Emerge in EM

E9: EtCO2 vs PCO2 in critically ill prehospital patients

• Mohamed Hagahmed

Share episode

Send us a text

In this episode, I'm  joined by Dr. Chris Martin-Gill to discuss the critical evaluation of the correlation between end-tidal CO2 and partial pressure of CO2 (PCO2) in acutely ill patients. The episode covers the importance of end-tidal CO2 in the prehospital setting, primarily for verifying airway placement and assessing ventilation adequacy. Dr. Martin-Gill discusses a significant study that he co-authored that analyzed data from over 6,400 patients, finding large discrepancies between end-tidal CO2 and PCO2, particularly in critically ill patients. The discussion highlights the implications of these findings on EMS protocols and the importance of considering patient subpopulations. The episode concludes with recommendations for EMS agencies on how to use end-tidal CO2 readings effectively and the need for further research.

Paper reference: 

PMID: 39546437 DOI: 10.1080/10903127.2024.2430394



Mohamed (2):

Welcome folks. Back to another great episode of Emerge. This time I have one of my colleagues and my best friends that I admire always, Dr. Chris Martin-Gill. Please tell us more about yourself and what you do.

Chris:

Thank you for having me, Mohammed. I appreciate being able to be on with you. So I'm an emergency and EMS physician here in Pittsburgh working at UPMC in the University of Pittsburgh. And I work in EMS medical direction and with a variety of colleagues such as yourself.

Mohamed (2):

Thanks Chris. I specifically chose this topic because I felt it was very relevant to the prehospital and critical care transport community. This is a topic of the end-tidal CO2 and the partial pressure of CO2 and the correlation between the two in acutely ill patients. Now Chris, obviously you happen to be one of the authors of this great paper. First, let me just talk about the basics a little bit. Why is end-tidal CO2 so important in the prehospital setting?

Chris:

Yeah, so that's a great question. So just working as an ALS practitioner in the field, we know that end-tidal capnography has become a cornerstone of the care that we provide whenever we're managing somebody who's got, respiratory issues, and particularly anybody with an advanced airway or undergoing mechanical ventilation. And we use end-tidal capnography most often often to ensure that we have placed an advanced airway correctly. We use it to assess the adequacy of ventilation. We have even used it to predict outcomes in cardiac arrest or sepsis or trauma. And the assumption has generally been that end-tidal capnography is serving as a good or reasonable surrogate for the partial pressure of carbon dioxide in blood, the PCO2, and in some studies, particularly studies of normal humans, that correlation is actually really good and has been cited to be about within about four millimeters of mercury difference between end-tidal and PCO2 and so when we're thinking about trying to keep that within a range of about 10 points, that correlation is good and we can use it as a pretty good surrogate ventilation and things like that. However, some small trials had suggested that correlation could be highly variable, particularly in sick patients,and anecdotally in our own critical care transport environment. Once we initiated the use of blood gases very actively several years ago anecdotally we realized that we were finding substantial discrepancies between the end-tidal readings we were getting and the PCO2 readings that we were getting from point of care blood gas assessments. And so what we wanted to do once we had been using that clinically for some time is to, go from that anecdotal experience now that we had the availability of a very large data set to try to figure out what exactly that relationship is between the values of end-tidal and PCO2 values.

Mohamed (2):

Truthfully, Chris back in my paramedic days, which feels like a long time ago I feel like the learning of end-tidal came in two ways. The numeric value, which is I. This number range that always was like, instilled in our brain, which are 35 to 45, and then the graphic representation of the entitled CO2 when we were told to always be scared of the shark fin. You know what I mean? In COPD patients, but it wasn't until recently a few years ago, since we started using end-tidal CO2 in a broad range of patient presentations, like you mentioned TBI, cardiac arrest. Shock acidosis, all these things, and it makes sense to keep it within an a, a specific value or range. But this paper, as you mentioned gives us a different story, and that's the discrepancy between the end-tidal CO2 and the PCO2 in a subgroup of sick, critically ill patients with an advanced airway. Can you give us a little bit more about the paper that you co-authored?

Chris:

Yeah, so you bring up some very good points that I think we need to tackle first, which is that, the end-tidal measurement, the assessment continuous waveform capnography really brings up a lot of important a lot of important information. So when, if you're looking at the shape of that waveform, you can gain a lot of information about somebody's pathology. Just the fact that you're getting some end-tidal reading is clearly a measure for us to be able to determine, for example, if airway placement. And that is critical in our environment. But really what we wanted to get down to is, was the specific numbers. What are those readings. And especially when we recognize that in our environment we have protocols that really try to get us to titrate our ventilation to a very finite tidal reading. And we really wanted to figure out how well that correlated with. True PCO2 values that we would think is what we really want to target. we ended up with a fairly large data set to look at this. We had over 6,400 patients that underwent critical care transport, and these were all patients with an advanced airway in place that were undergoing. Mechanical or BVM ventilation mostly with mechanical transport ventilator. And what we're able to capture is that those patients had an tidal reading and a PCO2 from a point of care blood gas assessment within five minutes of each other so that they should be correlating temporally, certainly. our goal was to analyze the delta CO2. So essentially the difference between the end-tidal reading and the PCO2 to try to figure out what that variability is.

Mohamed (2):

And obviously you found some differences, and we're gonna talk more about those results specifically. But I couldn't help myself. When I looked into the paper further, it looks like you had some side questions that you attempt to answer by exploring some of these analyses. What are these questions and what were you looking for?

Chris:

I'm usually very mindful of what the impact of these questions can have on clinical protocols at, being involved in the development of protocols for our critical care transport agency, as well as thinking about the protocols that we use, from a state basis or, across the nation. Where they wanna think about if there are discrepancies between the end-tidal and PCO2 readings. Is that really primarily in certain subpopulations? Is it that we can believe that there's a good correlation in most patients, but in certain patients, there's a reason that there's a discrepancy and we can't really go off of the end-tidal for an exact. Number or is this a more global issue that we need to look at? And specifically if you think about how you use the research in writing the actual protocols, you wanna try to figure out which patient populations that really needs to apply to and how you may be able to use it. We wanted to look at what that gap was across all of our patients, and then we wanted to specifically look at some of those populations. And the subpopulations that we looked at were, comparing folks that underwent inner facility transport versus scene transport. We wanted to look at those of different medical categories. Thinking may be those with respiratory complaints versus trauma or neuro patients that you could assume had no primary respiratory issues, were similar or are otherwise correlated, we looked at patients with hypotension. If you think about the decreased perfusion of the lungs may be a reason that you have a mismatch there, or even folks with an elevated respiratory rate or hypoxia where it is more singling out those that have a respiratory issue versus other patients.

Mohamed (2):

So these were mostly intubated patients. Some of them had an advanced airway like supraglottic airway other than an ET tube. And they were transported by a critical care transport unit in our agency that was STAT Medevac, which is the flight critical care team, and a combination between interfacility transport and also scene transport. So what did you find, Chris, when it comes to looking at those data?

Chris:

The results we thought were quite striking. So the median difference between the end-tidal reading and the PCO O2 was 12.9 millimeters of mercury. So it, 13 point gap found in 50% of our patients, right? And 60% of the patients had a gap that was greater than 10 points. And even 30% of our patients had a gap that was greater than 20 points. And really, that's a significant deviation, particularly as you think about EMS protocols often wanting us, and I say that writing, I. Some of our own protocols, we write our protocols and we work in our environment where we try to ventilate patients for an end-tidal goal that is fairly narrow from 35 to 45. Recognizing now that at least in these critical care transport patients, the gap itself. Is 10 points in 60% of patients in 20 points. Outside of that 10 point range that we're trying to ventilate somebody within in 30% of our patients. So that's quite striking in the implications that it can have in terms of how we use that end-tidal reading for how we adjust our ventilation is really profound. And we found even greater discrepancies, which actually make physiologic sense in those that were hypotensive, those that were hypoxic, where there was at least a 10 point gap. Two times the likelihood odds ratio for patients that were hypotensive to have at least a 10 point gap and even a greater likelihood of that discrepancy existing and those that were hypoxic, IE those that had a specific respiratory complaint.

Mohamed (2):

Wow, that's a striking difference the delta between those subset of patients. Now it makes sense for scene runs for those subset of patient population to have a larger delta because they're acutely ill and they have other comorbidities. What about the inner facility patients?

Chris:

Yeah, it was interesting the way that, that we ended up asking that subpopulation question was actually flipped. So I was interested in trying to figure out if this was really just an interfacility problem. Once a patient had been sick for some period of time, now you have a ventilation:perfusion mismatch, and maybe on scene runs, which you think might be closest to your ALS population. Maybe the gap wasn't that. Significant in scene runs, and so from a practical standpoint, maybe us doing a point of care blood gas assessment really only matters so much when it's an inter-facility transfer. And maybe we can just base our treatment on an end-tidal reading on those quick scene runs when we're trying to get off that scene fairly quickly. Fortunately, or unfortunately, it turns out the data told us otherwise. It actually found that for scene patients, the likelihood of at least a 10 point gap was greater compared to inter-facility cases, though it was an issue across the entire population. And in fact, there wasn't any discreet subpopulation that we looked at where we could say, oh, okay, in this little subpopulation, we can just go based on the end-tidal'cause it always correlates or there's not a substantial subset where it isn't correlating very well. Now it turns out that in the whole population, in any way that we looked at our subpopulations this was an issue to actually think about.

Mohamed (2):

I guess that brings up very important question, at least for me when I go out there on the field, and obviously I tend to over utilize end-tidal CO2 in a lot of situations and even in the hospital for you and I, Chris, as well. We use end-tidal CO2 a lot in sedation and other procedural stuff as well, but for sick patients. Does that mean we should just ignore end-tidal CO2 altogether? How do you see us implementing end-tidal CO2 in a useful fashion in critically ill patients?

Chris:

I can tell you what we've already done in our own regional critical care transport service. And that's obviously knowing that these data were coming for some time and knowing our own QI experience. Even before this publication, we had already altered our own clinical protocols. To base a lot of our ventilatory management on blood gases when we're able to assess those and we assess them at this point in most of our patients that are undergoing mechanical ventilation, because we understand now that we really need to assess what the pH is and the PCO2 are in the process of deciding what, how to ventilate somebody and really try to gauge whether in that patient there's a good correlation between PCO2 and end tidal before we use the end tidal as a surrogate to make further ventilatory management changes during the transport. The real question that remains, and this is a limitation of our study, is that I can't tell you what this gap is in your average ALS patient population. However, if you think about it, I. A post cardiac arrest, which is the most common patient that's intubated in the ALS environment. A post cardiac arrest patient is not less sick than any of the other critical care transport patients we're talking about. So if you think about the likelihood of there being a substantial VQ mismatch, when you're talking about hyperperfusion, and we're talking about aspiration or other reasons for there to be a a ventilation issue. We're really talking about the high likelihood that these discrepancies also exist in ALS patients, if we had the opportunity to check point of care blood gases in that environment. But that's really where we need to look at next. I think that the reality is that we need to reconsider closely how well we're able to realistically titrate somebody's CO2, the true value We're trying to end up in what's actually in their bloodstream to a narrow range between 35 and 45, or even narrower in certain patient populations, like some protocols for traumatic brain injury, where we might say target right at 35, 35 to 40, 30 to 40, you're trying to get to a very narrow range or even think that we're going to adjust by five points one way or the other. When actually the gap of what the measure is substantially higher. and we're really gonna have to one, hopefully have more data for an ALS population if we're able to get point of care testing in that environment. and two, really think about how we might adjust our ventilation parameters for other components that we're assessing on patients, not just assuming that the end-tidal is that exact correlation of what we think that it is.

Mohamed (2):

So the idea that I'm getting from you, Chris, is that for those agencies, EMS agencies that don't have accessibility to a point of care gases, they only have end-tidal CO2, and that's what they rely on for all of their patients, right? So the idea is don't be falsely reassured by the number right in front of you. Just look at the patient who is in front of you and try to manage their hemodynamic stability or whatever physiology, whether it's shock acidosis post cardiac arrest when it comes to trauma. So not just solely rely on the number in front of you, however, just make sure that all of the other things are addressed first, like hypoxia, hypothermia, hypotension.

Chris:

Yeah, for sure. And I don't, I don't wanna overstate any results, particularly because I very much recognize that we did not study an average ALS population. And I also want to emphasize that end-tidal readings have great value in our environment, and particularly end-tidal readings tend to underestimate PCO2 values, and so if the end-tidal reading is quite high. Then we can generally infer that the PCO2 is also going to be high. So I think it has general practicality to continue to using those values, and it needs to continue to be a critical part to our protocols and the way that we ventilate patients in the absence of other information. It's also impractical to think that most ALS services are gonna be able to go out and get a point of care blood gas analysis assessment, nor that it is necessary in all these cases, particularly if we have shorter transports to a hospital. but I think that, we need to look at this more closely I think there clearly needs to be more research for us to look at the broader EMS population. And it may be that as we gain more information, we may be able to adjust not whether we use entitle or not, but what the optimal ranges will be based on how they tend to correlate with PCO2 values. And in fact, it may be that in the future we need to adjust the range that we're aiming for in specific patient populations recognizing that it isn't correlating exactly.

Mohamed (2):

Oh man, this is fascinating discussion. Chris, I know you addressed some of these, questions, but do you anticipate any next steps in future research in this topic?

Chris:

So I think we've fairly well answered this question for the critical transport environment. It certainly can be reproduced by others. I would strongly encourage that. With a data set of over 6,400, I think that we have gotten the information that we're able to get. The baton is next to those that can evaluate this in an ALS population. We may have some surrogate data from large data sets that we could look at, but next it is to really assess this in a broader patient population and whether the ability for us to target. Potentially alternate goals for end-tidal is the right answer, is a question that we need to try to figure out and if there are indeed subpopulations of the ALS oriented patient population that this is of greatest importance in, or other questions that we'll need to figure out.

Mohamed (2):

Yeah, so I guess stay tuned for more coming on this side of the world. I would tell you like just few hours ago I got a consult about a cardiac arrest and the crew was telling me that the end-tidal CO2 is in this patient specifically was minimal eight, and then it went up magically to 40, so they were confused about what to do next. What, just your final thoughts about cardiac arrest and the utility of end-tidal CO2 in cardiac arrest patients.

Chris:

Because we found these discrepancies between end-tidal and PCO2, I just can't emphasize enough that end-tidal, CO2 continues to be extremely important, continues to be very useful for us, and as of today, I wouldn't change necessarily any protocols that we're currently using. I think this certainly adds a lot of questions. I can tell you that for an environment where we have point of care testing, we have identified that it is very important for us to check that and to use it because we can, for the cardiac arrest population, I think that is one in particular where we're often trying to titrate that ventilation for a specific goal. If you happen to have a blood gas analysis, then use it. But we're gonna need to, that is a population that is similar to other critical care patients. And we just need to be thoughtful about the impact of perfusion on end-tidal. And in fact, whereas we might want to titrate to that goal of, 35 to 45. If the blood pressure suddenly tanks and the end-tidal matches you won't it's not necessarily correlating to PCO2 at that time. So focus on the rest of the resuscitation of that patient and try to get that blood pressure back up. And finally the last thing that we also looked at in our paper, and I think is of great value for those that are managing a patient with a longer transport time is to see if we can, make a judgment about that correlation initially and then compare it over time. Meaning if you see that, as you de you describe, if you have a patient that you first notice that there's a substantial gap between the end-tidal and the PCO2 if you're managing this patient for the next hour that you might have a longer transport for them in a facility or otherwise, can you just use that gap that you found initially to titrate based on end-tidal over time? And so we actually looked at a subset of patients where we had end, we had multiple end tidal and PCO2 readings within five minutes of each other to try to see if that gap was fairly consistent over time and it really wasn't in a lot of patients. And that makes sense because as you just described, in that post-arrest patient, when you're fixing somebody's profusion, when you're actually a, addressing the issues of VQ mismatch, if you have that respiratory patient that you're now ventilating better while you've had hands on the patient for some period of time, then that gap might narrow. Alternately, if that patient is getting sicker, that gap may widen. And it's really difficult to correlate that over time. But this, these are the areas where we need more information. We need more data to figure out how exactly do we do the, do we use these values? How often might we have to recheck a blood gas assessment for the inter-facility transport in which patient populations is this of greatest importance?

Mohamed (2):

It brings up the importance of also reassessment, right? Reassessing your patient and make sure that whatever you provide for an intervention actually works. So continue with reassessment as well. So this is a very important point too. Thank you so much, Chris. I appreciate that. Any last pointers you, man, this is. so many tips over here and

Chris:

No, this is. Yeah, this has been very exciting to be able to finally have this data set together to share with everybody. I hope that it opens up folks' minds for those EMS agencies that have that capability to, to have point of care assessments recognize that there is true value in taking a look at that and making sure that it's well integrated into your protocols. And more to come, I hope along this line. So thank you very much for having me on. Mohamed.

Mohamed (2):

I hope so too. When. Thank you for all of you for tuning in and learning from our amazing educators here at University of Pittsburgh. And thank you so much Chris, for discussing this paper. I look forward to the next episode, and please stay motivated, stay compassionate, and I hope to talk to you soon. Take care.