Latest / CRITKNOW / Bicarbonate for in-hospital cardiac arrest
Transcript
- Bharath Kumar: Hello everyone and welcome back to Critno. This is Bharat Kumar, and I'm joined by my co-host Ramesh Venkatraman. We want to start by apologizing for the relatively long gap between our previous episode and the current one. We were both held up with some personal and professional commitments, and we are excited to be back here again. We're also very grateful for all of the support and encouragement that you continue to provide us. So thank you so much. In today's episode, we are going to look at another. bicarbonate trial. this time evaluating bicarbonate in the context of cardiac arrest. I'm referring to the bicarbonate in in-hospital cardiac arrest trial that was pu published in Jama a couple of months back. So Ramesh, interesting question. Ramesh Venkataraman: Absolutely Bharath. I think it's become a norm in most clinical practices to give bicarbonate during a cardiac arrest resuscitation if ROSC is not obtained within a reasonable time. And it's become ingrained as being one of the five Hs that we have to tick off, especially in a rhythm that is non-shockable. And therefore, this is a practice that has been happening for no decades. Studies show that almost up to 50 % of cardiac arrest patients in the in hospital setting do get bicarbonate somewhere during their resuscitation. But if you just step back a little bit and think about why we give bicarbonate, the commonest reason anybody would give is if there is acidemia contributing or pre-existent prior to cardiac arrest, I need to correct it. And in the presence of acidemia, the catechol responsiveness is going to be very, very low and giving bicarb would help me make sure that the other drugs, especially the adrenaline that is given actually works. But there are also people who feel that bicarb not only may not be helpful, but could potentially harm. The commonest mechanism being it causes paradoxical intracellular acidosis. generates, especially boluses generate acute CO2 production that can actually diffuse into the cells. and that can cause it. Two, it can push the potassium into the cell calcium and therefore you can have hypokalemia and hypocalcemia. All of that can be bad during a resuscitation and for achieving a stable rhythm. But if you look at the guidelines, both ELCA or European Resuscitation Council have clearly made a statement that bicarbonate therapy is not recommended routinely in in in inpatient cardiac arrest. except in conditions where there is a clear reason there is a hyperkalemia pre-existent tricyclic antidepressant overdose etc things like that. But this recommendation is not based on any evidence it has just been extrapolated from out of hospital cardiac arrest studies. And coming to the out of hospital cardiac arrest studies there have been three major studies two studies a long time ago and the most recent study, all of that together basically showed that bicarbonate does not really improve outcomes even in out of hospital cardiac arrest. So that's where the literature has brought us so far. So there is physiological rational, but ambiguity about its benefit in out of hospital cardiac arrest set up, doesn't work. And people who believe in bicarb would say in out of hospital, Maybe that there is a delay in time to give bicarbonate and if I'd given it earlier, it will help. in hospital cardiac arrest, patients are more likely to have pre-existent metabolic acidosis. And therefore, it is a different category of patients who may benefit from bicarbonate. So that kind of puts things in perspective. And I Bharath Kumar: Mm-hmm. Ramesh Venkataraman: think this is a very relevant question that we commonly see. And I think this study kind of answers that very pertinent question. Bharath Kumar: Yeah, of course. I think like as you said, there's a strong rationale, some potential downsides. you know, there is of course evidence of equipoise and uncertainty on whether we should be using this. So I'll dive into the PICO of this trial. So this is a trial from Denmark across 21 hospitals, and the population they enrolled is adult patients with an in-hospital cardiac arrest and having received at least one dose of epinephrine. Now there are some exclusions, most of which are reasonable, DNR you know, directives prior to already in place, prior enrollment in the trial, if the patient was on invasive mechanical circular to support the time of the cardiac arrest, if they were known or suspected to be pregnant, if there was a known objection to participating in any clinical trial. And as you said, if there's any prior clinical indication for giving bicarbonate, or the indicate or bicarb is indicated. For the etiology of the arrest, then those are exclusions. The intervention is 50 ml of 8.4% bicarbonate after the first dose of epinephrine, and they allowed for one additional dose to be given after the second dose of epinephrine. The control was a matching placebo, which is just 0.9% saline. And the primary outcome, important to note, is the return of spontaneous circulation and it has been sustained for at least 20 minutes. number of secondary outcomes, which are the usual ones in postcardiac arrest trials, which is survival at different time points and then survival with a good neurological outcome, quality of life, and so on. And coming to the sample size, so as we said, the primary outcome is ROSC. So they postulated that their assumptions were that in the control arm this would be 33%. And they postulated a 10% benefit in favor of the bicarbonate arm. So they said it would be about they anticipated it to be about 43%, a 10% difference. And we can of course talk about whether that's appropriate or not. standard alpha and a standard power of 80%, and this gave them a sample size of 780 patients. in the on the consent, I just wanted to you know highlight something that is slightly different. so obviously incapacitated, these patients are typically incapacitated, so they could take consent from the next of kin, but they also had this provision for taking consent from an independent physician, which I think is different than probably allowed in the jurisdiction where the trial was conducted. And their analysis I wanted to again the analysis is generally straightforward and and I think pretty appropriate. but the only thing that is important to highlight is that their ITT is already a modified IT population in that they only included in their primary analysis those that received at least the first dose of the trial drug. And met all the inclusion and exclusion criteria. So they did have to exclude some people to arrive at this. This would sound almost like a per protocol, but it's a modified ITT that the approach that they have taken. And their justification is in this case because it's a blinded trial, it perhaps doesn't you know bias the estimates. they had a number of subgroups by age, witnessed and unwitnessed arrest, and then based on the initial rhythm. So that's you know, sort of in a nutshell, the PICO. the sample size and the analysis. so what did they find, Ramesh? Do you want to summarize the Ramesh Venkataraman: Yeah, so interesting though, I found two things interesting in the methodology. One is the consent by an independent physician that is unique to their setup and that's novel for our listeners to kind of know about. Second thing I thought they did a great job. They were very transparent about why they excluded people. And it included all kinds of logistical reasons which I can relate to, right? The drug was not available, there was nobody to give the drug, the doctors just forgot during the CPR. I mean, all of that could happen in real life with us and I could relate to it. And I actually felt that they were very, very honest and investigators had to be applauded rather than blindly shoving everything under the carpet saying logistical reasons, these were protocol violations. So I thought that was something worth pointing out and something that I could relate to. And this would happen in my ICU too. So I think kudos to them. So coming to the results, Bharath, I felt the patients were fairly representative. and generalizable to our ICUs. The age of the patients were somewhere in the 73 to 74 range. 86 % of the patients are non-shockable with them compared to shockable with them. About 13, 14 % of patients had metabolic acidosis prior to the cardiac arrest. And if you think about the time to adrenaline was... about five minutes on both groups and time to bicarbonate is only about eight minutes from the onset of cardiac arrest to be able to give the first dose of bicarbonate. So I thought that was quite impressive. All the intracardiac co-intervention seemed to be similar between the two groups in terms of amiodarone, lidocaine or whatever they had to give was very similar. And if you look at the primary outcomes, The ROAC was sustained, ROAC of 20 minutes was obtained in 39 % in the sodium bicarbonate route and placebo group, which is basically no big difference. All the secondary outcomes were almost similar between the two groups. Time to ROAC, 30-day survival, 30-day survival with good neurological outcome and so on, all of which was very, very similar between the two groups. At the end of resuscitation when they looked at the patient variables the pH was 7.11 in the bicarb group and 7.03 in the placebo group. The standard bicarbonate was 17.7 in the bicarb group and 13 in the placebo group. So there was clearly a separation between the two groups in terms of how that happened. There was obviously a higher incidence of hypernatremia, alkalosis. hypokalemia and hypocalcemia in the group that got bicarbonate. But as I said, no difference in the sustainable sustained ROSC return between the two groups. So that was the final results Bharat. What do you think about the results at this point of time? And I mean, one thing I forgot to tell you all the subgroups that you mentioned that absolutely no different between the Bharath Kumar: indifference. Ramesh Venkataraman: two groups either. So essentially no no differences in the bike card versus no bike card in any of the outcomes that they Bharath Kumar: Yeah. thanks Ramesh. So I was Ramesh Venkataraman: Thanks. Bharath Kumar: to me this is I mean, I think it's a great trial and well done and a really important question, whereas you were saying at the beginning that, you know, we use it routinely without any data. But looking at the results, I'm also wondering whether we shouldn't be surprised because of a number of you know points that that occurred to me and and I'm sure you will you will have some additional thoughts. So one is the dose, right? so they are only giving two doses of fifty milligrins of bicarbonate. And there's a I think a proportion of patients that don't get the second dose, some for the right reasons, you know, either ROC is achieved or they have you know stopped CPR, but there's also other reasons why second dose doesn't get administered. So hmm. Second dose. Ramesh Venkataraman: About 30 % of patients in both groups didn't receive the second dose. They got only one dose. Bharath Kumar: Right. So really, you know, a large at least a substantial portion of patients are not getting even the second dose. So it's just about 50 milli equivalents of bicarb. So is that, you know, is that enough? Right? that is one question that comes to mind to make a difference. the second point that came and occurred to me is that this is obviously all patients that they're choosing and they don't have, you know, I don't think they have done a ABG or a gas during the resuscitation to decide. who should you know get more bicarb? Should everybody get bicarb? Right? So, this is being given to everybody, but who really which group is going to need it the most? I don't think this trial can can get at that, right? So, which population or subpopulation needs it the most, and whether ROSC is the most appropriate outcome. And and I I can see the rationale. this is an intervention that is, as you mentioned at the beginning, right? It helps the myocardial contractility. Helps the catecholamines to act. So maybe ROSC is the right outcome. But typically we are interested in more patient-centered outcomes, more longer-term survival, survival with a good neurological outcome. So those are some of the thoughts that occur to me. I mean, maybe the results are not surprising given these limitations, which of course have been acknowledged at least partially by the investigators themselves. I'm curious to know or hear no what what were your thoughts on what you know how you interpret the results. Ramesh Venkataraman: No Bharat, think all of your points are very well valid. So I'll just take points from your thoughts and then I'll expand on them. One about the dose of bicarbonate. It gets difficult to kind of titrate bicarb that have been during a code to do something. So that's the rationale for a fixed dose. And most of the time in codes, people just arbitrarily give whatever bicarb they have, at least they've standardized that. Like you rightly said, whether the dose is adequate, even though there has been a difference between the two groups in terms of the pH and the bicarbonate, as I told you, the bicarb group had a post resuscitation pH of 7.11 and the placebo group at 7.03. To me, it means that both of them had bad enough acidosis and the bicarb really did not correct the acidosis to improve the myocardial contractility or catecholamine. So maybe there was an under dosing. to the physiological endpoint was not achieved despite the bicarbonate and that could be because a lot of patients received only one I mean good good number of patients could not receive the two doses. So that just to add on your dose issue is something that I would think of. Second thing you talked about whether RASC is the right outcome. I do agree with you that RASC seems to be a relevant outcome because we looking at the process of resuscitation and the first clinically relevant endpoint would be achievement of a stable ROS for 20 minutes. Here I would like to add that they only looked at the first cardiac arrest. If the patient arrested again, those patients were not included in this trial. that again looks at they are stable rhythm for 20 minutes and this is a blinded trial. So hopefully no any other confounding factors were removed. So that is the first primary relevant clinical endpoint for all of us as clinicians. You have to survive and get a return of spontaneous circulation to be alive at 30 days. But the problem, Bharath, is that there are so many interventions we do in the post-cardiac arrest state if somebody develops a ROSC that could affect the 30-day outcome. Temperature management, for example, CO2, hyperoxy, glycemic control, shock management. So there are so many other things that do that. Reason for shock, revascularization, so things like that. So when we have so many other things that could potentially impact the 30 day mortality, to expect bicarb to show a difference there would be unreasonable I felt. So I think 20 minute of stable ROSC I think seems to be a very valid outcome in my mind. So that's number two. I wanted to pick your thoughts, Bharath. Do you think a 10 % absolute risk reduction in power analysis I felt was a little bit optimistic to say? Now I will show a difference of 10 % of RASC more in the bicarb group compared to the placebo group. What do you think? Bharath Kumar: Yeah, I I mean I sort of was trying to process whether we should call this too optimistic or not. But then looking at the, you know, in cardiac arrows literature, we know that the 30-day survival and survival of the good neurological outcome is really bad. As even we can see in this trial, right? Even under the best control conditions, it's about ten percent or so. So with the ROSC as the primary outcome, I mean, if the intervention is not really doing something substantial, right, at least achieving know 10% higher Rosk, when you look at it from that context, if it doesn't even do that to out in terms of a 10% increment in Rosk, then it's going to be a challenge to advocate for rolling out such interventions because Rosk is really the the first step in a long journey in the road to recovery. So if you look at it from if I looked at it from that framework, I was thinking maybe 10% as reasonable as you know as as an as what they should be looking at. yeah, I again w wha what were did you did you have a counter thought to that or yeah. Ramesh Venkataraman: No, no, I think your points are valid. No, if I get 10 percent raw, if I don't even get a 10 percent, so I could have powered the study for a 5 percent or a 2 percent increase. One, it would have led to substantial amount of patients that were necessary, which is just not pragmatic. And two, if I can't get that reasonable difference in raws, then the proportion of patients I'm going to say with intact neurological outcome is going to be minuscule and doesn't matter anyway, studying this question probably doesn't matter anyway. So I agree with you on that thought, Bharath. But I felt that the study was overall very, very well done. Doing cardiac arrest studies are not easy. I think their milieu allowed them to get consent from a close family member. And I think the independent physician was a very, very good cross check in terms of making sure that's the right thing to do for this patient. I think the time to... giving the test drug is very, fast. Eight minutes to get to the bicarbonate was very, very good, and it's not that easily achievable. So I think they have done really well there. And then I think their outcome of rask of 20 minutes seems reasonable to me. I think that that's also something that I would go ahead. And yes, we did discuss the limitations, and I think the take home message is very straightforward, right? At this dose, bicarbonate doesn't really help. if you, like we always talk about, the study should not be looked in isolation. It should be looked at the context of the previous studies. We have had consistent studies in out of hospital cardiac arrest where bicarb didn't help. It's the first study in hospital cardiac arrest, again, echoing the same results. And looks like probably bicarb therapy is probably ineffective in terms of... improving RASC or a 30-day mortality outside the classic indications of a pre-existing cause like a hyperkalemia or a tricyclic overdose that we thought about. But Bharath Kumar: Yeah. But Ramesh Venkataraman: it opens a few questions. One, can I titrate my bicarb therapy to a pH? Now, if I'm able to do know ABG's during the arrest and somehow get more bike up in to get the pH above 7.25, 7.2 in the study which they couldn't do. Would that make a difference? I don't know. It seems unlikely because that would take more time and more effort. It's not pragmatic. It's not going to be applicable at the bedside across all centers. So that kind of study is going to be very difficult. But I think this study will probably be the nail on the coffin in terms of bicarb in cardiac arrest, which is pretty much now the practice for most people. I think the Hail Mary bicarb can safely be stopped. I feel at this point of time, a desperate bicarb can be stopped unless you have a strong reason to suspect hyperkalemia in that particular clinical scenario. my take home is protein bicarbonate in in-hospital cardiac arrest is ineffective and I would probably not use it at this point of time, Bharath. What do you think? Bharath Kumar: Yeah, no, I was actually going to ask you where you thought the research would go next. And I agree. I mean, I think as you said, should it be titrated based as based on you know during arrest ABGs? The other question, of course, is is the effect different if you already have acidemia to begin with, right? At the point of arrest, which is not the question that this trial addresses because Ramesh Venkataraman: This patient population, only about 14 % of patients had metabolic acidosis going in. Predominantly, they didn't have. So that question still remains, Bharath. It's a valid Bharath Kumar: Yeah, yeah. But I otherwise agree with you that, you know, for all arrests and sort of throwing bike up perhaps does not make sense. no no other additional thoughts and maybe it's a good point for us to stop this episode. we wanna thank everyone that's as I said supported us on this journey. This is also our tenth episode. I think we started way back in March and and we're very happy that we managed to get to this point. once again, if you've been enjoying what we've been doing, please do comment, like, subscribe, share. thank you so much. Ramesh Venkataraman: I also want to add Bharath that there are several new exciting trials that have just been published yesterday and day before from the European Society of Intensive Care Medicine meeting. And we were excited that there is enough material for us to kind of go through, read in detail and then come back to you all with our thoughts on it and hopefully, you know, enjoy the process as well. So. Thank you all and we see you with the next podcast sooner than this time, think. Bharath Kumar: Thank you. Ramesh Venkataraman: Bye-bye.