#415 - Finding Optimal PEEP at the Bedside With Electrical Impedance Tomography?

In this episode, we sit down with Dr. Jessica Shui, attending neonatologist at Mass General for Children, to explore the game-changing potential of Electrical Impedance Tomography (EIT) in the NICU. We dive into her recent paper in the Journal of Perinatology on using non-invasive EIT to identify optimal PEEP in infants with severe bronchopulmonary dysplasia. Dr. Shui explains how this real-time, radiation-free technology allows clinicians to visualize lung mechanics, dynamically titrate ventilator settings, and confidently reduce PEEP without risking atelectasis. Join us as we discuss moving beyond blind adjustments and stepping into the future of personalized neonatal respiratory care.

Transcript

Ben Courchia MD (00:00.6) Hello everybody, welcome back to the incubator podcast. We're back today for another interview and we are joined today by Dr. Jessica Shui. Jessica, welcome to the podcast.

Jessica Shui, MD (00:09.136) Thank you. Thank you so much for having me. I've been really looking forward to this. Big fan of your show.

Ben Courchia MD (00:13.582) We've been working on scheduling this interview for some time and we have been talking about this, I believe it was at one of the PAS meetings. And you had the release of a recent article that came out in the Journal of Perinatology on the topic that we're discussing today, electrical impedance tomography. So we wanted to coincide with the timeline, and this brought us here today.

Jessica Shui, MD (00:24.076) Yes, exactly.

Ben Courchia MD (00:42.838) Jessica, you're an attending neonatologist in the NICU at Mass General for Children. You trained in the Northeast Tri-State area. You got your medical degree from New York Medical College, did your residency at Connecticut Children's, and then did your fellowship in neonatal and perinatal medicine at New York Presbyterian Columbia University Medical Center. Currently, your research is taking place in the Lorenzo Berra Laboratory. You're looking at respiratory strategies for neonatal lungs, and you are the founder and the leader of the Encore team, a consultative service using electrical impedance tomography that is currently one of, maybe the only one in the US doing that. I don't know if at the time this bio was written, maybe you were the only one. Are you still the only one? We will talk about that. You may not know. Maybe some people are working on this at this moment. Very happy today to talk to you about your work on electrical impedance tomography. I have reviewed a recent publication that you authored in the Journal of Perinatology called Identifying Optimal Positive End-Expiratory Pressure with Electrical Impedance Tomography Guidance in Severe Bronchopulmonary Dysplasia. The first question I would like to ask on behalf of the audience is, what is EIT? What is electrical impedance tomography? EIT for dummies, if you will.

Jessica Shui, MD (02:12.272) I know it's a mouthful, but it's really fascinating technology that I'm very excited to use. It is a real-time lung physiology monitoring tool, and it's fantastic for our population because it does not involve any radiation. It is non-invasive, and it can be used right at the bedside. It doesn't require any in-hospital transportation like we need to do with other imaging for our patients. The way it works is there is a band of electrodes wrapped around the chest.

Jessica Shui, MD (03:06.458) Different companies make it, but it can range from 16 to 32 electrodes. It injects imperceivable, very small amounts of electricity through the thorax, and it gets picked up by neighboring electrodes. Because the impedance—or you can also think of it as electrical resistivity, it's related—as it goes through tissue, fat, and muscle is different than when it goes through air, which is usually in the lungs. The computer is able to use its algorithm to see where air is located in the chest. It captures around 50 images per second. So you really get breath-by-breath feedback on where air is entering the lungs.

Ben Courchia MD (04:13.388) This is really neat. I feel like with these small electrodes, maybe in the early days, you automatically think, is it like an EEG? The technology has evolved now so that it's more like a little belt that you can adjust, just strap on the baby's chest, kind of like those heart rate monitors for people who run. I think it looks quite similar in the way it's done.

Ben Courchia MD (04:39.284) It's just this little belt that you put on the baby's chest that then gives you the measurements you need. And the output of this is you get on a screen, basically a round shape, which represents the chest. In it, you can see the lungs and you can see where the areas of the lungs are well inflated versus not so well inflated. Is that correct?

Jessica Shui, MD (05:09.418) Yes, the orientation is kind of like a CT scan. You see it in a cross-section of the thorax and it takes an average of around where the belt is. This technology started in the adult world. For adults, this is considered regional measurements of ventilation because it ends up looking like a slice of the chest.

Jessica Shui, MD (05:36.9) What's really great for our population is because the belt takes up so much real estate on these small chests, it really ends up getting most of the lungs. Maybe you miss the very apex or the very base of the lungs, but it's quite representative, more so than in the adult or older children population. I think that's what's special about using this tool for babies.

Ben Courchia MD (06:01.602) This is very interesting. You mentioned that it's an electrical signal, so there's no radiation involved. I think that's always a concern with so many X-rays and imaging. This is completely radiation-free.

Jessica Shui, MD (06:23.424) Exactly. There are different kinds of EIT, but all of them are non-invasive. Either it uses a hydrogel similar to the EKG leads we use clinically, or other companies make it with a fabric-lined belt with a contact gel. I really appreciate that there are different considerations for our preemies and the fragile, immature skin in our population to make sure it's tolerated by them.

Ben Courchia MD (07:09.736) We feel that as neonatologists, we're very good at neonatal respiratory management because that's the bulk of our work. The tools we use currently to get information on the baby's degree of lung inflation and respiratory mechanics really involve two things. You have your traditional chest X-ray and your ventilator. The ventilator provides you with a lot of information as well, those flow-volume loops and all these different measurements. But can you tell us a little bit about what kind of information you are practically gathering at the bedside? I will refer people to your paper as well. Obviously, if you do a simple Google search, you will find more images. The way the information is provided to you as a clinician is very interesting. You have it divided into quadrants and you can really see the degree of inflation. Can you walk us through what you're looking at at the bedside? Is it qualitative data that you're gathering or is it truly quantitative data that you then trend? Can you walk us through the data acquisition you're doing at the bedside when the EIT is on the baby?

Jessica Shui, MD (08:30.818) That's a great question. It really should be used as a trending tool for each baby because, for each patient, the body mass is different and it affects the measurements. You can't compare absolute values between patients. With amounts of edema, there are just different factors that come into play. Keep in mind that these are arbitrary units. If you use it as a trending tool—for example, changing different ventilator settings to see a change in inflation or new development of collapse—that's where it's useful for that specific baby. My goal for my research is to show the utility of giving personalized, individualized care as our approach to respiratory care. The other key thing is patients don't need to be on specific respiratory support to use this tool. It could be invasive mechanical ventilation, non-invasive ventilation, or they could just be on room air, the oscillator, or ECMO. It's a really great physiologic monitoring tool.

Ben Courchia MD (09:58.498) You've been familiar with the technology and the tools for some time now. You've used it in the context of research. One of the reasons we're doing this interview this year is because I believe these devices are actually going to reach the market this year. We'll have separate episodes about that down the road when that happens. It's great to talk to someone who has had some mileage with the devices and has knowledge and experience to share with the community. In the paper you published recently, you mentioned that you tend to do these studies for about one to two hours. Is that the amount of time you leave it on to acquire a certain amount of information? I think we have some experience with devices that are similar. I would think like transcutaneous CO2 monitoring, for example, where we leave them for specific durations of time. How long do you leave them on until you say, 'We have enough data'? Can I potentially recycle the belt to another patient, or give the baby a break and come back later?

Jessica Shui, MD (11:21.968) It really depends on your clinical question and what you're looking for. Depending on the type of belt, they could be worn for 24 to 72 hours at a time. They are single-patient use.

Jessica Shui, MD (11:46.212) In our paper, where we were looking at identifying optimal PEEP in our babies, we kept it on longer to look for any evidence of late-onset atelectasis that may happen in this population because of their challenges with lung compliance.

Jessica Shui, MD (12:15.364) Certainly, if you have infants who have more homogeneous lung disease, like RDS, it just depends on the question. You may not have to keep it on for a couple of hours because you get real-time data. If you change different respiratory settings, you can see it happen instantaneously.

Ben Courchia MD (12:39.854) So you could keep it on the baby for 15 minutes and get plenty of data.

Jessica Shui, MD (12:42.936) Yes. You just want to make sure you have a steady signal. If the baby is crying or fussy, it's hard to make judgments on that. Sometimes it takes time to settle a baby. When we take segments of what we feel is representative data to make comparisons, we take two-minute stretches of data to compare.

Ben Courchia MD (12:52.694) Of course.

Jessica Shui, MD (13:11.856) So it could be as little as five minutes. It depends on your question.

Ben Courchia MD (13:11.906) Very cool.

Ben Courchia MD (13:19.404) I want to get into the practical applications. Obviously, this was one of the main points of the paper you recently published. One of the painful processes with babies with evolving chronic lung disease, bronchopulmonary dysplasia, is that they're often on a high degree of PEEP, and everybody's terrified of weaning that PEEP. You get a patient handed off to you, and they say this patient is on a PEEP of eight.

And they've been on it for a week and a half. You wonder, am I going to just leave it for my time on service, or am I going to be the brave one who does it? The real issue is that we don't know how these infants react to changes in their mean airway pressure, specifically when we talk about positive end-expiratory pressure. In the past, guiding this type of decision with data meant doing dynamic studies at the bedside with bronchoscopy to actually look at airway collapse as we were titrating the PEEP up and down, which is obviously very invasive and stressful to the baby. But this is really an avenue where EIT could be a game-changer for us. It gives us information about what effects we're seeing when we titrate the PEEP up or down.

Jessica Shui, MD (14:39.02) Exactly. The benefits are that it can be done at the bedside. In some hospitals, you would have to transport the patient to have dynamic airway bronchoscopy done. We eliminate that because these babies are very fragile, and transportation can be dangerous. They don't need any sedation.

Jessica Shui, MD (15:07.024) You can see what their physiology is in their usual state with spontaneous breathing. The invasive nature of a bronchoscopy can cause atelectasis because the actual scope can obstruct 50% of their airway and affect your measurements. Because this is non-invasive, you can see how they are in their usual state. Whether it's their awake state, their sleeping state, or during feeds, you can just keep the monitor on and watch for a longer period than you can with a bronchoscopy. You can look at different positioning; sometimes gravity positioning can affect recruitment. Or if you want to look at times of giving a bronchodilator to see if that changes things. There are so many clinical questions you can assess to see how your interventions help or may not help the baby, and you can see that in real-time without the invasive nature. I think that's what's really exciting.

Ben Courchia MD (16:40.334) Your data showed me that even those of us who feel careful enough and capable of titrating PEEP well, aren't brave enough to make the changes that positively reflect on lung mechanics. You have these beautiful figures in your paper where...you show the titration process of the positive end-expiratory pressure and the effects both on the electrical impedance tomography, but also on hyper-distension and compliance. You have two cases that struck me. In one, you begin with a PEEP of 14, which is quite high to begin with, and try to find the optimal PEEP. If I were a colleague in the unit with a patient on a PEEP of 14, and I felt I could wean it, I would probably try a PEEP of 12. That would be the brave soul who goes down on the PEEP by two. But your paper shows that with the information gathered from EIT, the optimal PEEP is 10. I would have never gone to 10 without the data to make me feel better because it's a massive jump. It's the same thing for another patient you describe where...

Jessica Shui, MD (17:49.199) Yes.

Ben Courchia MD (18:09.398) ...you start with a PEEP of 12 and have to go all the way down to eight. When you see the images, that makes absolute sense because you can see the tomography changing, and you see the hyper-distension going away without causing collapse. Without that tool, I don't think many people would turn it down by four centimeters of water, even gradually. This is not a change I would have done in a single day of weaning a PEEP on a chronic lung disease baby. Do you find people are surprised when you find that optimal point?

Jessica Shui, MD (18:52.464) Absolutely. We're so limited by the tools available at the bedside. A common tool used is X-ray, and that's just a snapshot in time with an arbitrary breath. You could see some evidence of over-distension there, but it can be scary to make changes of that magnitude because there's a historical recollection of how sick the baby was and why the team escalated up to that high in the first place.

Jessica Shui, MD (19:18.58) It makes the physicians, respiratory therapists, and nurses nervous to try to make a change because you don't want to 'break' the baby. But this is so empowering to be able to see at the bedside specifically for this baby instead of choosing arbitrarily. Maybe we're ready to make a wean today as a trial-and-error approach, but you can actually see in real-time at what point you start to lose lung volume and have collapse. You could just go back up right away. Usually, you make a change, come back a half hour later, see how the baby's breathing, look for more work of breathing, get an X-ray, get a gas. By then, the baby could have been de-recruiting, and you end up losing more ground than you gain. Here, because you see it in real-time, you can make the change, go back up to where you last minimized the over-distension without inducing collapse, and try to find that sweet spot. You can keep the monitor on to make sure that still is the right setting for the baby, watch for late-onset atelectasis, and adjust again. Perhaps even quicker than the next care time because the data is always displayed on the screen.

Ben Courchia MD (21:03.118) Jessica, what have you found the response from your colleagues to be? I feel like it's one of these things where it's seen as another gadget. I'm sure as soon as somebody sees the data points, they'll be calling you back multiple times for every patient on that team's census.

Jessica Shui, MD (21:20.002) I really appreciate our team. We've had such an outstanding team open to introducing this technology because it is so new and different. We are using it in our most fragile babies, our sickest babies, because this is the population that we think would benefit the most. We don't necessarily need to use EIT to measure a typical RDS baby weaning a PEEP from CPAP from six to five centimeters of water.

Ben Courchia MD (21:52.782) Yeah, a 34-week-old on Bubble CPAP 6 is not the patient that you get called for.

Jessica Shui, MD (21:58.978) Right, we're using this in our sickest babies, the babies with congenital diaphragmatic hernia, the babies with severe BPD. We have a multidisciplinary team for the buy-in. We have champions in our unit from the NICU nurses, the respiratory therapists, as well as physicians, working together to ensure we're all on the same page. The clinical team on service agrees that these are safe adjustments to make, and we're all watching together and having discussions because physiology is dynamic and always changing. Just because in one assessment we see a PEEP of eight is optimal, that can change if there's an aspiration event or abdominal distension. The team knows they can use their clinical judgment if things change.

Ben Courchia MD (23:05.259) In the field of extubation medicine, an important criterion is: after how many days can I say that my extubation was successful? People argue three to seven days. In your paper, the changes you made when it came to the PEEP were sustained, and that is one of the main conclusions for weeks. It's not like you got the baby well that day, and then the next morning the baby is 'broken'.

Jessica Shui, MD (23:11.354) Right.

Ben Courchia MD (23:35.118) These changes were sustained for significant periods of time. Obviously, the clinical status of a baby can evolve quickly for various reasons. Every time those changes happen, it resets everything, and you have to reassess from the holistic standpoint of every system.

Jessica Shui, MD (23:57.614) To get back to your question earlier, I think a lot of us were surprised with how low we were able to go. We thought we understood the baby; we had been taking care of the baby for months, and it seemed the baby did better with higher amounts of PEEP. Then to see that we could go lower, and to see how long it lasts over a couple of weeks or a month on this lower setting. It's really changed our approach, and seeing these success stories helped with buy-in in our unit to see how this device can guide us.

Ben Courchia MD (24:46.446) That was one of the most important lessons from Dr. Ben Kalari at the University of Miami when I was a fellow. Initially, we would sign out patients and say, 'Patient requires a PEEP of 14.' And he would say, 'Does he require it, or do you just have him on a PEEP of 14?' These are two very different things. A lot of times patients are on certain vent settings, but do they actually require that amount of support? You don't really know until you try to take them off.

Jessica Shui, MD (25:01.188) No.

Ben Courchia MD (25:13.858) But how do you do that safely? This is where EIT has such an interesting promise. When we get X-rays in centers where we don't have this technology yet, we sometimes see a baby has a bit of localized atelectasis, like a right upper lobe collapse. We say, 'Let's try to put the baby a little bit right-side up to see if we can get that air to go,' assuming all the other mechanical interfaces are properly set. If you do this on EIT, have you tried to see if the ventilation actually changes with these positional changes of the baby?

Jessica Shui, MD (25:55.6) It's actually exciting; you can see the gravity-dependent changes and the recruitment in real-time.

Ben Courchia MD (26:16.494) Because we know that physiology makes sense, but to actually see it on the screen is vindicating. Do you think that in the future, we might actually be able to reduce the amount of exposure to radiation from X-rays if we use more information from impedance tomography?

Jessica Shui, MD (26:22.382) Yes, it's really satisfying in that way. I think that would be a great goal if this could be like another vital sign for us. Just as we have EKG leads and SpO2 probes, we could have the EIT monitor on. You'd be able to see malposition of the endotracheal tube—that you're right mainstemming and only ventilating one side—so you can adjust and not have to keep repeating X-rays. Depending on the neck positioning, you can just adjust it in real-time.

Ben Courchia MD (27:23.16) How would you see this on the EIT? From the images I've seen, you see the two outlines of the lungs in a cross-sectional view. Would you infer the position of the ETT based on the asymmetry between the ventilation of either lung?

Jessica Shui, MD (27:42.512) On EIT, you can see what's called the distribution of ventilation. It looks at the percentage of the air on the right and left halves of the chest. It's not going to distinguish the right and left lung if you have a mediastinal shift; it's just dividing the thorax into the right and left halves.

Ben Courchia MD (27:55.181) I see.

Jessica Shui, MD (28:11.472) You also have to pay attention to the clinical context because this is measuring air. If you have a tension pneumothorax, it's going to show there's more air on that side of the chest, but it doesn't mean the baby is more recruited on that side. It's something to pay attention to for kids who perhaps have unequal distribution of exogenous surfactant and different compliances.

Ben Courchia MD (28:22.828) I see.

Jessica Shui, MD (28:39.652) This is a potential tool to help pay attention.

Ben Courchia MD (28:45.816) That's very interesting. So it cannot ultimately substitute itself for a radiograph. But maybe like we were saying earlier with transcutaneous CO2 monitors, because of their trending ability, it might help you reduce exposure. You can just trend them and use spot checks for the gold standard, like a blood gas or a serum bilirubin level.

Jessica Shui, MD (29:23.022) It's not going to localize anatomically, like a specific right upper lobe collapse, because you're looking at an average of the chest in a cross-section, or for smaller babies, a more global representation. You'll see decreased aeration on the right side, but it would be a different plane compared to an X-ray.

Ben Courchia MD (29:33.591) Mm-hmm.

Jessica Shui, MD (29:52.346) You get different information. It's complementary to what we have; we're not expecting it to replace X-rays.

Ben Courchia MD (29:54.889) For sure. Jessica, how did you find yourself working on this?

Jessica Shui, MD (30:11.568) I started in the basic science lab, doing translational research looking at lung cells and their development. I was actually on service and we had one of the babies from the case report with severe BPD, and we were struggling with figuring out the right setting. We threw the whole kitchen sink at him trying to improve his oxygenation. I am so thankful for our respiratory therapists—so shout out to Carolyn LaVita. It's too bad that we work in silos in our hospitals, but the respiratory therapists are the ones who go to all the units: the NICU, the PICU, the adult units, and the operating rooms. They saw on the adult side that there is this technology called electrical impedance tomography. During COVID, when they were trying to find optimal ventilator settings and see if proning patients helped with recruitment, they were able to see that in real-time with EIT.

Ben Courchia MD (31:13.922) Wow.

Jessica Shui, MD (31:33.8) They do PEEP titrations, particularly in obese patients because it's hard to see chest wall movement to identify the optimal PEEP. My mentor, Lorenzo Berra, has been using EIT clinically at Mass General for 11 years now. They told us about this non-invasive technology and wanted to see if it would help our patient. It has just been an elevator ride away.

Ben Courchia MD (32:11.33) Hahaha. You had this on the sixth floor?

Jessica Shui, MD (32:28.624) We tried it together with the specialists who have been using EIT so that we put all our expertise together. Seeing it work in real-time, and seeing how impactful it was clinically for this baby to improve on a lower PEEP, was amazing. The FiO2 dropped from 90 to 100% down to 40% because we dropped the PEEP from 12 to 8 in just a two-hour session. Like you mentioned earlier, it would take many weeks for someone to be brave enough to keep chipping away at that setting, if at all.

Ben Courchia MD (33:14.464) Any sign of desat would send the troops running to go right back.

Jessica Shui, MD (33:20.932) Exactly. But because we felt so empowered at the bedside, and everybody was on the same page to see the measurements, we made these moves coming down without causing more collapse. We actually improved lung compliance. That excited me so much, and I felt like this is what I need to do once I saw how impactful it was. I made a big career research pivot from the lab to clinical research to increase awareness about this technology and find applications for our patients. We adapted our protocol for our babies because they are not just smaller adults; they have their own protocol on how to do PEEP titration and what is considered optimal PEEP. This is just the beginning, and it's really exciting.

Ben Courchia MD (34:26.91) To give credit, Carolyn LaVita is the second author on the paper you recently published. My last question for you is, what's coming next from you and your team? The paper that you released in the Journal of Perinatology was a case series looking at specific cases where EIT was used for optimal PEEP titration. What can people look forward to coming out of your lab?

Jessica Shui, MD (34:57.168) It's been absolutely fascinating to see the different changes in lung physiology. This technology has been more popularized in Europe and Australia; it just hasn't reached the US until recently.

Ben Courchia MD (35:18.947) Mm-hmm.

Jessica Shui, MD (35:25.668)There is a lot of literature from institutions in those places about the different physiologic changes we can see in infants. For us, we want to go beyond looking at respiratory changes and see if we can use EIT to measure pulmonary perfusion changes as well, to measure ventilation-perfusion mismatch.

Ben Courchia MD (35:54.904)VQ mismatch. Beautiful. Well, good luck with that. We will eagerly be scouring PubMed for your next publication. Dr. Jessica, thank you so much for taking the time to talk to us today about your work and good luck with everything else related to pulmonary research.

Jessica Shui, MD (36:18.01) Thank you so much. This has been a lot of fun. Thanks for having me.

Ben Courchia MD (36:20.246)Yeah, same.