• 
• 

Three Part Question

In [patients presenting with acute agitation in an emergency setting], is the use of [benzodiazepines as a single agent, or in combination with an antipsychotic] [more optimal in producing effective and safe sedation]?

Clinical Scenario

A 45-year-old male presents to the Emergency Department with an Altered Mental State. He is acutely agitated and aggressive, and attempts at verbal de-escalation are unsuccessful. A decision to pharmacologically sedate is made. As benzodiazepines are commonly used first line to sedate acutely agitated patients, you wonder whether using them as a single agent, or in combination with an antipsychotic, will result in more effective (faster and minimal need for re-sedation) and safer (fewer adverse events) sedation.

Search Strategy

The following search strategy was used to find Medline publications via the PubMed database (1980-2025):
(((((((("altered mental state"[Title/Abstract] OR "psychosis"[Title/Abstract] OR "psychotic"[Title/Abstract] OR "agitation"[Title/Abstract] OR "aggression"[Title/Abstract] OR "delirium"[Title/Abstract]) AND "tranquillisation"[Title/Abstract]) OR "sedation"[Title/Abstract]) AND "benzodiazepines"[Title/Abstract]) OR ("chlordiazepoxide"[Title/Abstract] OR "midazolam"[Title/Abstract] OR "diazepam"[Title/Abstract] OR "clonazepam"[Title/Abstract] OR "nitrazepam"[Title/Abstract]) OR "temazepam"[Title/Abstract] OR "estazolam"[Title/Abstract] OR "alprazolam"[Title/Abstract] OR "lorazepam"[Title/Abstract] OR "oxazepam"[Title/Abstract]) AND "antipsychotic"[Title/Abstract]) OR ("haloperidol"[Title/Abstract] OR "chlorpromazine"[Title/Abstract] OR "clozapine"[Title/Abstract] OR "risperidone"[Title/Abstract] OR "olanzapine"[Title/Abstract] OR "aripriprazole"[Title/Abstract] OR "quetiapine"[Title/Abstract] OR "benperidol"[Title/Abstract] OR "prochlorperazine"[Title/Abstract] OR "typical antipsychotic"[Title/Abstract] OR "atypical antipsychotic"[Title/Abstract] OR "typical antipsychotics"[Title/Abstract] OR "atypical antipsychotics"[Title/Abstract] OR "first generation antipsychotic"[Title/Abstract] OR "second generation antipsychotic"[Title/Abstract] OR "first generation antipsychotics"[Title/Abstract] OR "second generation antipsychotics"[Title/Abstract] OR "neuroleptic"[Title/Abstract]) OR "neuroleptics"[Title/Abstract] OR "antipsychotics"[Title/Abstract] OR "sedative agent"[Title/Abstract] OR "sedative agents"[Title/Abstract]) AND "emergency"[Title/Abstract]) AND ((excludepreprints[Filter] OR medline[Filter]) AND (meta-analysis[Filter] OR randomizedcontrolledtrial[Filter] OR systematicreview[Filter]) AND (humans[Filter]) AND (english[Filter]))

Exclusion criteria: Preprint publications, Non-English publications, Paediatric studies (<18 years)

Search Outcome

172 publications were identified. Following screening of Titles and Abstracts, 10 papers underwent a full-text review.

Studies were included if they were RCTs comparing the use of benzodiazepines alone and with an antipsychotic, in acutely agitated patients in the Emergency Department (ED), minimally comparing parameters relating to time-effectiveness of sedation, and adverse effects of the drugs. Studies focusing on patients with particular diagnoses only e.g. schizophrenia were excluded.

Four publications were found to be directly relevant. Amongst the 10 was a systematic review and meta-analysis comprising seven studies, of which one other RCT was found to be directly relevant. The five publications are reflected below.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Geoffrey K. Isbister, Leonie A. Calver, Colin B. Page, Barrie Stokes, Jenni L. Bryant, Michael A. D 2010 Australia	Patients ≥18 years, presenting at any time (24-7) to an Autralian urban ED between August 2008 to July 2009, with violent and acute behavioral disturbance requiring both physical restraint and parenteral sedation, as assessed by ED nursing or medical staff. Patients received an IM injection of either: (i) 10mg droperidol, (ii) 10mg midazolam, or (iii) 5mg droperidol/5mg midazolam	Double-blinded RCT: n=91; 33 received droperidol-only, 29 received midazolam-only, 25 received combination treatment	Effectiveness of Sedation	No difference in median duration of violent and acute behavioral disturbance: 24min (IQR 13-35min) for midazolam, 25min (IQR 15-38min) for combination. On AMSS, combination treatment caused deeper and more unpredictable sedation than other groups. Patient or staff injuries and further episodes of disturbance did not differ between groups.	Duration of violent and acute behavioral disturbance was used as primary outcome instead of time-effectiveness with a sedation scale (secondary outcome). Unclear whether the former is more objective given the proxy measure of sedation. Wide IQRs suggest inconsistent sedation within groups. Incomplete stratification of patients intoxicated by alcohol prior to enrolment may confound adverse event data regarding respiratory depression with midazolam-only.
			Requirement for Further Sedation	Additional sedation was required in 18 (62%; 95% CI 42% to 79%) midazolam patients, and 12 (41%; 95% CI 24% to 61%) combination group patients.
			Adverse Events	There were eight adverse effects for midazolam (28%; 95% CI 13% to 47%) and two for combination (7%; 95% CI 1% to 24%). In alcohol-intoxicated patients, midazolam-only was found to be associated with oversedation and more respiratory depression.
Esther W Chan, David M Taylor, Jonathan C Knott, Georgina A Phillips, David J Castle, David C M 2013 Australia	Adult (18-65 years) patients determined by ED registrars or consultants to be requiring IV drug sedation for acute agitation from in three metropolitan EDs from August 2009 to March 2011, were given: (i) 5mg IV droperidol, (ii) 5mg IV olanzapine bolus, or (iii) control: IV saline solution Each of these were immediately followed by incremental IV midazolam boluses (2.5 to 5mg) until sedation was achieved.	Multi-center, randomised, double-blinded, placebo-controlled, double-dummy RCT: n=336	Effectiveness of Sedation	Difference in median for time to sedation between saline control (ie midazolam-only) and droperidol group was 4 minutes (95% confidence interval [CI] 1 to 6 minutes). At any point, patients in droperidol group were around 1.6 times more likely to be sedated compared with control; droperidol group hazard ratio was 1.61 (95% CI 1.23 to 2.11).	Primary objective of study was to compare IV droperidol or olanzapine as an adjunct to midazolam, hence data extrapolated is a very indirect and likely imprecise way of studying effect of single-agent benzodiazepines (saline control group) versus in combination (droperidol group) via addition of incremental midazolam boluses. No objective measure of initial state of agitation. Need for parenteral drug sedation and incremental IV midazolam boluses relied entirely on different individual clinicians‘ judgement. Dosing variability due to unstandardised sedation protocol possibly confounded outcomes. Little explanation as to whether Soluvit N as double-dummy placebo-olanzapine is completely inert or will have any direct biological impact thus confounding results. Limited conclusions drawn regarding prolonged QTc’s association with adverse events due to many not receiving ECGs, especially in midazolam-only group. Possible selection bias introduced as ECG testing not standardised but determined by clinical judgement.
			Requirement for Further Sedation	Patients in droperidol group required less rescue or alternative drug use after initial sedation compared to saline control.
			Adverse Events	The groups’ adverse event profiles and lengths of stay did not differ.
S A Bieniek, R L Ownby, A Penalver, R A Dominguez 1998 USA	Acutely agitated patients aged 18-50 presenting in Psychiatric Emergency services of a large hospital, with psychometric evidence of agitated or aggressive behaviour*, meeting clinical criteria for use of chemical restraints. *defined as achieving a total score of ≥4, with 2 or more on at least one item, of the Overt Aggression Scale (OAS). Patients received an IM injection of either: (i) 2mg lorazepam or (ii) 5mg haloperidol/2mg lorazepam	Double-blinded RCT: n=20; 11 receiving lorazepam-only and 9 receiving combination treatment	Effectiveness of Sedation	Significantly greater percentage of subjects on combined treatment improved on OAS and visual analogue scale 60 minutes after dosing (p<0.05). Kaplan-Meier survival analyses showed significant differences between groups in survival curves plotted for the entire study period (p<0.05). However, there was no difference recorded between both groups on the CGI severity scale. The equivalent Kaplan-Meier analysis for CGI severity also was not significant (log rank statistic = 0.00, df=1, p=0.980).	Very small sample size meant results obtained were not statistically significant. Single centre municipal hospital study likely not representative of EDs in other settings. External validity of study called into question due to exclusion of multiple patient groups including patients with psychopathology due to suspected General Medicine conditions. Possible data-driven bias due to post-hoc change of definition of “clinically-significant improvement” on visual analogue scale. Two patients with continued severe symptoms given second injection outwith key 60-minute observation period (time period with strongest evidence of group differences), potentially confounding results. Patients on combined treatment improved on some specific measures but not others, calling conclusion’s validity into question.
			Requirement for Additional Sedation	Only two patients required a second injection, both from lorazepam-only group.
			Adverse Events	No adverse events observed or reported in either group.
J Battaglia, S Moss, J Rush, J Kang, R Mendoza, L Leedom, W Dubin, C McGlynn, L Goodman 1997 USA	Psychotic, agitated, aggressive adult patients (73 men, 25 women) presenting to EDs across five university or general hospitals in an 18 month period, with a score of ≥5 on a scale of 1 to 7 for three or more of 11 psychosis/anxiety items on the Brief Psychiatric Rating Scale (BPRS), and with behavioural dyscontrol to the extent that they were capable of harming themselves or others. Patients received an IM injection of either: (i) 2mg lorazepam, (ii) 5mg haloperidol, or (iii) combination 2mg lorazepam/5mg haloperidol	RCT: 24h double-blind, parallel, multicenter, randomised prospective study n=98 patients; 31 lorazepam-only, 35 haloperidol-only,32 combination	Effectiveness of Sedation	Significant (p<0.05) mean differences on the Agitated Behaviour Scale or ABS (hour 1) and modified BPRS or MBPRS (hours 2 and 3) suggest that tranquilisation was most rapid in patients receiving combination treatment.	Potential confounding of outcome interpretation due to “falling asleep” being considered a therapeutic endpoint, which may not reflect agitation resolution in itself. Study only evaluated patients for 24h; no reporting of patients’ conditions post-incident or presence of relapses.
			Requirement for Additional Sedation	91% of combination group received three or fewer doses, with this occurring for only 74% of the lorazepam-only group.
			Adverse Events	Incidence of EPS symptoms in lorazepam-only group was 3%, compared to that of combination group at 6%. However, Fisher's exact test revealed no statistically significant differences between groups.
David McD Taylor, Celene Y L Yap, Jonathan C Knott, Simone E Taylor, Georgina A Phil 2016 Australia	Patients aged 18 to 65 presenting to two metropolitan EDs between October 2014 to August 2015, requiring IV sedation for acute agitation, were given IV boluses of either: (i) 10mg droperidol, (iii) 10mg olanzapine, or (iii) 5mg midazolam/droperidol 5mg	Double-blinded, triple-dummy RCT: n=349 patients; 117 droperidol-only, 124 olanzapine-only, 120 combination	Effectiveness of Sedation	10 minutes after first dose, signicantly more patients in the combined group were adequately sedated (25.4% difference, with 95% CI 12.7% to 38.3%) compared with olanzapine-only group. Difference in median time to sedation between combined group and olanzapine-only was 6 minutes (95% CI 3 to 7).	No midazolam-only arm, preventing effective comparison given different benzodiazepines for monotherapy and combination. Study enrolment not based on scoring system but solely patient and staff safety considerations. Much of study protocol taken from prior study by Chan et al2 with its own limitations (see above). Possible confounding with percentage of patients having taken drugs, alcohol, and sedatives prior to enrolment. Four protocol violations due to age still included in data analysis (though re-analysis post-exclusion was also conducted).
			Requirement for Additional Sedation	Patients in combined group required fewer extra doses or alternative drugs to be adequately sedated.
			Adverse Events	Groups’ adverse event rates did not differ.

Comment(s)

Current evidence indicates that combination therapy with a benzodiazepine and an antipsychotic may be more effective than benzodiazepine monotherapy in achieving faster sedation and reducing the need for re-sedation in acutely agitated patients in ED settings. Existing data also suggests no significant difference in the frequency and type of adverse events between groups, with respiratory adverse events most commonly occurring. However, the evidence remains inconclusive due to limitations such as small sample sizes and significant heterogeneity in how agitation and sedation outcomes are defined and measured across different studies. Notably, baseline agitation levels varied considerably across studies and in numerous cases were not quantified by standardised scoring systems. Disparate methods were used to determine sedation effectiveness, including time to sedation, proportion sedated at any given time, and improvements in agitation or sedation scores over time. The method of establishing treatment endpoints, criteria for re-sedation, total dosage of medication given, and routes of administration differed as well. To strengthen future research, there is a clear need for homogeneity in measuring and reporting outcomes, as well as larger, adequately-powered trials.

Clinical Bottom Line

Given the comparable incidence of adverse events between treatment groups, and the observed benefits of faster sedation with a reduced need for re-sedation, it appears that the benzodiazepine-antipsychotic combination therapy is more effective than benzodiazepine monotherapy in sedating acutely agitated patients in ED settings.

References

1. Geoffrey K. Isbister, Leonie A. Calver, Colin B. Page, Barrie Stokes, Jenni L. Bryant, Michael A. D Randomized Controlled Trial of Intramuscular Droperidol Versus Midazolam for Violence and Acute Behavioral Disturbance: The DORM Study Annals of Emergency Medicine October 2010, Pages 392-401.e1
2. Esther W Chan, David M Taylor, Jonathan C Knott, Georgina A Phillips, David J Castle, David C M Intravenous droperidol or olanzapine as an adjunct to midazolam for the acutely agitated patient: a multicenter, randomized, double-blind, placebo-controlled clinical trial Annals of Emergency Medicine (Mosby Inc.) 2013;61(1):72–81
3. S A Bieniek, R L Ownby, A Penalver, R A Dominguez A double-blind study of lorazepam versus the combination of haloperidol and lorazepam in managing agitation Pharmacotherapy 1998 Jan-Feb;18(1):57-62
4. J Battaglia, S Moss, J Rush, J Kang, R Mendoza, L Leedom, W Dubin, C McGlynn, L Goodman Haloperidol, lorazepam, or both for psychotic agitation? A multicenter, prospective, double-blind, emergency department study American Journal of Emergency Medicine 1997 Jul;15(4):335-40
5. David McD Taylor, Celene Y L Yap, Jonathan C Knott, Simone E Taylor, Georgina A Phil Midazolam-Droperidol, Droperidol, or Olanzapine for Acute Agitation: A Randomized Clinical Trial Annals of Emergency Medicine (Elsevier) 2017 Mar;69(3):318-326.e1.