (n = 10)	-		L: 2.00	H: 5.00	-
			L: 2.00 (n = 5)				L: 2.00

^a L = lorazepam, H = haloperidol, IM = intramuscular. IV = intravenous. - = medication not given.

minute training sessions. A written definition of each variable guided the data abstraction (see Appendix). Early in the course of abstraction, a second author (JR) independently abstracted a random sample of 5 charts. Out of a possible 235 data points, there were 3 items of disagree- ment (1.3%). None of the discrepancies were related to the primary or secondary outcomes involving medication, doses, sedation scores, or adverse events.

Sample size

In order to calculate our necessary sample size, we assumed that 5 min was the smallest difference in time to sedation that would be clin- ically significant. Based on an expected mean time to sedation in the standard sedation groups of 15 min, with an estimated standard devia- tion of 5 min, 80% power, and alpha of 0.05, we determined that 17 par- ticipants were needed in each group.

Analysis

All data was entered into an Excel spreadsheet (Microsoft, Redmond, WA) where basic descriptive statistics were performed. Means, medians and standard deviations were computed for continuous variables and percentages for categorical data. Bivariate analyses of categorical vari- ables by treatment were conducted using chi-squared statistics. Univar- iate analysis of variance was used for comparing continuous variables by treatment. All analyses between medications were adjusted for mul- tiple comparisons. Analyses were performed in Stata v. 14.2 (College Station, TX) and two-sided p b 0.05 was used as the criterion for statis- tical significance.

Results

During the study period, 106 patients were enrolled and 98 met eli- gibility criteria. Demographic data is presented in Table 1, with groups based on initial medication given. Those receiving ketamine were signif- icantly younger than those receiving other medications (p = 0.033) and the entire study cohort was mostly male. There were no significant dif- ferences based on race, stated use of substances, or previous psychiatric history. There were no significant differences in disposition based on study group (Table 2).

Mean doses of medications administered are presented in Table 3. There was no significant difference between groups in initial agita-

tion scores (Table 4). Most patients were “highly aroused” or “violent”

prior to receiving medication. Based on agitation scores at 5-, 10-, and

15-min after receiving medication, more patients in the ketamine group were no longer agitated than the other medication groups (p = 0.001, p <= 0.001, p = 0.032). There was no significant difference between groups in provider's reported time until agitation was subjec- tively controlled.

There was no significant difference between groups in the require- ment for subsequent redosing of sedative medication (Table 5).

There was a significant difference in the pulse rate reduction within the first hour seen with midazolam (Table 6). No other significant differ- ences in pulse rate or blood pressure were found. The single greatest in- crease in SBP in a patient given ketamine was 75 mm Hg. The patient received naloxone just prior to ED arrival from EMS.

Two patients receiving ketamine were intubated. One patient each receiving midazolam, lorazepam, haloperidol, and combo haloperidol plus benzodiazepine were intubated (Appendix).

Discussion

In this prospective evaluation of ketamine as a first line agent for se- dation of agitation emergency department patients, we found that sig- nificantly more patients receiving ketamine had their agitation controlled at all study time points. Though not powered to detect sec- ondary outcome differences, patients receiving ketamine had similar rates of redosing, changes in vital signs, and adverse events.

A vast body of literature exists supporting the use of ketamine for procedural sedation and intubation [13-16]. While the literature on its use in agitated patients is less robust, ketamine has been used to control agitation in prehospital, aeromedical, military, and ED settings [3,4, 17-22].

The only study that has evaluated the effectiveness of ketamine se- dation for agitation in the ED comes from a subgroup analysis of patients who received droperidol or midazolam and then had intramuscular ke- tamine added after the other medications were ineffective [5]. The mean time to sedation post-ketamine was 20 min, and they concluded ketamine appears to be a reasonable 3rd line agent for sedation of pa- tients with acute behavioral disturbance. Our study adds to this by sug- gesting that ketamine is effective as a first-line sedating agent. Our mean time to control of sedation was much faster. This may be ex- plained by the high number in our cohort receiving the medication in- travenously (18/24, 75%), though our time was consistent with a prehospital study that found a mean time to sedation post-intramuscu- lar ketamine of 5.5 min [4].

A recent retrospective study evaluating adverse events in 27 agitated patients receiving ketamine found 62.5% of patients required additional

Table 4

Agitation scores and time until control of agitation.

	n	0 min (mean +- SD)	5 min	10 min	15 min	Time until control (min) (mean +- SD)
Ketamine	24	4.29 (0.91)	1.25 (1.73)	0.71 (1.08)	0.79 (1.14)	6.57 (8.65)
Midazolam	19	4.58 (0.77)	2.90 (1.56)	2.58 (1.54)	1.95 (1.51)	14.95 (10.47)
Lorazepam	33	4.24 (1.06)	2.51 (1.71)	1.85 (1.58)	1.45 (1.52)	17.73 (24.78)
Haloperidol	14	4.29 (0.91)	2.79 (1.63)	2.71 (1.32)	2.14 (1.66)	13.43 (15.36)
Combo	10	4.80 (0.42)	3.60 (1.26)	2.30 (1.83)	1.10 (1.37)	23.30 (25.12)
p-Value		0.386	0.001	b0.001	0.032	0.107

Table 5

Redosing of sedative medication.

	n	Number needing rescue medications	p value
Ketamine	24	14 (58.3%)
Midazolam	19	15 (78.9%)
Lorazepam	33	26 (78.8%)
Haloperidol	14	7 (50.0%)
Combo	10	7 (70%)	0.199

sedating medication [17]. Similarly, 14/24 (58.3%) of our patients given ketamine required redosing, though there was not a significant differ- ence in redosing among groups. Ketamine is unlikely to resolve the un- derlying processes causing agitation and in this context it is used to gain rapid control of violent patients so that safe medical evaluation can pro- ceed and treatment of the underlying cause can commence.

Only 2/23 (8.7%) of our ketamine patients were intubated, which is significantly lower than in prehospital studies, with rates as high as 63% [23]. A retrospective case series of 32 emergency department pa- tients receiving ketamine for agitation reported no intubations [17]. The ED Intubation rate may be lower than prehospital studies due to ED providers’ familiarity with the dissociative properties of the medication.

In line with previous studies, which have shown median systolic blood pressure changes from +5 to +17 mm Hg and heart rate changes from 0 to +8 beats per minute [5,17], we found ketamine to have rela- tively stable hemodynamic effects. It is possible that the agitation-con- trolling properties of ketamine counteract its sympathetic effects, leading to a relatively neutral hemodynamic profile in this patient population.

Limitations

This was a single center study in a population exhibiting a high per- centage of methamphetamine abuse. Data may not be generalizable to populations with different toxicological profiles. It was not possible to randomize patients and so some selection bias may exist. It is important to note though that without randomization, groups were similar in ini- tial agitation scores, gender, race, alcohol abuse, and psychiatric history.

Table 6

Change in vital signs in first hour after medication administration (mean +- SD).

While blinded to hypotheses, physicians were not blinded to the medications patients received which may have biased their assessment of time to sedation. The retrospective collection of the secondary out- come variables carries with it the limitations inherent in chart reviews. Dosing was not uniform and varied among medications making di- rect comparisons imperfect. Mean medication doses administered were below current recommended doses for ketamine and midazolam. Our study is limited by its sample size. Although ketamine adminis- tration had similar adverse events as other sedating medications, a larg- er sample is required to reliably confirm its safety profile. The change in vital signs data is limited to what was charted by nursing staff in the first hour after medication administration. This study was not powered to

detect differences in adverse events or vital signs.

We also did not account for pre-hospital treatment. It may be possi- ble that some patients received medication prior to presenting to the ED, as is the case with a patient who received naloxone just prior to ar- rival. Prehospital providers were not able to administer ketamine at the time the study was conducted.

Conclusion

In summary, in highly agitated and violent emergency department patients, significantly fewer patients receiving ketamine as a first line sedating agent were agitated at 5-, 10-, and 15-min. Ketamine appears to be faster at controlling agitation than standard ED medications.

Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ajem.2017.02.026.

Funding

This work was supported by the University of California San Francisco Clinical & Translational Science Institute.

Acknowledgements

The authors thank Svetlana Bagdasarov for her help in the adminis- tration of this study.

References

1. Rossi J, Swan MC, Isaacs ED. The violent or agitated patient. Emerg Med Clin North Am 2010;28(1):235-56 [x].
2. Battaglia J, Moss S, Rush J, et al. Haloperidol, lorazepam, or both for psychotic agita-

   Ketamine

   Initial vitals^a Mean change within first

   hour^b

   p value

   tion? A multicenter, prospective, double-blind, emergency department study. Am J Emerg Med 1997;15(4):335-40.

   Le Cong M, Gynther B, Hunter E, Schuller P. ketamine sedation for patients with acute agitation and psychiatric illness requiring aeromedical retrieval. Emerg Med

   Pulse rate	102.3 (25.7)	92.7 (21.7)	0.087	J 2012;29(4):335-7.
   Systolic blood	139.8 (18.7)	140.3 (24.0)	0.834	[4] Cole JB, Moore JC, Nystrom PC, et al. A prospective study of ketamine versus haloper-
   pressure				idol for severe prehospital agitation. Clin Toxicol 2016;54(7):556-62.
   Midazolam				[5] Isbister GK, Calver LA, Downes MA, Page CB. Ketamine as rescue treatment for diffi-
   Pulse rate	110.4 (30.0)	96.9 (22.9)	0.026	cult-to-sedate severe acute behavioral disturbance in the emergency department.
   Systolic blood	129.3 (18.9)	123.8 (24.8)	0.226	Ann Emerg Med 2016;67(5) (581-587.e1).
   pressure				[6] Green SM, Roback MG, Kennedy RM, Krauss B. Clinical practice guideline for emer-
   Lorazepam				gency department ketamine dissociative sedation: 2011 update. Ann Emerg Med

   ^a Pulse rates are in beats/minute. Systolic blood pressures are presented in mm Hg.

   Pulse rate	114.7 (32.5)	109.9 (25.0)	0.377
   Systolic blood pressure Haloperidol	134.7 (26.5)	133.4 (26.5)	0.665
   Pulse rate	104.7	104.6 (20.9)	0.974
   	(14.80)
   Systolic blood pressure Combo Pulse rate	129.2 (24.7) 99.6 (19.1)	144.5 (30.3) 90.5 (12.7)	0.058 0.197
   Systolic blood pressure	140.7 (20.1)	129.0 (24.7)	0.195

   ^b This represents the mean of the largest change from initial vitals that were docu- mented anytime in the first hour after receiving sedative medication.

   2011;57(5):449-61.

   Green SM, Rothrock SG, Harris T, Hopkins GA, Garrett W, Sherwin T. Intravenous ke- tamine for pediatric sedation in the emergency department: safety profile with 156 cases. Acad Emerg Med 1998;5(10):971-6.

3. Clinton JE, Sterner S, Stelmachers Z, Ruiz E. Haloperidol for sedation of disruptive emergency patients. Ann Emerg Med 1987;16(3):319-22.
4. Calver LA, Downes MA, Page CB, Bryant JL, Isbister GK. The impact of a standardised intramuscular sedation protocol for acute behavioural disturbance in the emergency department. BMC Emerg Med 2010;10:14.
5. Knott JC, Taylor DM, Castle DJ. Randomized clinical trial comparing intravenous mid- azolam and droperidol for sedation of the acutely agitated patient in the emergency department. Ann Emerg Med 2006;47(1):61-7.
6. Battaglia J. Pharmacological management of acute agitation. Drugs 2005;65(9): 1207-22.
7. Richards JR, Derlet RW, Duncan DR. Chemical restraint for the agitated patient in the emergency department: lorazepam versus droperidol. J Emerg Med 1998;16(4): 567-73.
8. Green SM, Rothrock SG, Lynch EL, et al. Intramuscular ketamine for pediatric seda- tion in the emergency department: safety profile in 1,022 cases. Ann Emerg Med 1998;31(6):688-97.
9. Reich DL, Silvay G. Ketamine: an update on the first twenty-five years of clinical ex- perience. Can J Anaesth 1989;36(2):186-97.
10. Jabre P, Combes X, Lapostolle F, et al. Etomidate versus ketamine for rapid sequence intubation in acutely ill patients: a multicentre randomised controlled trial. Lancet 2009;374(9686):293-300.
11. Hughes S. Towards evidence based emergency medicine: best BETs from the Man- chester Royal Infirmary. BET 3: is ketamine a viable Induction agent for the trauma patient with potential brain injury. Emerg Med J 2011;28(12):1076-7.
12. Hopper AB, Vilke GM, Castillo EM, Campillo A, Davie T, Wilson MP. Ketamine use for acute agitation in the emergency department. J Emerg Med 2015;48(6):712-9.
13. Melamed E, Oron Y, Ben-Avraham R, Blumenfeld A, Lin G. The combative multitrauma patient: a protocol for prehospital management. Eur J Emerg Med 2007;14(5):265-8.
14. Hick JL, Ho JD. Ketamine chemical restraint to facilitate rescue of a combative “jump- er.”. Prehosp Emerg Care 2005;9(1):85-9.
15. Burnett AM, Watters BJ, Barringer KW, Griffith KR, Frascone RJ. Laryngospasm and hypoxia after intramuscular administration of ketamine to a patient in excited delir- ium. Prehosp Emerg Care 2012;16(3):412-4.
16. Roberts JR, Geeting GK. Intramuscular ketamine for the rapid tranquilization of the uncontrollable, violent, and dangerous adult patient. J Trauma 2001;51(5):1008-10.
17. Scaggs TR, Glass DM, Hutchcraft MG, Weir WB. Prehospital ketamine is a safe and ef- fective treatment for excited delirium in a community hospital based EMS system. Prehosp Disaster Med 2016;31(05):563-9.
18. Olives TD, Nystrom PC, Cole JB, Dodd KW, Ho JD. Intubation of profoundly agitated patients treated with prehospital ketamine. Prehosp Disaster Med 2016;31(6): 593-602.