JC: Should we premedicate for ketamine sedation? St Emlyn’s

Here in Virchester we believe that we were early pioneers of conscious sedation in the UK. It’s something we adopted as a routine procedure back in the 90s, in both adults and children, and so over the years we have tried to refine and adapt our techniques as the evidence changed. If you want to read more about our philosophical approach to sedation in the emergency department then you can read this blog and listen to the podcast on the concept of ‘balanced sedation’.

One question that often arises is whether to give patients pre-treatment with midazolam in order to smooth emergence and decrease side effects. In children, and via the BestBets methodology we developed here our view is that it probably does not work, but what about adults? I will admit that I did go through a phase of using small doses of midazolam (1-2mg) before the procedure, in particular when the patient appeared anxious. However, in recent years I’ve used it less and less. I’ve also had the experience of dealing with some really disturbing emergence phenomena. I remember deciding to use ketamine for the sedation of an elderly patient with a significant limb injury. All went well, but on recovery she became psychotic, believing that she was dead and that we were all dying. It was rather chaotic and distressing for all involved, so if there is something we might do to avoid this then all the better.

This month there is a randomised controlled trial in the Annals of Emergency Medicine that addresses this question, not only looking at the use of midazolam as a pre-treatment, but also in the use of haloperidol in patients undergoing ED ketamine sedation. The abstract is below, but as we always say, please read the full paper and make up your own mind about the strength of the evidence.

What kind of paper is this?

It’s an RCT which is entirely appropriate. We are looking at an intervention and so an RCT is the best design.

What did they do?

Patients over the age of 18 undergoing procedural sedation were recruited. There is a long list of entirely reasonable exclusions for the study, although interestingly there are a lot of patients in the list who I still use ketamine with – e.g. hx of alcoholism). This is a single centre study in a relatively small US emergency department. Patients were randomised using block randomisation, and then I think an envelope sequence of packs to dictate the next allocation with prepacked syringes of either haloperidol, placebo, or midazolam in blinded syringes, together with an unblinded syringe of ketamine.

The premeds were given five minutes before the ketamine which is a reasonable length of time to ensure that they would have had an effect. The ketamine dose was standardised at 1mg/kg which is a fairly robust dose for sedation in my opinion (my practice is to titrate to effect), but it’s in keeping wth standard recommendations. However, I wonder if some patients may have received more ketamine than they need as midazolam or concurrent analgesia may reduce the required ketamine dose.

What were the outcome measures?

So the key points here were the use of the Richmond Agitation Score at 5, 10 and 50 mins post ketamine, and then the Pittsburgh Agitation Score (PAS) during recovery which seems to be used in a variety of settings notably dementia(?).

And the main results?

They randomised 185 patients in total, with pretty good follow-up as you would expect in a trial that takes place entirely within the ED. 185 is a pretty small number for a three-armed trial. The power calculation was based on a five-fold decrease in the incidence of agitation at a score of 3 on the PAS. A score of 3 is defined in the paper as disruptive behaviour though if you look at the scoring system it’s arguably a fairly low level (as the PAS goes from 0-16) There were no major differences between in the groups in statistical terms, but with small numbers it’s difficult to tell.

In terms of sedation as measured on the RASS score the results suggest that patients with placebo were less sedated and sedated for less time. Indeed recovery time was significantly longer in the haloperidol and midazolam groups. Recovery time for placebo was 18 mins on average compared to 35 mins for midazolam and 50 mins for haloperidol (Ed – wow!). That’s a big difference in recovery times in my opinion. An important point if you are thinking about doing this.

In terms of agitation then the authors state that there is less agitation in the midazolam and haloperidol groups, and that looks to be the case. Agitation scores were only really seen in the placebo group with a median of 3 and an IQR of 0-5 (so still fairly low). Interestingly the overall level of agitation as defined in the study was just under 64% in the placebo group vs 25% for midazolam and 20% for haloperidol (defined as a PAS>2). This outcome is the one used for their power calculation. Clinician satisfaction was the same with all agents.

I do find it interesting to reflect on how we commonly report studies of this type where analysis is often targeted at differences in average scores. Arguably this is less important than looking at the number of patients who have very severe reactions. It’s a similar argument in studies about average pain scores, where I might be more interested in reducing the number of patients with severe pain as opposed to very small (arguably inconsequential) differences in average scores. In this study that finding is there in the adverse events data where there was a 10% risk of severe agitation (PAS>8) in the placebo group, but no severe agitation in the others. That’s an interesting finding, but the number is small and does not reflect my experience in using ketamine in the ED.

Any limitations?

The authors do a fair job in highlighting the limitations of this study. It’s a small single centre study which is always a bit of an issue with generalisability. The RCT design is good, but I wonder how effective the blinding will have been with the very different emergence times recorded (after a while I think we could have worked out who had placebo as they would have recovered far faster). The reported changes in agitation are interesting, but seem rather higher than in my practice which perhaps reflects the threshold of what was considered a significant change on the PAS (a score of 3 being regarded as important). I am interested in the apparent higher number of severe agitations in the placebo group, but the data here is not conclusive.

We would also look at the 60 min limit on data collection. Although this seems a long time it is pretty close to the average sedation time for haloperidol and so it might be possible that some patient outcomes may have occurred after data recording stopped.

What is really missing is the patient perspective. Does it matter if the patient is a bit agitated on emergence but that they have no memory of it? We can’t decide in this study, but my experience is that ketamine is a fabulous amnesic agent and so we may be pursuing something that does not matter from a ‘patient orientated outcome’ perspective. The PAS is a measure of expressed emotion/distress which may not be the same as the patients’ experienced emotion/distress. This is a subtle but important distinction that should be addressed in future trials

So what’s the bottom line?

This is a good paper, it’s a good design and the patients are seemingly similar to those in my practice, but is it enough for us all to adopt midaz or haloperidol? I’m not so sure at the moment. It’s important that we weigh up the potential for less agitation against the much longer recovery times.

However, there may well be a group of patients whom I might feel are at risk of emergence phenomena where I might think that trade off is worth it. Paradoxically though, those might well encompass many of the patients who were excluded from inclusion in this trial. Such are the difficulties and vagaries of trying to practise EBM.

So well done to @pooya_mehr and co-authors for moving the evidence on. I agree with them that larger and multicentre studies would help clarify those questions that remain.




How you can support St Emlyn’s


  1. Balanced Sedation in the ED. St.Emlyn’s http://www.stemlynsblog.org/balanced-sedation-in-the-ed-st-emlyns/
  2. Midazolam use in children undergoing ketamine sedation to reduce emergence reaction https://bestbets.org/bets/bet.php?id=3044
  3. Richmond Agitation Score https://www.mdcalc.com/richmond-agitation-sedation-scale-rass
  4. Pittsburgh Agitation Score https://www.researchgate.net/publication/230557061_The_Pittsburgh_Agitation_Scale_A_User-Friendly_Instrument_for_Rating_Agitation_in_Dementia_Patients
  5. Premedication With Midazolam or Haloperidol to Prevent Recovery Agitation in Adults Undergoing Procedural Sedation With Ketamine: A Randomized Double-Blind Clinical Trial.Akhlaghi N1Payandemehr P2Yaseri M3Akhlaghi AA4Abdolrazaghnejad A5. https://www.ncbi.nlm.nih.gov/pubmed/30611640
  6. JC: Intranasal Ketamine vs. Fentanyl for kids. St. Emlyn’s https://www.stemlynsblog.org/jc-intranasal-ketamine-vs-fentanyl-for-kids-st-emlyns/
  7. JC: K is Good For You – Subdissociative Ketamine vs Morphine in the ED https://www.stemlynsblog.org/jc-ketamine/
  8. JC: Is Ketofol worth the hassle? St.Emlyn’s https://www.stemlynsblog.org/jc-is-ketofol-worth-the-hassle-st-emlyns/
  9. Is Ketofol the milk of human kindness for procedural sedation. St.Emlyn’s https://www.stemlynsblog.org/ketofol-milk-human-kindness-procedural-sedation-st-emlyns/

Posted by Simon Carley

Professor Simon Carley MB ChB, PGDip, DipIMC (RCS Ed), FRCS (Ed)(1998), FHEA, FAcadMed, FRCEM, MPhil, MD, PhD is Creator, Webmaster, owner and Editor in Chief of the St Emlyn’s blog and podcast. He is Professor of Emergency Medicine at Manchester Metropolitan University and a Consultant in adult and paediatric Emergency Medicine at Manchester Foundation Trust. He is co-founder of BestBets, St.Emlyns and the MSc in emergency medicine at Manchester Metropolitan University. He is an Education Associate with the General Medical Council and is an Associate Editor for the Emergency Medicine Journal. His research interests include diagnostics, MedEd, Major incidents & Evidence based Emergency Medicine. He is verified on twitter as @EMManchester

  1. Thank your for your informative analysis of this interesting study. Especially in the prehospital environment I’ve had good experiences with (time and hemodynamics permitting) a combination of fentanyl followed by ketamine. Is anyone else doing this?


  2. Hi Simon, I think this topic needs to be framed in the context of patient-centered care (performing the intervention with the patient’s best interest in mind) and of course, patient safety. Ketamine was marooned in US anesthetic practice many years ago, and only now are we carefully returning to this drug for sedation purposes. I am not here to debate the above article, and this is more of a discussion of actual practice philosophy than evidence-based medicine.

    The old adage one of my senior colleagues in his early 70’s shared with me was “…the key to ketamine…is valium.” For patient-centered care issues, the sedation regimine with ketamine must include a drug to ‘calm the mind’ before its administration. You cannot lay a serious mind-expanding, hallucinatory trip on most adults without a severe amount of dissatisfaction and weird, disruptive behavior, especially if they are coming to you in a distressed, injured state. The trip will not lead to enlightenment, it will lead down the funnel of despair, and push every single psychologic button in those people that is related to their preexisting negative self-images.

    So thank goodness for benzodiazepines. I have been finding that in a ‘balanced’ sedation regimine with ketamine, less…is more. The 1 mg/kg dose is too high for my needs and uses, but then again, I am not performing the painful procedures that you are tasked with performing. My recent experience performing a ketamine supported tracheal intubation of a difficult airway (this dose not constitute medical advice) utilized midazolam 3 mg, titrated to minimal to mild sedation, fentanyl 25 mcg, glycopyrolate 0.2 mg, and aliquots of ketamine in 5 mg increments until onset of desired depth of sedation for the procedure. I began to reach that depth at the 15 mg mark, and utilized 25 mg total during the entire procedure (0.5 mg/kg).
    I encourage an ongoing discussion on this topic, and I thank you for the thoughtful FOAMed post.


    1. Hi James, You may well be right and that’s my sadness with this paper in that there we don’t really know what the experience is for the patient. I take your point about calming the mind, and that’s certianly in keeping with our philosophy of balanced sedation. Wouldn’t it have been great if the study had looked at patient scored outcomes too? That may have really made this definitive.



  3. […] St Emlyns; JC: Should we premedicate for ketamine sedation? […]


Thanks so much for following. Viva la #FOAMed

Translate »