Latest / CRITKNOW / Ketamine vs. Etomidate: the RSI trial
Transcript
- Ramesh Venkataraman: Hello everyone, welcome back to our podcast. This is the second podcast of CRIT-NO. I am Ramesh Venkatraman, a senior consultant of critical care medicine at Apollo Hospital, Chennai and my co-host Dr. Bharath Kumar here is a consultant critical care medicine at Apollo Hospital. First of all, I want to say that our CRIT-NO now is available both on YouTube and Spotify. Bharath, I was quite excited to see that almost about 85 people have subscribed. to this podcast and we have had at least 500 views in the last two weeks, which is quite encouraging for beginners like us. These are small steps and we hope we can keep polishing and keep going forward. I also want to take a moment to thank a lot of the students, friends and colleagues who have taken the time to listen to our first podcast and actually called us back or took time to text us with very constructive feedback. And we'll try to incorporate those as we go and hopefully keep getting better with what we do. So thank you everyone. ⁓ so going for this week Bharath, thought, you know, there a lot of good trials out there. I was just thinking something that would again help me impact my day-to-day clinical practice. So the paper I thought would be worth discussing is the RSI trial that compared it on the date versus ketamine as an induction agent for ⁓ emergent intubations in the emergency room and intensive care units. This paper was published in NEJM in December, 2025. It's about three months old, but I still thought it had enough impact that we should be discussing it in a little bit more detail and our audience would benefit from that. So with that prelude, ⁓ why don't you Bharath just give us a brief overview of the context of why the study was done and then before we delve into the paper itself. Bharath Kumar T V: Yeah, absolutely. Thank you. Thank you so much, Dr. Ramesh. And of course, I echo all of the gratitude and all of the sentiments that you shared. ⁓ So, airway management and specifically intubation is something that, you know, as ICU doctors, we deal on an everyday basis. And I think in that sort of context, this trial is super relevant and super important for us. And as critical care physicians, when we're thinking about integrating somebody in the ICU, think anatomical difficulty is something that occurs to all of us. And depending on the paper that you read, anatomical difficulty can range from anywhere between 10 to 25 % in the context of critically ill patients. But what is more important is the fact that nearly everybody that we intubate is physically physiologically challenging or physiologically difficult and for a variety of reasons. The most important of course is the fact that patients have a low physiological reserve, but there are additional factors at play such as suboptimal environment, the variation in ⁓ the operator skills and experience and even in well-resourced units, these things can vary from day to night and shift to shift. So there are lot of reasons for the physiological ⁓ challenges that we encounter. In the Intube study, which was published in 2021 in Jhama, which specifically looked at peri-intubation adverse events, the rate of adverse events was over 40 % if you take all adverse events put together. And cardiovascular instability was the top adverse event, again, occurring in over 40 % of patients. And when they say cardiovascular instability, they were referring to either hypertension or cardiac arrest or increased need for vasopressors. This study was conducted in, I think, at least 28 or 29 countries, over 200 centers. So a big effort at understanding peri-intubation adverse events. And obviously when you're thinking about cardiovascular collapse, one of the key drivers is the choice of serrative or anesthetic that we use. And in Intube, Propofol was the most commonly used agent followed by Midazolam and Itoamidate and ketamine were less commonly used. And the study also showed that Propofol was an independent risk factor for poor outcomes. One other important point that the study brought out is if you have a peri-intubation adverse event, it actually increases your chance of having a poor outcome as well. So I think in this sort of a background, this RSI trial comparing two strategies or two sedative choices for intubation, I think is very, important. Ramesh Venkataraman: I think this is fantastic wrap up just to add on to you. like almost 40 % of the patients have hemodynamic instability. And like you said, it almost doubles the mortality if you do have peri-intubation hemodynamic instability. So it is not a trivial event. And when we talk about physiologically difficult airway, I want to acknowledge two leaders from within our country who have been talking about it for almost a decade now, both from colleagues, respected colleagues and ex-presidents of IACCM1, Dr. G.K. Diwatia, and then Dr. Sheila Nayanan, who's actually a co-author of the InTube study. And they have done a lot of work over the last 10 years to get us forward in this field. So I think it's now due that we acknowledge them in this podcast for their contribution. So when I started my training Bharat, I think the answer was we use it use it to automate it before the in-tube study. So clearly in-tube study showed that propofol is bad. It showed midazolam causes variable amounts of hemodynamic instability. So by exclusion and the wisdom over a period of time has always been intermediate and ketamine are probably the two. relatively safe agents to induce somebody when you do an emergent intubation. When I trained, we've always used it to amputate back in the US, but then there was emerging literature that it inhibits cortisol synthesis, right? Even a single dose by blocking 11 beta hydroxylase decreases cortisol synthesis and this suppression lasts for as long as 72 hours. So that was a big fear. Now, what am I doing to those patients by suppressing cortisol? Even though there has not been a direct study associating this adrenal suppression to increase mortality with uterine bidet that has always been a concern and people have slowly moved towards ketamine as a more hemodynamic friendly way of intubating people. As most of our listeners would know ketamine has two effects here it's worth just quickly going into it for the sake of the trainees who are listening to it. One it is a directly it causes a negative inotropic effect. a vasodilatory effect. So, if you think of ketamine as a molecule it has got ⁓ cardiac depressing effect and vasodilatory effect, but it also releases catechols by various mechanisms and this catechol release causes an alpha and beta effect where it neutralizes some of the vasodilatation and negative inotropic effect and therefore, we get this relative hemodynamic stability. So, the traditional belief is ketamine because of all this seems to be better compared to the fear that we had with Etonidate. There have been two trials from what I could say worth mentioning here. One is the Ketacet trial that was done in 2009, which enrolled about 460 odd patients, which compared ketamine to Etonidate, showed that the mortality was worse with ketamine. The second study is EVK study done in 2021 of 800 patients. which showed decreased survival with the etomidate. So we have two major studies. One showed worsening mortality with ketamine, the other with the etomidate. And I think in that context of peri-intubation hypotension, ketamine versus etomidate and the pluses and minuses of both drugs and two studies that have shown conflicting results, I think the RSI study came in to basically say, look, can we get a definitive answer? Is there a good way to intubate? Which agent should I use? And that I think that's where the RSI study. comes into context for most of us. Bharath Kumar T V: Yeah, absolutely. No, it's very interesting that, you know, in your practice in the US, I think you've used a lot of itomidate. Whereas for me in India and Canada, where I sort of we use itomidate only in the context of cardiovascular OR. In Canada, I've never used itomidate. And then in India, of course, we predominantly use ketamine, at least in our unit. So it's interesting to see how there are differences, ⁓ I think. Yeah. Ramesh Venkataraman: Yeah. And I also think the difference is related to the time because when the itomidate cortisol literature started emerging around the time when I trained and I think that became a big factor later on and people are moving towards ketamine probably when you're trained it's possible that one there could be a geographical differences in how we practice and also maybe the time and the beliefs of itomidate being bad would probably hurt the use of itomidate. So with that prelude. Bharath Kumar T V: Mm-hmm. Yeah. Yeah, true, Ramesh Venkataraman: Do you want to just take this trial on and tell us what actually the authors did? Bharath Kumar T V: Yeah, sure. So in the RSI trial, the population included as adult critically ill patients undergoing intubation either in the emergency room or in intensive care unit. Ramesh Venkataraman: All Bharath Kumar T V: They enrolled patients from ⁓ six EDs and eight ICUs across six medical centers in the US, all in the US. The key exclusion that we noticed that they excluded pregnant patients and patients coming with trauma. ⁓ And what helped them to enroll, I think, is the specific regulatory ⁓ provisions that allow for exception from informed consent. Ramesh Venkataraman: Thank Bharath Kumar T V: because it would be very difficult to go through the conventional informed consent process here. So I think they had those regulatory sort of exceptions that are available. Ramesh Venkataraman: Just to kind of interject you, I think this is a good example of how, despite a waiver of consent, a study can be done as ethically and as neutrally as possible. People generally believe not having consent is something not ethical. And I think the listeners should understand that, in these situations, getting consent is difficult. And you can still do a multi-center randomized control study and get waiver of consent if the rest of the methodology is tight. I think it's commendable the authors were able to do that. Bharath Kumar T V: Yeah. Absolutely and and all the safeguards that protect patients are very much of course in place So, you know, it's not that any in all the processes that are needed to protect patient safety are very much in place and the regulatory Processes are also very much in place ⁓ The intervention in this trial is they randomized patients in a 1 s to 1 ratio to either ketamine or it on with it They use permuted blocks of variable sizes and they ensured ⁓ allocation concealment by using sealed opaque envelopes and the randomization was stratified by sight there was no blinding that's important to know it's a pragmatic trial with no blinding the doses that they allowed were ketamine either 1, 1.5 or 2 milligrams per kg and the itomerate it was either 0.2, 0.25 or 0.3 milligrams per kg The clinician could choose the dose, but they'd given nomograms by weight and for all of these dose ranges. All other aspects of care were left to the ⁓ discretion of clinicians. other important, ⁓ so coming to the outcomes, ⁓ the primary outcome here was in-hospital death by day 28. And this is death from any cause between enrollment and 28 days. ⁓ But the outcome ascertainment would end at hospital discharge. So that's important to note. and they had one pre-specified secondary outcome which was cardiovascular collapse and they defined this as a systolic blood pressure less than 65, receipt of new or increase in the vasopressors and cardiac arrest. There were a number of exploratory outcomes. Specifically for this outcome of cardiovascular collapse, they had a trained observer who would assess this outcome who was not involved in the intubation but at the same time was also not blinded. That's again important to note and all assessments of cardiovascular instability were from time of induction until two minutes after the intubation. So that's the time point until which they assessed. Ramesh Venkataraman: I think the two minute windows is so that they only capture the effect of the drug and not the entire physiology of positive pressure ventilation and acid doses and the rest of the stuff that could cause hypotension. I think that's the rationale for the outcomes to be judged within two minutes of the induction agent being given. Bharath Kumar T V: Yeah. Yeah, yeah, I think they were trying to isolate the, you know, the sort of the pure effects of the sedative or the anesthetic, think. Yeah, and they, with this approach, they enrolled 2,365 patients and we can sort of talk a little bit about the results now. Ramesh Venkataraman: Yeah, so like said, ⁓ it's a pretty well done trial, I thought. About 2,400 patients enrolled, roughly 1,200 in each group. We just look at it on any trial. I just look and see if those patients are similar to the patients that I see in my ICU. If you look at it, the average age in the study was about 60. 40 % of them were women. Apache's score was 18. So not really very sick patients, but moderately sick. About 22 % of the patients were on vasopressors at the time of intubation. And almost 45 % of the patients had sepsis or septic shock. So that sounds like a ⁓ patient population that I would know deal with in my ICU and the most common reason for intubation is acute respiratory issues rather than no primary cardiac issues or primary neurological issues. So that's again something that most of us would encounter in our ICU. So if you look at the primary outcome, like you said, the 28-day mortality, there was no difference between the itomidate group and the ketamine group. But then going down to the next most important secondary outcome, which is the cardiovascular collapse as defined, you clearly mentioned the three criteria that was used. The ketamine group had about 22 percent cardiovascular collapse compared to the itombidate group which had only about a 17 percent cardiovascular collapse. So, there was an absolute risk reduction of 5 percent with itombidate compared to ketamine and that was statistically significant with confidence intervals are comfortably above 1, clearly showing that this results were good. So when I dwell in the data a little bit more in depth, one of the things that I was quite uncomfortable with is defining cardiovascular collapse as a systolic blood pressure of less than 65. I mean, we would never let systolic blood pressures go to less than 65. Cardiac RSTS, new or increasing vasopressor sounds reasonable around the time of intubation, but less than 65 is kind of odd. But the authors have given in the supplement and other materials very clearly that if you use a cutoff of systolic of 85 or a 30 percent drop in systolic blood pressure whichever criteria you use the signal was similar ketamine seems to be doing more cardiovascular collapse that kind of reassured me saying that no matter which criteria you used ketamine seems to cause more cardiovascular collapse at least in this study compared to itomidate. So, that is the kind of the gist of the study and do you have any thoughts on anything else or. The other two things I thought worth discussing Bharat are they throw these words one is sensitivity analysis other is subgroup analysis. I think for the audience to understand why would somebody do sensitivity analysis what does it add to the paper and when do you look for subgroup analysis and how do you take those results of subgroup analysis. Could you please elaborate on those two. Bharath Kumar T V: Hmm. Yeah, yeah, yeah, yeah, So on your first question, nothing much to add on the results. We will talk a little bit about which components seem to be driving the results and so on. Coming to subgroup and sensitivity. So subgroup analysis are typically exploring whether the effects of an intervention are differing in subgroups of populations, right, within the trial. And now these subgroups could be based on age, they could be based on sex, they could be based on illness severity. They could be based on a diagnostic category like sepsis or not. So it's trying to tease out whether there are effects of the intervention are differing by these subgroups. Now, it's important to note that most trials are not going to be powered to detect subgroups because obviously within each subgroup, are lesser number of patients. So they aren't really set to detect those differences. And even when there are differences detected, they are to be considered as hypothesis generating. So that's something that as readers of the literature, we need to be aware of. Ramesh Venkataraman: It just gives you a signal to say a particular group is more prone for this and maybe we need to look at this group in more detail in the subsequent studies. Bharath Kumar T V: Correct. Absolutely, ⁓ So that's definitely the reason for doing it. In this case, the researchers have taken a variety of approaches to exploring subgroup effects. We don't need to get into all of them, but the conventional approach is actually published as existing in the main manuscript itself. Now coming to sensitivity analysis, this is slightly different. Here what the investigators are trying to do is they are saying if we change ⁓ some of our assumptions in the methods or in our analytical approach, Do the results still stay the same? Do the results still stay consistent? So that is what a sensitivity analysis is trying to do. And when the results stay consistent, or if the results stay consistent, it reassures me that the primary approach that I've taken is reasonable, and it increases the internal validity of whatever my findings are. So again, in this paper, start their primary approach or their primary model has only the intervention and the site into the model as covariates. and then they do a number of sensitivity analysis. They do one unadjusted analysis, they do one with additional covariates, they do a best case, worst case and so on and so forth. And importantly, all their sensitivity analysis are agreeing with the primary results, which again sort of reassures me that the primary approach is robust. Ramesh Venkataraman: So, sensitive analysis kind of gives us confidence that this study results are robust and I can kind of trust the results that the signal is consistent no matter how I look at it right. And subgroup analysis helps me to say okay which patients are more prone or less prone for this particular thing and should I evaluate that particular population in detail going forward if it is worthwhile. The only thing I would like to add is subgroups should be pre-hawk decided right before that ideally the subgroup should have been defined Bharath Kumar T V: Yeah, absolutely. Yeah. Ramesh Venkataraman: ahead of time, which the authors have done here and can't be added after the study is done for the sake of just kind of splitting the data into multiple things to get results. So that's something the readers have to be very clear. Watch out for subgroup definitions that have been done pre-hoc. having gone... Yeah, go ahead. Bharath Kumar T V: Yeah. Yeah, yeah, Sorry, no, I was just going to say that that's a very important point that you raised that they have to be pre-specified and there should be some biological rationale for testing in those subgroups as well. So that's again something to think about. Ramesh Venkataraman: Absolutely. So here I think it made sense. They looked at sepsis, sepsis-6 shock, high Apache score and so on. So all of those made sense to me. I thought overall it was a very well done study, Bharath. I think there's a lot of positives in the study. One, two, like you said, the waiver of informed consent. It's a tough study to do. You have very little time to intubate somebody and then within that time you got to somehow manage and do the randomization right. So they got waiver of consent. The randomization was already done with envelopes and Bharath Kumar T V: Mm. Ramesh Venkataraman: And the amazing thing was 99 % they got it right. Now that is really, really impressive in a multi-center study across emergency rooms and ICUs. And the trained observer who was unbiased and who's not a part of the intubation team quietly observing and basically looking for the outcomes that have been defined, I think is again another strength of this particular study. And ⁓ I thought about 70 % of patients who were screened were enrolled. That's a very good number because a lot of times you exclude a lot of patients and then I wonder whether this is the kind of patient that I would intubate in my ICU. But if you enroll 70 % of the patients, no more often than not, maybe except pregnancy and trauma, looks like they've included everybody. And that kind of generalizes and makes me believe that the study is something that I can take to my ICU and use it for practice. So I thought it was a greatly done study, Bharath. What do you think? Bharath Kumar T V: Yeah, yeah, absolutely. I think it's a fantastic trial. And as we said at the beginning itself, it's a super important area. So I echo all of what you said. I had a couple of maybe concerns to point out. ⁓ One is the fact that the cardiovascular collapse, the outcome is primarily driven by an increase in vasopressors or new vasopressors. Now in an unblinded trial, obviously this is... Ramesh Venkataraman: Just to pause you and recollect for the audience, cardiovascular collapse is defined as one, cardiac arrest, two, systolic blood pressure less than 65 and three, ⁓ increase or new addition of vasopressor after the intubation has happened within the first two minutes. from like, mean, what the point you're trying to make is most of the difference in this group, the 5 % difference happened in the vasopressor criteria rather than the cardiac arrest and the systolic blood. pressure, right? Bharath Kumar T V: Correct, correct. And my concern stems from the fact that this is an unblinded trial, right? Everybody is aware of obviously the drugs, the assignment. So the thresholds for increasing vasopressors particularly is obviously going to vary between clinicians and is, I guess, subject to some bias creeping in. Ramesh Venkataraman: Thanks. Bharath Kumar T V: So that's something that I think again readers should, I mean listeners should sort of consider. The other concern that I had is I was keen on understanding what is the steroid usage post intubation until 28 days. We know that the number of people taking steroids to begin with was sort of similar between the two groups. but I was interested in knowing what proportion got steroids after. If you go to the supplementary appendix, there is this information provided. So I think 49 % in e-tomidate receive and 46 % of patients in ketamine receive. So there's a slightly higher use of steroids in the e-tomidate group. This is not statistically significant. So obviously, maybe you can't read too much into it, but there is some sort of a trend. But what is the... Ramesh Venkataraman: Thank Thank you. And if 45 % of the patients had septic shock, they would have received steroids anyway, more likely. I mean, very difficult to say, Bharath Kumar T V: Correct. Exactly, yeah. So I'm not sure what we can read into it. But what is reassuring is that the mortality is not different. So that's something that sort of reassured me. So those are a couple of concerns that I thought I wanted to highlight. ⁓ those are my thoughts. But I want to get back to you, Dr. Amation, and get your final thoughts on the trial and what's your takeaway, I guess, from the trial. Ramesh Venkataraman: Only addition to what you said Bharath is I agree that most of the results are driven by new or increased vasopressor, but in the itomidate group before intubation 2 % of the patients, mean 2 % more patients got vasopressors already started prophylactically before intubation. So it is possible that maybe they they averted increase of vasopressor after intubation by doing it just before intubation 2%. Bharath Kumar T V: Yeah. Mm-hmm. Hmm. Ramesh Venkataraman: I do know whether that is going to make a difference or not. And the second thing is the contrary, right. If I have already prophylactically increased my vasopressor, my tendency to increase the vasopressor post intubation may be slightly lower. If ketamine group did not have the prophylactic increase in vasopressor as a clinician, no, I may just say, okay, tweak the norepinephrine a little bit higher right after intubation to avoid the hypotension and by definition that is cardiovascular collapse. In a open label study this is possible but I not saying this is what drove the outcomes in one place or the other. I think overall I think the study is very well done. We discussed the details of the study. My personal take home message out of all this Bharat is that traditionally at least before the study I thought ketamine was a very hemodynamically friendly drug and I can probably get away intubating most ICU patients with ketamine. Bharath Kumar T V: Mm. Ramesh Venkataraman: Now I have a no ⁓ extra respect for ketamine induced hypotension. is real, especially if you have patients who are already catechol depleted, right? And patients who are in septic shock were already on high dose vasopressors where the ketamines effect to release catechol is gone. And then all you're with is the negative ionotropic effect and the vasodilatory effect causing cardiovascular collapse. So the first thing is the first take home message is irrespective of the drug you use. I think as clinicians we got to be ready for cardiovascular collapse and have your catechol purges or your infusion ready to go or if you would want to start a low dose vasopressor and really hypotensive patients or increase the dose by all means go ahead and do that. Second like the common sense thing I can say is if this study doesn't say ketamine is bad and it dominate is better it only says that No, do use one drug for all inducing patients for across all diagnosis, right. I would say ketamine is probably something I would avoid after reading the study in patients with severe septic shock on multiple vasopressors. They already have very severe RV or LV dysfunction and their chance of cardiovascular collapse is very high. They are higher risk of ventricular arrhythmia say acute coronary syndrome in periacute coronary syndrome patients with high risk of ventricular arrhythmias. maybe I can electively stay away from ketamine after the study because it can exacerbate the effects. In all other patients ketamine especially if it's state as asthmatic, state as epileptic, somebody where I want to delete sequence intubation for whatever reasons I think ketamine may be a great drug. And in patients where I want to avoid ketamine I would probably go with the Itomidate. As you said the steroid use is not a big concern there is no mortality difference I think Itomidate may be a good drug. So my take to Bharath Kumar T V: Yeah. Mm. Ramesh Venkataraman: My main two take home points are one, don't use one drug, choose a drug that will suit the physiology of the patient, especially after this. Two, whichever drug you use, a healthy, I have respect for the drug induced hypotension and cardiovascular collapse and be prepared ahead of the time. think those two are the two things I would probably change my practice after reading the study. How about you Bharath? Bharath Kumar T V: Absolutely, I echo absolutely echo whatever you've said. I think we need to choose these drugs in the context of the clinical situation. ⁓ Broadly, I would say that if you are somebody who is comfortable using Itomerate, have access to Itomerate, it makes sense to use Itomerate for most patients with some specific exceptions. But if you're going to continue using ketamine, think heightened vigilance, better optimization, all of these are going to become very, very important. And as you said, ⁓ I think we have an increasing sort respect for the kind of ⁓ consequences of using ketamine and we definitely need to better prepare for as we intubate through the peri intubation period. ⁓ Ramesh Venkataraman: Only one point, itombutate ⁓ used personally and what literature says is you do get into poor laryngoscopic view compared to ketamine because of mesy trespassin. Right, commonly suboptimal doses especially can cause mesy trespassin and make it difficult. In this study, almost everybody got paralysis, succinylcholine or rocoron in both groups. So, be careful that no this study did not look at that particular outcome and did not Bharath Kumar T V: Mm. Mm. Ramesh Venkataraman: even though they give you data, there was no difference, understand that patients were completely paralyzed. In patients where you are not able to completely paralyze for whatever reason, because you think it is an anatomically difficult airway, itomidate can probably make it slightly more difficult to intubate compared to ketamine. Just a point I want to throw there out there for people to just think about the various factors. Bharath Kumar T V: Absolutely, and I think that becomes very relevant if you have never used Detomidate before and decide to start using it now after this trial, right? Because you may not be familiar with what are the consequences and what challenges you may come into. So anyway, on that note, again, a fantastic trial. Congratulations to the investigators. it's as always fun talking about all of the clinical but also the methodological nuances. We're going to stop there. We want to thank everyone again for listening in, for viewing, for providing us lots of feedback. Thank you so much. Ramesh Venkataraman: Thank you. Bye bye.