SCCM is performing maintenance on its websites. For the best browsing experience, please use Microsoft Edge or Safari. Those using Chrome or Firefox may experience access issues at this time.

SCCM Pod-507 PCCM: Pediatric Neurologic Challenges Unveiled

visual bubble
visual bubble

Host Maureen A. Madden, DNP, RN, CPNP-AC, CCRN, FCCM, FAAN, is joined by Dr. Lakshmi Raman, MD, to explore critical insights into pediatric ECMO, addressing neurological complications and their potential mitigation strategies and other topics associate with the article "Early Changes in Arterial Partial Pressure of Carbon Dioxide and Blood Pressure After Starting Extracorporeal Membrane Oxygenation in Children: Extracorporeal Life Support Organization Database Study of Neurologic Complications" (Pedtr Crit Care Med. July 2023; 24(7):541-550).  Dr. Raman is Professor of Pediatrics and Medical Director of ECMO at the University of Texas Southwestern Medical Center and Children's Medical Center in Dallas, Texas.

*If you are unable to play the podcast please click here to download the file.

Category: PCCM Podcast


Dr. Madden: Hello, and welcome to the Society of Critical Care Medicine Podcast. I’m your host, Maureen Madden. Today I’ll be speaking with Dr. Lakshmi Raman, MD, and we will be talking about the article, “Early Changes in Arterial Partial Pressure of Carbon Dioxide and Blood Pressure After Starting Extracorporeal Membrane Oxygenation in Children: Extracorporeal Life Support Organization Database Study of Neurologic Complications.” This was published in the July issue of Pediatric Critical Care Medicine. To access the full article, visit Dr. Raman is a professor of pediatrics at the University of Texas Southwestern Medical Center and is the medical director of ECMO at Children’s Medical Center in Dallas, Texas. Welcome, Dr. Raman. Before we start, do you have any disclosures to report?

Dr. Raman: Thank you, Maureen, for having me. I’m really excited to talk about this article, a topic that is very close to my heart. The only disclosure I have is I’m a coinvestigator on an NIH-funded grant that looks at neuromonitoring using optical instruments. And the second is I’m a principal investigator on an NIH grant that just got recently funded, working on a prediction model for neurological injuries on ECMO.

Dr. Madden: Congratulations on both of those. Those are wonderful. Great achievement. I can’t wait to see when all that work is done as well.

Dr. Raman: I can’t wait myself. Yes. I’m already worried about the deliverables. Thank you.

Dr. Madden: It’s clear that ECMO is something that you are very passionate about. First, I want to congratulate you on the article being published and being selected as an editor’s choice.

Dr. Raman: Thank you. Kudos to the first author, Neel Shah, who did a lot of the groundwork, and to the team of people who helped, including statistical support, looking at the interaction between CO2 and blood pressure from Dr. Venkataraman and all the ideas from Dr. Thiagarajan as well. It could not have been done without the effort of the team.

Dr. Madden: That’s wonderful. I appreciate you acknowledging all of them. To start, I want to get a little bit of background from you in terms of how you became so involved in ECMO and developed this passion to start doing research with it.

Dr. Raman: How I became involved with ECMO was a little bit of a chance. In 2008, I moved to UT Southwestern as a faculty and Children’s Medical Center of Dallas. It’s a high-volume ECMO center. In this program, since about late 2009, early 2010, they had a group of unique physicians who would be the ECMO team. Then in 2013, I was asked to be on the team and then two months later I suddenly realized I’m also the medical director of the program, a high-volume program, so now I had to really get my act together to know the ins and outs of the therapy and now this one. My research interest was always in neurocritical care. My initial work from my fellowship was bench research looking at chronic hypoxia in rats and mice. I looked at neurogenesis and all that and then realized I wasn’t going to do well in bench science and there was a lot of clinical time that I had to cover.

When I became the ECMO medical director, I said, how am I going to transition that interest? Then I said, these patients have neurological injury, why don’t I start focusing on that? So my initial five, seven years of work was mostly on neuromonitoring. We started doing cerebral oximetry. We made that the standard of care. I published some work related to that. Then the idea of, these are patients who have such a wealth of data, both from the clinical side and from the pump side, and I always felt like when I put a patient on ECMO, I quite didn’t know who would suddenly get neurological injury because it would be like the kid would be doing okay, then suddenly they’ll have a pupillary change and I would be like, what did we do wrong? I quite couldn’t place my finger on it.

Then somebody suggested to me that there is this machine learning. You can look at all of the data. So we actually published our single-center work. Again, a lot of the groundwork was done by Neel Shah, the first author for this paper, and Dr. Verhaert, who had done some MRI work, and we published that work in, I believe, 2020. Then we said, oh, maybe we will now look at the ELSO registry because it has a wealth of data, wealth of patients, though the amount of data is not as granular as you would be able to get in a single center.

Dr. Madden: Isn’t that the same? No? Every single time. Once you create this idea and you start to do it and you dig down into whatever data that you have and you realize, nobody read my mind and put the pieces in there that I actually want to look at.

Dr. Raman: Then, as we were thinking about this, the adult paper got published from Eddy Fan’s group where they showed that rapid changes in CO2 in adult VV ECMO was associated with increased neurological complications and we were like, okay, there we go. Before that, there have been already case series from single centers, periodic case reports. Personally, at the bedside, we already had experience and we started having an algorithm for our patients, so we said...

Dr. Madden: Before you go into that, I want to interrupt you, I want to go back a little bit more. Because when we think about this, and as you said, most people are very aware that ECMO is a therapy, the children who are able to successfully come off of it, many of them have sustained some neurologic injury. And people are looking at the long-term pieces of this in terms of their developmental components and their cognitive pieces. So, thinking about how the adults had published, looking at the PaCO2 and the change in regards to ECMO and how it clicked in your mind, tell me what you thought, how you would put all this together in terms of not just the PaCO2, but you also talked about the MAP and the relative changes of the MAP as well.

Dr. Raman: Yeah, exactly. When we got the data, our plan was certainly to look at the PaCO2, but I’m always intrigued because I feel like many times, as we know, these are some of the sickest patients in your ICU. They are acidemic, they are hypotensive. Then, when we put them on ECMO, everybody has a big sigh of relief. Oh, yes. We are weaning the vasopressors, the blood pressure is good and people walk out of the room. For me, I feel like now starts the clock, the ticking time bomb. Often we feel blood pressure happy because they, who are hypotensive patients, now we are able to come off vasopressors. Our blood pressures are generous.

Dr. Madden: To the point, as you said, the blood pressures are generous and we’ve normalized it or even gone beyond the threshold we wanted to achieve. In that respect, we want to correct everything. We love that they go on, and we have this therapy that they’re responding to and we can normalize items. As you said, these are the sickest of the sick children, so their acuity is very high. In terms of talking very specifically about your paper and the focus on the changes in the PaCO2 and the MAP and actually you created it that there is a relationship between both of those, so it’s not just one but it’s actually both. I’d love to hear some of the thought process about that.

Dr. Raman: We said, you know, the increase in MAP is also going to affect their neurological injury but then I didn’t know statistically how can we show that. Shalon was our statistician who was doing a lot of the work and Dr. Shah spent innumerable hours going over the data, normalizing the blood pressure and everything. Then Dr. Venkataraman, who actually happens to be my husband, we were on a walk one evening with our dog, and I was like, you know, I do think this blood pressure matters too but I just don’t know how to. He said, no, Lakshmi, we can. There are ways that we can show the interaction between the two variables after controlling for other things.

So that is how we said, okay, we are going to, and a lot of groundwork went into this because we had to normalize the patient’s blood pressures to their age because, unlike adults, we were looking at pediatric patients from 30 days of age to 17 years of age. So you can imagine the wide range but luckily for us, just before that, there was a paper published that showed what are the normal ranges for critically ill patients in the hospital. So we took that and then we had to extrapolate the data because if they had only mean blood pressure, only if they had systolic and diastolic, we had to come up with the mean blood pressure, things like that. So we had to do a lot of groundwork to create that.

Then we were able to group their MAPs and look them up as continuous variables. The PCO2 also has continuous variable, and then we divided them as categorical variable based on ARDS greater than 70, less than 70, and then showed interaction was very meaningful. And that, of course, obviously, it was only in VA patients because that is where you see the significant change in blood pressure within the first 24 hours where you dial up the pump.

Dr. Madden: Right. That’s a piece then I wanted to ask you about as well. You chose to pull out very specifically about individuals who were cannulated in the carotid and then the differential response to the PaCO2. So talk about some of those statistics because, when you looked at it, some of the premises didn’t seem to be as significant.

Dr. Raman: I think the carotid cannulation, the renal replacement therapy, the pre-ECLS arrest, there are various previous both single-center studies and ELSO registry studies as well that have shown that these are some risk factors associated with neurological injury. When the ELSO registry gives us the data, there’s just such extensive data, so we sifted through a lot of that data and found out which variables are of significance. Carotid artery cannulation, we are not the first ones to show in the ELSO registry.

There are a lot of other studies that have shown that carotid cannulation is a risk factor. Similarly, renal replacement therapy. We also looked at other ventilator factors, which did not turn out to be, for instance, inhaled nitric oxide, mean airway pressure, PEEP, and those didn’t pan out in our statistical modeling. So, yeah, I think cardiac cannulation, pre-ECLS arrest, renal replacement therapy were some important factors that showed to be of significance.

Dr. Madden: Your article demonstrated that the relative change in the PaCO2 and the relative change in the MAP and showed the discrete areas of that change for each of those elements that was statistically significant for what was to be less of a precipitating factor or an association with neurologic injury. But what I was interested in, and it wasn’t really discussed as fully, as I said, we know that 15.6% of children have a neurologic complication association with ECMO. What’s or how did you discuss the timing of that event when it occurred? We know the acuity of this patient, there’s minimal ability in terms of clinical signs to determine that something has happened, you know, because oftentimes they’re so sedated and neuromuscular blockade and all of those elements.

So I was interested to discuss, you’re talking about early, within the first 24 hours, careful adjustment of PaCO2 and MAP, and maybe not, as you said, normalizing them to age or normal PaCO2. But you’d looked at it in the first 24 hours and you felt that there was a significance to impact neurologic injury, meaning lessen that impact. But tell me, how do you correlate that to the evidence, if there is any, about when that injury occurred?

Dr. Raman: Yeah. I mean, those are very good points and unfortunately cannot be answered with the ELSO registry data. For one, remember we have one time point before ECLS and one time point after ECLS. And what happens, even within that 24 hours, is not granular in the data. So that is for a start. Second, you brought up a very important point, 15.6% of the neurological injury and when it happens. First of all, I think people who have done both single-center studies and other multicenter studies and where we have sought to look for neurological injury, we know that the neurological injury that’s reported in the ELSO registry is underreported. We believe that it is in the order of 30 to 40% actually. And when they happen, that I don’t think we know yet because, as we all know, these are some of the sickest patients.

Other than one adult study where they routinely got CT imaging as soon as they went on ECMO within six hours and said, okay, these are patients who already potentially had an injury even though we still don’t know if the six hours mattered. Unfortunately we don’t know how much of this injury is pre-ECMO, how much of this is peri-ECMO, and how much of this is within the first 24 hours, or is that a continuum? The reason to focus on this time period is multiple. For one, in the ELSO registry, we had only that data. Second, we do think that is the time of maximum changes. If you think about it, you take a patient who is hypoxic, acidemic, hypotensive, hypercarbic, put them on ECMO and we rapidly correct everything and we hold that place for the next several days until they can come off ECMO. So we do think that is the maximum time when things change rapidly and can potentially be the source of the problem.

Dr. Madden: It makes sense. We as practitioners have an understanding about cerebral autoregulation, and you’re talking about an approach to mitigate that disruption based on the acuity of this population. You have some elements that there’s other science out there. There are other papers that have been published to support focusing on that. And I think that we have a lot of room or research to do in terms of trying to optimize outcomes on this, not just as we’ve shifted our focus over time for, not just survival, but to actually thrive and have optimal outcomes.

So this is critically important, and I really appreciate that this was published and you’re continuing to work on this. As I was reading this, one of the points was discussing the ability to analyze the change in the PaCO2. You had to draw the gas, you had to come back, you had to draw the gas, you know, all of those changes. I’m envisioning about a continuous blood gas analysis that is somehow within the circuit. You’ve talked about now how you had started continuous neurologic monitoring. Tell me some of the other pieces that you have thought about or that you’ve implemented in your environment.

Dr. Raman: Actually, when you say continuous CO2, we do have the CDI that monitors where we have continuous CO2. Based on my clinical experience at the bedside, how we had these case reports and my own personal experience at the bedside that what we have adapted in our unit is generally to say that if a patient has a high PCO2 and low pH, we try to match our circuit to match the patient’s blood gas by bleeding in CO2 because you have to remember that ECMO, even if you don’t give the sweet gas, the oxygenator itself is so efficient in removing CO2. So you don’t have to have much sweet gas. Just the pure blood flow through the oxygenator is very efficient in removing CO2.

What we have tried as a lead practice here in our institution is we can match the CO2 and then, sometimes the patients come to us and they have to go on ECMO immediately, but if they have been in our unit, they have been hypercarbic, if we have a sense of how long they have been hypercarbic, then we at least take two/thirds of the time to come down on the PCO2. We tell the specialist, this is how we want the CO2 to be dropped at the bedside. If you start off at 100 and we want it to come down to 50 and that’s a difference of 50. And if they have been like that for 24 hours, we are going to take at least 18 hours or so to bring down the CO2 from 100 to 50. You do the math, and that’s how we are going to come down.

Since we have the CDI that continuously monitors, we are able to look at it and we’ll also draw frequent blood gases. But you can come back to me and ask the question, have you made a difference? And that is a study that still has to be done. Since we have adapted this practice, have we seen? I’m hoping that the multicenter study that I’m getting ready to start may be a venue to maybe look at these things and answer these questions as well.

Dr. Madden: We’ve talked about the 15.6% with any neurologic complication, and your data showed, with the change in the PCO2 being minimal, that they still had approximately a 13.9% neurologic injury. So I’m going to comment that, percentagewise, it doesn’t seem to have made that much of a change. That’s me making a statement without a statistical analysis of it. But you were talking about this process then and how we need to continually improve and the discussion you had about how you set out with your specialist about how to change the PaCO2 over time. Is that the algorithm that you were talking about that you’re developing or have in place?

Dr. Raman: We have in place but, as I said, is that proof? But also remember, CO2 is not the only factor, and that is the impetus for my multicenter study as well. We did again an algorithm to find who are the patients at risk for neurological injury and that data is being currently reviewed. And it is not just the CO2, there are multiple other things. In our own paper, we showed not only CO2 but also blood pressure. But there are also other things that definitely impact renal replacement therapy. There are multiple other factors. Just because I have an algorithm for CO2, am I going to save all these kids from neurological injury? Certainly not. But at least this is something that we can potentially control. Because we said carotid cannulation is a very important factor. Can I not cannulate the carotid artery? I don’t think I have that luxury, especially for really small infants and children.

Dr. Madden: There are definitely parameters that we can’t change. You have to work with what you’re given and, as you said, all these other therapies come with benefits and risks.

Dr. Raman: Yes. I think those are some of the challenges, but at least I think one of the things that I am striving toward, and I think the entire ELSO community and ECMO community is striving toward, is, one, are we able to identify patients who are at risk, is the first question. Then, once we identify, are there modifiable risk factors that we have control of? Then, we do know we are going to have a set of nonmodifiable risk factors, so then what can I do?

Then the other thing is, can I intensify the neuromonitoring so that I do identify the injuries early on when it happens so that then I can potentially give some neuroprotective stratum therapies, i.e., keeping the head of the bed elevated, keeping the sodiums higher, maybe if they are already at risk for seizures, seizure prophylaxis?

Then, especially in the adult world, as soon as a bleed happens, they’ve been good about doing even decompression, potentially for strokes, thrombolysis. I think those are all still in very small case series and this has not become the standard of care. But I think it opens up a whole potential for thought process intervention. How do we take care of these children?

Dr. Madden: Right. Along those lines then, thinking about this, moving any individual who’s on ECMO is really a huge task. But at the same time, is there a component that we have to put in some standards or some thought process about routinely having other modalities in place such as, do they need an MRI, do they need a CT scan? Many of them go routinely on EEG.

Dr. Raman: Yeah, I think that is maybe because of my passion, you’ll see in my unit, we have intense neuromonitoring. Everybody, every bed space, has cerebral oximetry. We put everybody, as soon as they go on ECMO, within six hours we hook them up to EEG. As I said, we have another NIH study, but that’s a consented study, so we have to get parents’ consent, where we are doing optical measurements. One of the things we are working on with our cerebral oximetry is, because we have continuous bedside etiometry, which shows waveforms, can we come up with noninvasive cerebral autoregulation curves that we can come up with?

That’s certainly something that we are thinking about, but we don’t have it yet. There are many centers that have portable CTs, and there are now portable MRIs. In fact, we are working with our hospital administration to see if this is something that we can purchase. There are some case series using portable MRIs in pediatrics and in neonates and cancers in adults from Johns Hopkins. I think they are considered to be safe with ECMO machines, but it certainly takes a village if we have to transport them. In fact, in our NIH study, the consented study, we do take them for CT within 48 hours. So, yes.

Dr. Madden: That earlier identification.

Dr. Raman: Earlier identification, correct, of injuries.

Dr. Madden: Yeah. This is such a field that is wide open still to so many different research aspects and intervention aspects and, as you said, we’re trying to improve the neurologic outcomes, to have these individuals survive their illness and actually have good outcomes and thrive, not just survive ICU.

Dr. Raman: Exactly. Not just survive. That is why I think it has become important: one, to say, are there modifiable risk factors? If so, can how can I modify? Second, I know there are no modifiable risk factors. Can I identify the injury early and do some neuroprotective strategies? And lastly, can I at least recognize that they do have injury and put them early on a path to therapies so that we can improve their long-term outcome? As you say, we not only want survival to ECMO and survival to hospital, we want them to be integrated into society as normal as they can be.

It’s a big challenge but I think luckily ECMO is, as once somebody said, that we are in the phase of cancer treatment. In those days, if you survived cancer, that was considered a big deal. But today, cancer therapy is such that they are seeing, what is the long term? Are they able to have children? Are they able to go back to being in medical school? Do they have kids? I think ECMO is getting there, and that’s what we should strive for.

Dr. Madden: Right. We have really shifted, as providers, our focus, not just to good outcomes, as we said, but I’m hearing you say that if we can find the preventative strategies and have them in place rather than reacting to, unfortunately, something that is a risk that we knew going in and now it’s happened. But how can we mitigate that risk if it’s modifiable? But also some others, can we put preventative strategies in there because they weren’t an anticipated risk, like seizures? We can always talk about that. But I want to just also bring up the fact that if individuals weren’t aware, there was an editorial that accompanied this article as well. Dr., I apologize if I don’t say it properly, Anne-Marie Guerguerian?

Dr. Raman: Guerguerian, yes.

Dr. Madden: Yes. I really think that if other individuals haven’t seen this, it’s so complimentary to your research in terms of actually breaking it down to the cerebral autoregulation. I really feel that it’s so wonderful in supporting what you were studying that I would encourage other people to look at it, but I wanted to get your take on this as well, if you had anything you would like to say.

Dr. Raman: Yeah. Anne-Marie is a good colleague of mine and a great researcher. She has done a lot of good work in the cardiac arrest population and looking at long-term outcomes. And, yes, she was very complimentary and she said that is such important work. And, yes, I think it is so interesting. Cerebral autoregulation was something that, in the 90s they did LAM studies and had sham LAMs and LAMs on ECMO and looked at how they had autoregulation disturbances, particularly on VA ECMO. I think, for the next 20 years there hasn’t been much on that but now in the last five years there are multiple people working on this.

We have done some work using, as I said, cerebral oximetry. Now we just published our pilot study and we now have this NIH study using optical spectroscopy and I know the adult counterpart in upstate New York, they also got recently funded to do the same in adult patients. Then the group in Paris, others are also working on looking at these several autoregulation disturbances. And, you know, that’s always, is it the chicken or the egg, you know, do the autoregulation disturbances precede the injury, or is it because they are so sick they have autoregulation disturbances that then put them at risk for neurological injury, or is it because they are so sick they already have some neurological injury that we have not appreciated both by imaging and/or clinically, which leads to cerebral autoregulation disturbances?

I think, either way, at least if we have modalities of identifying, then it gives us room for intervention, you know, can we identify what is the optimal blood pressure where their autoregulation disturbance is the least? In fact, one of my fellows is doing a preliminary project on that and she got ELSO funding. She is looking at the same time period that we are talking about in this paper, the pericannulation period and looking at autoregulation disturbances using cerebral oximetry.

Dr. Madden: There’s so much research that’s coming out. It’s going to be wonderful to really be able to see how it impacts the care of our children. I hate to say that our time is coming to a close. But before we do that...

Dr. Raman: We have been talking about this topic for for days on end. This is very close to my heart.

Dr. Madden: That’s very clear, and I love that you have this passion for it because it we need it. Before we conclude, I just wanted to see if there was anything that we hadn’t covered maybe that you wanted to make sure the listeners heard.

Dr. Raman: Yes. For all the young people out there and the trainees out there who are interested, I think this is such an area for research and I would say there are lots of good people across the globe doing work and please, please take this upon you to do this work.

Dr. Madden: It’s the mentorship and the bringing up all of our individuals as they grow. I love that you are there with that encouragement. I want to thank you again for this article, you and your team. It really is such a beautiful study, and I appreciate the time and energy that went into doing it. I thank you for that. At this point, I have to say this concludes another episode of the Society of Critical Care Medicine Podcast. If you’re listening on your favorite podcast app and you liked what you heard, consider rating and leaving a review. For the Society of Critical Care Medicine Podcast, I’m Maureen Madden.

Announcer: Maureen A. Madden, DNP, RN, CPNC-AC, CCRN, FCCM, is a professor of pediatrics at Rutgers Robert Wood Johnson Medical School and a pediatric critical care nurse practitioner in the pediatric intensive care unit at Bristol Myers Squibb Children’s Hospital in New Brunswick, New Jersey.

Join or renew your membership with SCCM, the only multiprofessional society dedicated exclusively to the advancement of critical care. Contact a customer service representative at +1 847 827-6888 or visit for more information. The SCCM Podcast is the copyrighted material of the Society of Critical Care Medicine, and all rights are reserved. Find more episodes at

This podcast is for educational purposes only. The material presented is intended to represent an approach, view, statement, or opinion of the presenter that may be helpful to others. The views and opinions expressed herein are those of the presenters and do not necessarily reflect the opinions or views of SCCM. SCCM does not recommend or endorse any specific test, physician, product, procedure, opinion, or other information that may be mentioned.

Some episodes of the SCCM Podcast include a transcript of the episode’s audio. Although the transcription is largely accurate, in some cases it is incomplete or inaccurate due to inaudible passages or transcription errors and should not be treated as an authoritative record.



Knowledge Area: Pediatrics Neuroscience Procedures