Ketamine for Acute Pain in Opiod Users

This is the sixth in a series of blog posts on new research in emergency toxicology. The last post was about ketamine analgesia in chronic opioid users and can be found here. We deal with all sorts of poisons here in Virchester, so be prepared for anything.

In emergency medicine, we regularly encounter patients who are agitated or distressed to the point of severe distress and/or violent behaviour directed at themselves or others.. This situation is described broadly as ‘acute behavioural disturbance’ (ABD) and managing it safely is an essential part of our specialty. The typical causes are mental health disorders and substance misuse, but some cases have a “medical” aetiology, such as encephalitis or intracranial haemorrhage.

Whatever the underlying pathology, ABD is a high-risk scenario. Rapid de-escalation is required, and although some patients can be pacified verbally, the mainstay of treatment in ABD is sedative drugs. RCEM recommends intramuscular ketamine or droperidol first-line, with consideration of benzodiazepines or haloperidol if these agents are unavailable.

Here in Virchester, we see a lot of ABD. Much of it is related to cocaine use. I rarely see antipsychotics used for these patients. I suspect this is because most of us are more familiar with ketamine and benzodiazepines – and in fairness, there is good evidence supporting these drugs. However, there are situations where we might reach for antipsychotics. Patients who are agitated due to serotonergic toxicity, for example, may benefit from their antagonistic effects.

But does it have to be haloperidol? We know that this drug carries significant risks, including extrapyramidal symptoms, hypotension, and QT prolongation. It would be ideal if we could use droperidol instead , but it is not stocked in most UK emergency departments. A potential alternative is olanzapine, which has demonstrated similar efficacy to haloperidol – with fewer adverse events – in mental health settings. Helpfully, a trial was recently published in the Journal of Emergency Medicine on the use of olanzapine for ABD.

Abstract

Background & Objectives: Acute agitation is a common presentation in the emergency department (ED), often managed with haloperidol. This study aimed to compare the efficacy and safety of olanzapine versus haloperidol for the initial treatment of acute agitation in the ED. Primary outcome was adequate sedation at 15 min, defined as Altered Mental Status (AMS) score of zero or less. Secondary outcomes included adequate sedation at 30 min, need for rescue medications and reported adverse events.
Methods: This open-label, randomized controlled trial included adult patients presenting to the ED with acute agitation, defined with an AMS score ≥ 3. After taking surrogate consent from their legally authorized representative, patients were randomly assigned to receive either intramuscular (IM) olanzapine (10 mg) or IM haloperidol (5 mg).
Results: Of the total 94 patients, 47 received IM olanzapine and 47 received IM haloperidol. Similar proportions of patients were adequately sedated at 15 min (olanzapine 31.9% vs haloperidol 25.5%; relative risk [RR] – 1.25, 95% confidence interval [CI] 0.65 to 2.37; p – 0.494) and 30 min (olanzapine 61.7% vs haloperidol 48.9%; RR – 1.26, 95% CI 0.87 to 1.82; p – 0.213). The need for rescue medications was similar (olanzapine 12.7% vs. haloperidol 25.5%; RR 0.5, 95% CI 0.20 to 1.22; p 0.116). Adverse events were uncommon and similar across both arms (olanzapine 4.2% vs. haloperidol 10.6%; RR 0.4, 95% CI 0.08 to 1.96; p 0.238).
Conclusions: Intramuscular olanzapine performed better than IM haloperidol in the management of acute agitation in ED. However, the differences were not statistically significant.
Choudhary Y, Jamshed N, […] Khan MA. Olanzapine vs Haloperidol for Management of Acute Agitation in Emergency Department: An Open Label Randomized Controlled Trial.
Journal of Emergency Medicine. 2024 Nov.

What was the study design?

This was an open-label randomised controlled trial comparing intramuscular olanzapine to haloperidol in ABD. Patients were recruited at a tertiary centre in north India between 2021 and 2022.

Can you tell me about the patients?

The researchers aimed to recruit adults (>18 years) presenting to the ED with acute agitation. This was defined as agitation that scored >3 on the Altered Mental Status (AMS) scoring system. I had not heard of the AMS prior to reading this study, but after reading through the scoring (reproduced below) this seems like a sensible threshold.

Score	Responsiveness	Speech	Facial Expression	Eyes
+4	Combative, very violent, or out of control	Loud outbursts	Agitated	Normal
+3	Very anxious, agitated, mild physical element of violence	Loud outbursts	Agitated	Normal
+2	Anxious, agitated	Loud outbursts	Normal	Normal
+1	Anxious, Restless	Normal	Normal	Normal
0	Responds readily to name in normal tone	Normal	Normal	Clear, no ptosis
-1	Lethargic response to name	Mild slowing or thickening	Mild relaxation	Glazed or mild ptosis (<half eye)
-2	Responds only if name called loudly	Slurring or prominent slowing	Marked relaxation (slacked jaw)	Glazed or marked ptosis (>half eye)
-3	Responds only after mild prodding	Few recognizable words	Marked relaxation (slacked jaw)	Glazed or marked ptosis (>half eye)
-4	Does not respond to mild prodding or shaking	Few recognizable words	Marked relaxation (slacked jaw)	Glazed or marked ptosis (>half eye)

There were many exclusion criteria for the trial, including Parkinsonism, QT prolongation, alcohol withdrawal, and known intolerance of study drugs. Patients were also excluded if they were over sixty-five or had known cardiovascular, cerebrovascular, or respiratory disease. QT prolongation is always a tricky one, as most, if not all the ABD patients I’ve seen are not in a fit state to have an ECG taken.

In total, 256 patients were screened and 94 recruited into the trial. Consent was provided on the patient’s behalf by a legally authorised representative. The majority of patients were young (median age 33) and there was a slight (60%) male preponderance. The AMS was 3 in most cases (83%) and the two most common (79%) causes of agitation were medical or psychiatric disorders. Only a minority (15.9%) were intoxicated or withdrawing.

What was given in each trial arm?

All patients were given a single intramuscular injection of an antipsychotic: the intervention group was given 10mg olanzapine, and the control group was given 5mg haloperidol.

After thirty minutes, patients could receive benzodiazepines if the treating clinician believed that they required additional sedation.

What outcome measures were used?

The primary outcome was adequate sedation at fifteen minutes post-administration. This was defined as an AMS score of zero or less.

Secondary outcomes included use of rescue medications and AMS score at thirty minutes post-administration. The researchers also recorded the incidence of pre-specified adverse events, including hypotension, QT prolongation, arrythmias, seizures, and dystonic symptoms.

What were the main results?

After fifteen minutes, adequate levels of sedation were recorded in 31.9% of patients given olanzapine and 25.5% of patients given haloperidol. Chi-squared testing was used to determine if this difference was statistically significant, and it was not: relative risk was 1.25 but the 95% confidence interval was 0.65 to 2.37.

Benzodiazepines were administered as “rescue” sedation twice as often in the haloperidol group: 25.5% compared to 12.7%. But this finding was also statistically non-significant.

Two patients in the haloperidol group developed QT prolongation, and three required intubation in ED.

What should we take away from this study?

The authors of this study should be commended for conducting high-quality research in an area of emergency medicine that is sorely lacking in randomised trials. It is also helpful that they used inexpensive, commonly-available medications to do this. I feel that the results, although statistically non-significant, would probably have been sufficient to support the non-inferiority of olanzapine.

I also feel that the investigators set their sights a little too high by setting AMS <0 as their threshold for adequate sedation. An AMS score of two, for example, describes a patient who is ‘anxious [and] agitated’ with occasional ‘loud outbursts’ but no violence or other combative behaviour. I would be very happy with this outcome in ABD! If the authors had found a significant difference between groups using AMS<2 as their primary outcome, I would count that as a clinically relevant result.

The key limitation to this trial, in my view, is the open-label design. Staff were aware of treatment allocation, and this may have biased several decisions crucial to the study – such as the prescription of “rescue” sedation, or the AMS scoring. I do not see why this could not have been a blinded study. Both olanzapine and haloperidol can be drawn up as 2ml solutions, and although the typical medicinal forms are different – powder and liquid, respectively – the injection could have been prepared by an unblinded member of the research team.

Another drawback worth discussing is the eligibility criteria. A wide range of co-morbidities were excluded, as were individuals over sixty-five. It is unsurprising to me that the researchers found a low rate of adverse drug reactions, as they screened out the patients most likely to have them. This is a shame. If the study had shown that olanzapine was equivalent to haloperidol at reducing agitation, but caused significantly fewer side-effects, this would be a very important finding.

Should this study change our practice?

Probably. This trial has encouraged me to use olanzapine in the situations where I might want to give an antipsychotic for acute agitation. It is widely used by our critical care colleagues here in Virchester, and we keep it in on-hand for use in serotonin syndrome.

I would love to see more resources devoted to the wider management of ABD in emergency care. I have seen trials of verbal de-escalation techniques in psychiatric and care home settings but nothing in ED. It strikes me that the success of individual sedative agents is likely to be highly dependent on non-pharmacological factors like these.

Greg Yates