Ketamine – does it maintain protective airway reflexes?

photo (9)

Recently a twitter discussion on Rapid sequence airway and ketamine sparked an email conversation between myself, Dr Nicholas Chrimes and Dr Reuben Strayer. With their permission and my thanks to them, I will detail what we discussed by reproduction of the email content.


Nic I don’t totally understand your concern here. Ketamine is used constantly in dissociative doses on non-fasted (not that that matters) patients for procedures of widely variable lengths. There are no reports of clinically significant aspiration that I am aware of. I’m sure it has happened, but it is certainly exceedingly rare, and I’m confused as to why you would be concerned about it during DSI. During a 4-hour operation without an anesthesiologist, that would be a concern, but they do that all the time with ketamine monotherapy in resource-poor environments. But for 10 minutes of DSI?

I have a hard time following twitter conversations so maybe I’m missing something?
Hi Reuben
In brief here’s my issue with this…
…in which he states that ketamine will not blunt airway reflexes citing this article…
…which states that protective airway reflexes are not depressed (although it then goes on to say that cases of aspiration have occurred which is a bit contradictory) citing this article…
…which is a study of pharyngeal muscle tone with ketamine sedation and makes no reference at all to airway protective reflexes.
So that chain of references is a bit of a dead end. Ultimately there is nothing to support the contention that airway reflexes are preserved with ketamine there.
Anaesthetic teaching re ketamine is that while it is “less likely” to suppress airway reflexes you shouldn’t rely on it. In fact you could have the worst of both worlds, where you induce “abnormal” airway reflexes which allow aspiration but promote laryngospasm – this could put you in a position where you have the diabolical combination of impaired gas exchange and interrupted oxygen delivery.
As far as numerous cases without incident – again I would argue that most drink drivers don’t have car accidents (and that many who don’t drink drive do). Yet clearly the risk is increased with drink driving. I also can’t recall a situation in my life where wearing a seatbelt has been of any benefit to me personally – but that doesn’t mean I don’t always wear one. To take yet another example, CICO is a once to none event in the career of most anaesthetists – but that doesn’t stop us thinking it is a significant enough problem to learn how to manage it. We are talking about rare but potentially catastrophic events. We need to consider them as risks even if we rarely or never see them.
Don’t get me wrong. I think DSI is a useful technique in selected patients – as always you need to understand what risks you are trading off and weigh them up in the particular circumstances. It doesn’t matter what drug you give, if you give enough you will depress airway reflexes and put the patient at risk of aspiration. You always need to be giving the minimum possible to get the job done. Doses of 1-2mg/kg IV are induction doses and would significantly impair airway protection in many patients. That doesn’t mean they WILL aspirate, it means they COULD – and that you placed them at that risk. That may well be a valid thing to do but you have to justify the risk you are trading off. In the patient with a probable normal airway and a potentially full stomach it is hard to justify compromising airway protection for preoxygenation (a technique whose sole purpose is to protect against prolonged interruption to oxygenation due to difficult airway). Conversely in the patient with an pharyngeal abscess and very difficult airway in whom awake intubation was not feasible (as it often isn’t in these patients) I might keep them spontaneously ventilating and induce full anaesthesia (ablating airway reflexes completely and putting them at risk of aspiration) despite a potentially full stomach. It just depends on the balance of risks.
This doesn’t mean there isn’t a reasonable “happy medium” between enough sedation to allow preO2 whilst maintaining preservation of airway reflexes. I just don’t think that’s the 1.5mg/kg that Scott Weingart mentions.  I would have thought 0.1-0.5mg/kg IV ketamine would be a reasonable starting point depending on the patient (I’ve seen frail patients become unconscious with 0.1mg/kg IV).  You can put more in but you can’t take it out!
Happy to discuss further (will be in NY next week too!)
My main concern is that most of the FOAMed literature I’ve read simply seems to say that “ketamine won’t affect airway reflexes” – which means that clinicians aren’t even being given the opportunity to weigh up the risks because they don’t appreciate that they’re there.
like I suggested, perhaps we podcast this..wouldnt it be cool if Nic you travelled all the way to NYC, meet Reuben and both of you podcast with me on this!
Anyway regardless of whether that happens, my thoughts
lots of textbooks, guidelines and policies state the ketamine preservation of airway reflexes. Its on, clearly stating this. Unfortunately in my research since we started this topic discussion, I have been unable to find any good research data to back up this claim.
So essentially its consensus expert opinion.
Now here is my clinical experience.
I use ketamine at fairly decent dosages in unfasted patients..a lot.
Have been flying psychiatric patients and recording our data for 5years now, using ketamine sedation. We have had no aspiration, but one vomit case. No one has needed to go to ICU or be admitted for any complications from ketamine sedation, excluding their psychiatric illness of course!
The one vomit case was in a full stomach patient( he had eaten several pies and a carton of milk). He should have been intubated but my colleague decided to use ketamine infusion which controlled the agitation well but patient had a large vomit during the flight. However he was able to control the vomit enough that it all went into a container and he suffered no aspiration whatsoever.
The doses I give are at least 1mg/kg and yes we have given up to 2mg/kg IV bolus on loading the patient to gain control. This is essentially deep sedation to general anaesthesia at those doses, but transient and justifiable to gain control of a dangerous agitated state. Of course we are prepped ready to do full RSI at time.
I rarely now need to progress to intubate these patients, which was the norm when I started in aeromedical retrieval. I establish them on an infusion of ketamine at anything from 1-2mg/kg/hr. I did a job last week where it took 200mg/hr to get patient nicely sedated. 15 minutes after turning off infusion, patient was out of bed and walked from ED to psych ward.
Now there are some rules when doing this. Everyone I treat with this protocol gets an antiemetic and my choice is droperidol, for added sedative effect and antiemesis. Also often I give some small midazolam bolus like 2-4 mg, maybe once an hour during the infusion. We run these infusions on average from 1- 2 hrs for our flights.
I try to fast everyone as much as possible but its sometimes not possible. So I do the ketamine sedation with full RSI kit ready to go and wait 5 halflives for steady state..thats about 45 minutes. Almost 95% of time, its all ok by then , we apply restraints and capnography and transport. If it all goes to custard during that period, we proceed with full RSI.
So its my clinical impression that ketamine DOES preserve airway reflexes to a degree. Having said that like any general anaesthetic agent , the more you give, the more anaesthetised the patient will become. And thats where it can get very tricky. A colleague had a case of a combative patient with no IV access, IMI Ketamine 500mg given, patient sedated within 3 minutes, airway noises and gurgling..needed intubation. we have had 2 cases of mild laryngospasm with our IV ketamine infusion protocol ..none needed advanced airway interventions.
At Xmas time, I had to go get some trauma patients from a remote clinic. car accident. everyone was really drunk and combative. One guy was essentially comatose but still a bit agitated. He needed a tube but was too combative to prepare still. So gave him some ketamine and droperidol. which worked but he started going stridorous and needed jaw thrust to maintain airway patency. Ketamine dose was 40 mg IV.
Point is that intoxicated patients can be so anaesthetised already that any extra anaesthesia/sedation even with ketamine, may be enough to trouble the airway.
Nic you make a good point that clinicians do need to get adequate experience with ketamine before using it in a way like DSI  etc. It is a forgiving drug I admit that and of all the agents available I would choose it above others to use in the unprotected airway. Its preservation of respiration is remarkable even in very high dosages. It is quite plausible that given this brainstem sparing effect, that airway reflexes are maintained. But yes I can find no data on this for proof.
Only clinical observation
Podcast from NY would be very cool. Happy to participate if feasible (but insist you introduce us as “LIVE FROM NEW YORK!” so that I feel like I’m on David Letterman).

Agree with most of what you’ve said. I think we agree that its a matter of balancing risks in individual patients (as with the RSA). I think the main value of discussing this is to raise the awareness that airway protection with ketamine, whilst largely touted, has no good evidence base beyond anecdotal and if you overstep you’ll only know after the horse has bolted => need to make informed decisions about balance of risk rather than just assuming ketamine will spare airway reflexes. It’s one thing to be set up for RSI but if the patient regurg’s and aspirates solid matter there may not be much you can do. The horse has potentially bolted and the patient could end up dead. Many cardiac arrest patients don’t aspirate but I don’t think we’d argue that death spares airway protective reflexes (granted aspiration with ketamine seems to be less common than with cardiac arrest!).
Many anaesthetists would advise against giving a patient ANY sedative agent if unfasted, even 2mg of IV midazolam – though personally I think this is ridiculous. The illogical thing is that they’ll then put them on a morphine PCA and let them have 40mg of IV morphine over then next few hours (which in many patients will be fine – again everything needs to be assessed on an individual basis). Nothing drives me crazy like inconsistent reasoning (as is often present when people have concrete rules).

Minh :

yeah the whole fasting thing is as you know on a poor base of evidence and lots of anecdote!
the whole thing with avoiding ETI in the patient group I mentioned is that it makes for some very dodgey logistics of where to bring these patients and recover them safely once intubated. we have had nasty situations where ICU,ED and psych ward have refused to recover patients with suggestions of extubating them in the ED corridor!
Basically because the patient has a mental ( non critical care illness) health condition.
And we have had a few intubation disasters in this patient group.
when we started doing this ketamine sedation, I consulted an anaesthetist friend of mine who had worked in RFDS prior. He agreed with ketamine given the unique situations we are presented with and it is an incredibly safe drug for the purpose. I have one article published on this and another one underway.
At least one other Australia retrieval service is using our protocol now, and they are trialling even propofol infusions in the unfasted unprotected airway…I think in the right hands with right care, that is also ok.
I will send you some slides I have for an upcoming retrieval lecture I am giving to registrars

Reuben :

thanks to both of you for your thoughtful replies.

there’s a lot to talk about here, and a lot of it comes down to important conceptual questions, like, what does it mean to say that the patient is protecting her airway? what does the term general anesthesia mean? how can we predict who is at risk to aspirate?
nick, you note that there are no studies to support the claim that airway reflexes are preserved with ketamine. what would such a study look like?
aspiration risk is damn tricky. the studies that attempt to assess it using objective disease-oriented outcomes (like the presence of tracer dye in the trachea/lungs) do not seem to correlate with patient-oriented outcomes, and patient-oriented outcomes from aspiration are so rare as to be the exclusively the domain of case reports.
anesthesiologists are focused on aspiration risk, but this practice, which informs massive efforts to, for example, keep patients NPO, is based on shockingly thin evidence.
the term general anesthesia, and the entire sedation continuum for that matter, makes sense for sedating anesthetics  but does not really apply to ketamine. patients given ketamine can obstruct their airway, but if you take a dissociated patient and give her some water in her mouth, she will swallow it. ketamine produces anesthesia by an entirely different mechanism than essentially all other anesthetics, and has a different risk profile.
I certainly appreciate your ideas about individualized risk assessments and not being dogmatic. all patients who receive dissociative-dose ketamine should be monitored in full PSA setup – they can obstruct their away by malpositioning or laryngospasm, or occasionally copious secretions. but I cannot agree that aspiration risk is an important concern during a few minutes of ketamine dissociation for DSI. this is not like drunk drivers or wearing seatbelts, where there is ample evidence of cause and effect. clinically meaningful aspiration has never been reported with ketamine to my knowledge. given how long and how extensively ketamine has been used, that is pretty damn powerful evidence in my opinion.
certainly, there could be a case report tomorrow. but then we have one. in 20 million ketamine dissociations.
a much more credible concern around DSI, in my opinion, is the whether it’s safe to use NIV in a patient who has received PSA-level sedation, or dissociation. we know for sure that ketamine makes folks vomit. this generally occurs once mental status has returned, so aspiration is not a worry, but occasionally it happens during dissociation. this is usually fine as well because, umm, airway reflexes are preserved. but add to that scenario an NIV face mask, blowing air and now vomitus to the back of the oropharynx, and I suspect you can cause aspiration in the wide awake and alert.
I am in the middle of a rural emergency medicine stint, will return to NYC on 3/19. working a fair bit that week, but if you’re around, Nick, happy to meet up for a podcast or a discussion or whatever.
Really sorry I couldn’t make it to SMACC. everyone seems to have had an awesome time.

Minh :

thanks Reuben that sounds great if a podcast with you two is possible. Let me know how you both travel with the invitation.
I am aware of some cases of aspiration with ketamine reported in the literature
here is one for example
I think the key is in the dosing and administration, like any anaesthetic agent.
I agree with you on the lack of evidence for fasting..much robust evidence of lack of good protective effect..having said that it makes sense if elective procedure
Thanks Reuben. I think precisely because aspiration risk is “damn tricky” you have to assume that it is present with sedation. That doesn’t mean you can’t do DSI, it means you have to have a rationale to justify it if the patient were to aspirate beyond “I didn’t think that would happen” – because most anaesthetists will say they did think it could happen. I think there has to be a rationale for taking the risk of aspiration. This will obviously be based on what the perceived risk is – but it’s not zero, and it can kill.

Citing “no case reports” is also tricky. Firstly there doubtless are case reports and secondly case reports are the “tip of the iceberg”. Most things don’t get reported – unless they culminate in a death. I think the timing of fasting is based on “shockingly thin” evidence, I don’t think the concept of fasting is. There’s no doubt that fasting empties the stomach in most patients (it must at some point!) and that patients who suffer significant aspirations in an elective setting often turn out to have material in their stomach (despite fasting).
I think the fact that ketamine works by a different receptor mechanism to propofol is a tricky argument. Inhaled anaesthetics work by a different agent again and even amongst them there is a spectrum of airway reactivity. All of them impair airway reflexes but some also simultaneously promote laryngeal hyperreactivity – similar to thiopentone which suppresses airway reflexes much less effectively than propofol – yet both still can prevent airway protection. Some degree of airway reflexes does not imply airway protection.
Sounds like it will be a good discussion. I’ll email you next week and we’ll see if it can be done.
thanks Reuben that sounds great if a podcast with you two is possible. Let me know how you both travel with the invitation.
Nick when are you in NYC, exactly?
I am aware of some cases of aspiration with ketamine reported in the literature
here is one for example
well there you go. a lesser person would say, how does this report of a 6 year old with a brain tumor who had also received a barbiturate apply to our patients, but I wouldn’t say that, I would say, aspiration is possible during ketamine sedation, but that it is a distinctly rare event. I perhaps differ with Nick in that I see the benefit:harm balance in ketamine-based DSI, on an agitated patient who will not cooperate with pre-oxygenation, as overwhelmingly in favor of DSI.
I think the key is in the dosing and administration, like any anaesthetic agent.
not sure I can agree with this either. but we can discuss.
I agree with you on the lack of evidence for fasting..much robust evidence of lack of good protective effect..having said that it makes sense if elective procedure
maybe it makes sense. there is evidence that fasting might actually be harmful.
and there is no evidence that fasting prior to PSA benefits patients.
but we can discuss.
In the meantime I had a lot of time on the plane to write the following (bear in mind that I would have been mildly hypoxic at the time!)…
I think we’re muddling some very distinct issues here:
  • The process of fastING of patients and how long it takes to achieve an empty stomach (balanced against the potential accumulation of gastric acid and lowering of pH with excessive fasts).
  • Patients actually being fastED – by this I mean actually having an empty stomach (or being likely to have one) rather than just having not eaten for what I agree is a rather arbitrary fasting period.
I don’t think anyone is disputing the benefits of being fastED. Someone who has solid food in their stomach is at more risk of regurgitation/aspiration with airway managment than someone who does not. We know that patients who can’t protect their airway (be that from drugs or stroke) are at risk of aspiration if there is food in their pharynx. And we know from observation that patients with solid material/fluid in their stomach frequently regurgitate this up into their pharynx during the induction of anaesthesia, much more so than patients who are apparently fasted. So I assume we aren’t debating that having a full stomach doesn’t matter in terms of airway management. If you have a full stomach and you impair airway protetion, you have an increased risk of aspiration.
It is also reasonable to say that you are more likely to be able to prevent pharyngeal material going into the airway by placing a cuffed ETT than an LMA (irrespective of whether you make cricoid pressure part of that process. I don’t think anyone is disputing this either. They may be arguing LMA’s provide more protection than previously thought but I don’t think anyone’s arguing they’re as good as an ETT). There may not be RCT’s to show it but it makes physical sense that a pressurised sealed cuff in the trachea will protect the airway more than the tip of the LMA sitting in the top of the oesophagus. It would also make sense to say that the risk of aspiration is likely to be less as the period between the loss of airway reflexes and the insertion of the cuffed ETT becomes shorter, simply because there is a decreased “window of opportunity”. This is the whole principle of rapid sequence induction – making the time to insertion of a cuffed ETT as “rapid” as possible after airway reflexes are lost (cricoid is part of this to theoretically provide “cover” during this interval but we can ignore the controversy of cricoid at the moment as the “rapid” rationale applies equally even if you decide cricoid is of no benefit).
The remaining questions are then:
  • What does it take to make a patient fasted (ie. low risk of significant gastric content)?
  • When do airway protective reflexes become compromised.
The question of time to being fasted in elective surgical patients is really irrelevant to the RSA/DSI discussion. The issue in question there is not whether having an empty stomach is beneficial but whether the anaesthetic profession is fasting patients too long in order to achieve an empty stomach, and potentially decreasing pH of gastric content as a consequence. There is a lot of work being done about this with numerous similar studies to the one Reuben has included. I think that the majority of anaesthetists now wouldn’t  want to fast from clear fluids for more than 2 hours on that basis (although sometimes hospital protocols and practicalities subvert this). A two hour fast from clear fluids has become pretty standard practice for elective patients, at least amongst the anaesthetist I work with. The required time for food is still a bit unclear and there are many issues to be considered which affect the rate of gastric emptying: fat content of food, obesity, pregnancy, pain, opioids, sepsis, autonomic neuropathy, etc. The standard 6 hour fast from food may indeed be too cautious in many cases but it is designed to provide a margin of safety for many of these issues (there is evidence though that for high fat meals gastric emptying time may be much longer than 6hrs). In any case excessive fasting from food doesn’t seem to be a problem if clear fluids can be continued as described above. So the question of harm to patients resulting from fasting is a question of excessive fasting and unlikely to be relevant to the ED/retrieval patient group where the issue is not being fasted at all.
Also I think that most anaesthetists would consider that many emergency patients are never going to be fasted, irrespective of the time since last oral intake. Pain, opioids, stress, anxiety, intrabdominal sepsis, etc all delay gastric emptying. Irrespective of the duration of fasting, most anaesthetists would not consider these patients fasted. A patient with acute appendicitis or severe pain would always be treated as having a potentially full stomach no matter how long it has been since they ate/drank.
I’m not advocating fasting most emergency patients for airway management, not because having an empty stomach might not benefit them but because fasting is unlikely to to achieve gastric emptying in many of them. I don’t do that in theatre and I wouldn’t expect it to be done in an ER or retrieval setting. In fact this is the basis of my concern. It won’t make any difference to management as they will still be at high risk of having a full stomach. Emergency patients by their very nature often have fasting times that are unknown &/or factors which might delay gastric emptying despite prolonged fasting. As a result they usually need to be treated as having a potentially full stomach.
So to recap:
  • Patients with a full stomach are more likely to regurgitate/aspirate than those with empty stomachs if their airway reflexes become impaired.
  • Emergency patients often need to be considered to have a potentially full stomach.
  • An ETT is likely to offer better protection from this than an LMA
  • Minimising the time between loss of airway reflexes and insertion of the ETT is likely to decrease the risk of aspiration by shortening the period of vulnerability.
The decision to be made about emergency patients is not whether to fast them (as this is likely to be fruitless) but whether their airway reflexes are going to become impaired by the degree of sedation being provided (in which case RSI and insertion of a cuffed ETT is warranted). The final remaining issue therefore becomes “When do airway reflexes become impaired?”. This is the key issue as I agree it is flawed to necessarily apply fasting times for GA to light procedural sedation as airway reflexes are probably maintained (provided “light sedation” can be reliably provided – which it may not be in an emergency setting with compromised patients who can be unexpectedly sensitive to sedative agents). As Reuben points out, it’s difficult to know at what point airway reflexes become compromised. It is generally accepted that as conscious state decreases, so does the ability of a patient to protect their airway. Anaesthetists generally associate “purposeful responses” as correlating with the maintenance of normal airway protective reflexes (I’m not sure whether there’s any scientific basis to this though) whilst ED often uses the magic GCS 8 as the trigger to intubate a patient with deteriorating conscious state for the purposes of airway protection (which provides a similar level of consciousness). Whether these exact triggers are valid, I don’t think anyone is disputing the general principle behind this – that as a patient becomes more unconscious they are less likely to protect their airway against aspiration.
Reuben has argued that because ketamine acts via a different mechanism (NMDA) to propofol or thiopentone (GABA) its effects on airway protection should cannot be compared with them. I’m not sure there’s any basis for this statement. Whether it’s by NMDA/GABA mediated drug mechanisms, inhalational anaesthetic agents (mechanism of action unclear), ischaemic stroke, intracerebral mass, hypoglycaemia, seizure, cardiac arrest, etc the same general principle is likely to apply: as conscious state decreases, airway protective mechanisms become more compromised. The degree of impairment at different conscious states may vary (and I accept it is widely accepted to be less for a given degree of sedation with ketamine than with propofol or thiopentone) but the trend is still there. To assert that ketamine defies this general principle and that even in states of deep unconsciousness (as would potentially be achieved with doses of 1-2mg/kg IV) normal airway protective reflexes are reliably still maintained (therefore making it different from all other known causes of unconsciousness) would require strong evidence or at the very least a convincing pharmacodynamic rationale. I don’t believe either of these exists. I think simply relying on the lack of case reports of aspiration with ketamine is a flawed rationale on at least 2 grounds:
  1. Lots of stuff that happens doesn’t get reported – unfortunately especially where people feel they caused the problem by perhaps over sedating an unfasten patient.
  2. Most ketamine use over the years hasn’t been at the doses of 1-2mg/kg IV that I’m concerned about, so wouldn’t be expected to necessarily place patients at risk of aspiration. This is probably true of most of the procedural sedation cases reviewed in your article. There’s no doubt that there’s a safe level of sedation where airway reflexes are likely to be preserved – I just think that it’s unlikely that induction doses of ketamine reliably provide it. And that’s really the issue, we’re talking about it like it’s procedural sedation but at those doses it’s really induction of GA. Most ED/retrieval physicians I know that use ketamine for DSI wouldn’t use these doses. Similarly anaesthetists I know who use ketamine to settle agitated patients pre-induction (not using the DSI nomenclature but it’s the same process) wouldn’t use these doses.
  3. Airway manipulation is taking place during DSI (in terms of application of PEEP) and RSA (placement of SGA) and the article Reuben has provided emphasises the increased risk of aspiration when airway manipulation takes place. Most of the cases in the review aren’t cases in which airway manipulation was undertaken and therefore were at lower risk of aspiration anyway.
To summarise:
The issue of how long to fast patients isn’t in dispute here. Most emergency patients aren’t ever going to be fasted even though they may have been fasting. The issue of fasting patients forlight procedural sedation is also not in question, I agree it’s probably unnecessary. Some confusion arises from a lack of clarity about what constitutes “sedation”. The term is misused (by anaesthetists  as much as anyone else) with procedures such as gastroscopy or ECT being described as being performed “under sedation” when in fact the lack of responsiveness, awareness and airway reflexes induced by this level of CNS depression constitutes general anaesthesia. The fact that no anaesthetic machine, airway device or depth of anaesthesia monitoring is in place doesn’t change the fact that this constitutes general anaesthesia – but does lead to a decreased level of situational awareness of what is being undertaken.
The question with DSI is whether ketamine at the doses stated impairs airway protection and places them at increased risk of aspiration and how best to protect them from that. I think it is true that ketamine preserves airway reflexes for longer than other agents and thus if sedation is necessary it is a sensible choice as it provides a greater margin of safety. At doses of 1-2mg/kg IV, however, I think we are well beyond that margin of safety. The patient is now receiving general anaesthesia rather than light “procedural sedation” and is likely to have significantly impaired airway reflexes and be at risk of aspiration. Thus if you give these sorts of doses you need to get a cuffed ETT into the trachea ASAP. DSI with these doses does not provide this. In fact the “delayed” nomenclature is quite appropriate as the opposite is achieved. In contrast to the “rapid” sequence induction it “delays” the time between the loss of airway reflexes and the protection of the airway with a cuffed ETT, which is likely to increase the time available for, and therefore the risk of, aspiration (which can be associated with signifcant morbidity & mortality). As Reuben points out the concommitant application of PEEP in these patients, which is often part of DSI, compounds this issue.
I’m not saying they’re aren’t cases in which favouring the need to sedate over the risk of impairing airway reflexes is warranted due to the risks of a patient being agitated/not preoxygenated, esp. where there is the potential for a difficult airway &/or the SaO2 is significantly imparied at the outset. I’m saying that these risks need to be made explicit so that clinicians aren’t using DSI where the balance of risks doesn’t support it, simply because they believe that “ketamine won’t impair airway protection at any dose” – because that simply isn’t true and believing it will potentially place patients at risk. My concern is that this appears to be a common belief at the moment and that that is leading to a gradual upward “creep” in the doses of ketamine being used.
I think that at incrementally administered doses of  0.1-0.5mg/kg IV the principle of DSI is usually likely to be both safe & useful in well patients. In patients who are severely compromised or “bouncing off the walls” the dose may have to be adjusted down or up respectively – but regardless of dosing the fundamental principle remains: if you make them unconscious, you risk compromising airway protective reflexes and they need an ETT ASAP. In reality achieving the goals of DSI shouldn’t require this degree of sedation anyway, so with judicious, incremental titration of doses there shouldn’t be an issue.
As an aside, with respect to procedural sedation, I was discussing with Minh the use of alfentanil for this purpose. This has become my standard drug for “procedural analgesia” where a painful stimulus is being applied. Most anaesthetists seem to use propofol for this purpose (on fasted patients). Propofol has no analgesic properties and thus to be useful must be used at doses which significantly impair consciousness (ie. general anaesthesia). Ketamine is preferred in ED to maintain an adequate level of consciousness for airway protection in potentially unfasted patients but has the downside of a relatively prolonged recovery. Alfentanil offers the best of both worlds: as a potent, short acting analgesic it provides intense procedural analgesia with levels of consciousness likely to be associated with maintained airway protective reflexes and its duration of effect is <10mins with patients being completely clear headed beyond this.
I think we can have a really valuable discussion around all of this.
thanks Nic
were you travelling business or economy class? 😉
Its interesting what you are trying to convey/argue. To me it highlights some of the nuances between ED ,prehospital and OT care of sedation/anaesthesia
BUt at the end of the day, you are right. an airway is an airway and drugs we give can affect to varying degrees the ability to protect it.
Everything is a poison potentially!
the fact is in ED settings, we resort to alternative strategies that are goal driven rather than process mandated.
for example would you ever use thiopentone to sedate an agitated patient in ED, without ETI?
I regard anaesthetists such as yourself to be akin to Master Chefs who have the skill and experience to experiment with various ingredients and make stunning results
ED, prehospital folks like Reuben and I, I regard as more akin to industrial chefs who aim at safe, reliable, reproducible creations. Occasionally we experiment at novel creations but only for certain indications and occasions
At the end of day, we make decisions that involve some risk, partic in airway mgt. The point of disagreement we have is that you offer traditional techniques as being well founded/proven in minimising risk, RSI for example. You suggest that any deviation from the principle of RSI is unsafe. You know that there is no solid evidence that RSI prevents aspiration at all.
Dosing of ketamine is once again a fairly variable issue and so whilst you argue the risk at that dosing is too high, this is in fact a matter of opinion with virtually no evidence to support the claim!
The difficult airway in hypoxic combative patient , the agitated psychiatric patient needing air transport, these are some challenging situations that are not always if ever addressed entirely by traditional anaesthesia techniques like RSI
In these cases it is the akin to the metaphor of Buridan’s ass. You cant sit on your hands trying to decide whats best whilst the patient deteriorates. There are no zero risk optionsDSI, ketamine. they have a role in such cases, not always but can be helpful.
Proof? well give us time 😉
If I can convince my local psychiatrists and other retrieval services to use ketamine in psychotic patients needing air transport to avoid intubation when safe to do so..we might even be able to teach DSI to the anaesthetic community 😉
I look forward to your response piece written on the return flight home!
Nick that was a very clear, well-written exposition of the issues.

I agree with a lot of it. Most of it, even. There are a couple of areas of disagreement.
I think you are mistaken about the usual dose of ketamine for PSA. It is a full dissociative dose. 1-2 mg/kg. Same is the induction dose. Clinicians of all stripes, but in particular emergency physicians, have been performing PSA (or what was previously known wrongly and still often referred to wrongly as conscious sedation) using this dose, 1-2 mg/kg, for many years, and there is tons and tons of evidence to support its safety. At full dissociative dose. Only recently has low dose or subdissociative dose ketamine become fashionable. It is clear beyond any reasonable doubt that unfasted patients can be fully dissociated and continue to maintain their vital functions with little or no assistance. It is because of this fact that I was confused about why you could be concerned about airway embarrassment for the few minutes a patient is dissociated during DSI.
Dissociated patients can develop airway embarrassment. It doesn’t happen often, but it happens, by either head/neck malpositioning, excessive secretions, or laryngospasm. This is why all dissociated patients require full PSA setup and airway-capable physician at bedside. Certainly a dissociated patient who regurgitates is at higher risk to aspirate than an awake/alert patient. But we know from PSA literature that this risk is so small as to be nearly meaningless when compared to the risk of paralyzing an incompletely preoxygenated patient–I’m sure both of you have seen patients die from that–I certainly have.
The idea that an emergency or pre-hospital clinician will “incrementally administer doses of  0.1-0.5mg/kg” on the type of patient we’re discussing–dying patients–is unrealistic, probably unsafe, and in my opinion unwarranted. I’m not trying to sound flippant, but this really highlights the difference between anesthesiologist practice (focused on well, cooperative patients with understood physiology) and emergency practice (focused on unknown, violently agitated patients who are either merely psychotic or drunk, or minutes away from arresting, we have no idea which).
I would love to try alfentanil for painful procedures in cooperative patients. I know some folks are using remi. The main reason I’m interested in this approach is that we might be able to use it without activating PSA protocol, which is a million forms, nurses, etc. etc.
There are other points you raise that we could discuss that are off the topic of DSI. Happy to discuss any of this further. If you want to get together for a podcast:



  1. Ketamine maintains better airway protective reflexes than most standard sedatives and anaesthetic agents
  2. Its dose related = higher dose, less protective reflexes
  3. You still need to be fully prepared to take over a patients airway, even using ketamine.

3 thoughts on “Ketamine – does it maintain protective airway reflexes?

  1. minh: regarding your conclusion #2. to the (relatively small, in my opinion) extent that airway reflexes are abridged by ketamine, they are progressively abridged with dose *until full dissociation occurs.* Once a patient is fully dissociated, more ketamine does not _further_ dissociate or put a patient “deeper,” the only effect is longer duration of action. I mention this because providers who want their patient fully dissociated should not shy away from large doses; unless the procedure is very brief, I start with 2 mg/kg IV and go up from there, anything less doesn’t dissociate everyone. You can often get away with subdissociative doses for PSA, especially because ketamine also offers profound analgesia, but psychiatric distress is much more common using this strategy, because cognition and perception have been turned upside down, but external stimuli have not been entirely blocked, so the patient is in a state that has the potential to be particularly terrifying, much like what is often described as emergence, when psychiatric distress occurs after full dissociation, as external stimuli are re-integrated. for RSI, I typically give 400 mg (2 vials of the most common formulation here in the US) because it buys me more time to get my analgesic and sedative drips up.

    1. Thanks Reuben! You gotta come to #smacc2014! We can have a whole plenary session on ketamine!

      I must say my clinical experience is in accordance with yours but I shy away from 2mg/kg doses and 400 mg RSI doses..basically because I seem to manage with lower doses. It is true though that the most common novice error when using ketamine is to give too little and yes this does potentially worsen agitation and have a dysphoric patient who is neither well dissociated nor cooperative and alert!

      Its pretty clear though in the literature and my experience that the larger the ketamine dose, the more adverse events you encounter as well as the longer duration of action. I have seen this mainly with IMI ketamine but sometimes with IV route. 5mg/kg IMI can induce vomiting and laryngospasm..this is well reported. Vomiting is rarely encountered at lower dosing range and laryngospasm is rare regardless but it seems to not occur at lower doses as much as higher ones ( personal observation only)

      Therefore my conclusion that higher dosing requires caution, but as you say lower dosing equally requires careful judgement! So finding the K hole for the desired clinical effect, is indeed the art of anaesthesia.;-)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: