Dr. Raghu R. Seethala - Lessons learned while caring for COVID-19 patients on ECMO

All right, good morning everyone. We are very excited to have one of Brigham's own coming over for a special Grand Round session today. This is Dr. Raguse Thala. He hails from Stony Brook, New York. He received his bachelor's in science at Penn State, followed by his MD at Jefferson College. He then completed his emergency medicine residency at Mount Sinai, did additional research fellowship in resuscitation science at Penn. And that was followed by fellowship in anesthesia critical care at the Brigham. And that is where he is currently. He serves as the assistant professor of emergency medicine at Harvard Medical School. He's the section chief of critical care at the Brigham. He co-directs the cardiac surgical ICU as well as directs the thoracic surgical ICU there. And he serves as the associate medical director of Boston Med Flight. Particularly of interest today is that he is the medical director of the ECMO service at the Brigham. And in this role, he has been instrumental in developing what has become a pretty innovative and impactful COVID ECMO experience over at the Brigham. So in response to the COVID-19 pandemic, he developed an ECMO ECLS disaster plan that has helped shape the way COVID and ECMO has developed over the course of this two-year timeframe since the pandemic started in the New England area. And this has also involved a new ECMO transport program that has extended the capabilities across the New England region. So Dr. C. Val, thank you so much for all the impact that you have had in this role. And we are looking very forward to your talk today. Thank you. Sounds a lot better when he says it. So I'm invited here by you guys to talk about our experience that we had during the past two years. Here are my disposers. Here are the objectives. So I became the ECMO medical director in January of 2020. I thought it was a good idea to take this position, had no idea what was going to happen over the next two years. And it's been a roller coaster and the past two years of ECMO experience really vastly surpasses the prior 10 years of ECMO experience that I had. And we really saw an increase in capacity utilization, different modalities, cannulation strategies, collaboration across the New England region, internationally, and really I think has made our ECMO program much stronger and all ECMO programs, I think are much stronger as a result of this pandemic. So I kind of want to just go through some of the things that we learned. And we had Dr. Ferouk DeMerry, who is a fully trained pediatric surgeon, but he was my ICU fellow for a couple of months. And we did some cool things while we're taking care of these patients as well. All right. So when we first started hearing about the pandemic, end of December 2019, early January 2020, it seemed like this wasn't going to be a big issue here in the States and that it was going to be isolated mostly to Asia, particularly China. And the first reports that were coming out regarding ECMO utilization for severe COVID that dismal survival rates, they were zero or one out of six. So at the very beginning of the pandemic, we did not think that ECMO was going to be a big player in modalities of treatment. And lo and behold, what did we know? And we saw that in 2020, there was a huge uptake in ECMO utilization around the world. So this is data from the ELSA registry. You know, also is the largest extra corporal life support organization where many centers submit their data to them. And we see that ECMO utilization for respiratory failure took a big jump in 2009 as a result of the H1N1 pandemic. And continued to rise. And then we had another big jump in 2020 because of the COVID-19 pandemic. And the first wave outcomes were actually very good. They were comparable to non-COVID-AIDS ECMO. Survival rates at 90 days, over 60%. So with that, many centers started utilizing ECMO for caring for these really sick COVID patients. Centers that didn't have capabilities of ECMO were becoming ECMO centers that had ECMO capabilities were vastly expanding their capabilities to care for these patients. And to date, so this is the snapshot from the ELSA COVID registry from yesterday. There have been a total of over 16,000 reported cases too and also, so that means that there are more cases since the start of the pandemic for treating COVID with ECMO. And surprisingly, in North America, there still seems to be 131 patients still on ECMO as a result of severe COVID. So that brings us now to who are the patients that we are placing on ECMO. And this talk is largely about respiratory failure ECMO because that was predominantly what we saw in the adult world. Just a smattering of COVID-Myocarditis, but we actually did not see any over the two years at Brigham. And I think MGH only saw one or two severe COVID-Myocarditis requiring extra-corpo life support. So who are the patients that we're putting on? So when we first started coming up with our criteria, we decided that we would treat this as any ARDS. So we stuck with the Eolia criteria, which is kind of the standard ARDS criteria for adult respiratory failure. And this comes from the landmark trial done several years ago, published in New England Journal, that looked at ARDS, severe ARDS, defined as a PETA F ratio less than 80, randomized to getting ECMO or being treated with standard ARDS net criteria. And what they found is that the patients that were treated with ECMO had a large tendency towards doing better. It didn't quite reach statistical significance, but there was a lot of crossover in patients that were randomized to control ARMS actually ended up getting ECMO. And that was one of the big criticisms, but it looked like ECMO was a good idea for a certain sub-select population for adult ARDS. And that's how we stuck with this criteria when we first started making our ECMO COVID criteria. And this is what our first go at our Candacea criteria was. We published this in the annals of American thoracic society. And essentially, we stuck with that indications for ECMO, severe hypoxemia manifested as a PETA F ratio less than 80. We went a little bit more conservative with the pH set of 725, we said 715. And the rest of the absolute and relative contraindications really were trying to limit patients that had a abundance of other comorbidities. What we wanted is we wanted young, healthy patients that had single organ failure secondary to COVID. And so we were pretty strict early on. Also, in terms of capacity and not knowing where this pandemic was heading, we really wanted to pick the right patients that we thought would benefit from this. This is what also published us a year ago and kind of thinking about candidacy for ECMO in relation to the capacity at the individual centers level. So you have over here on the left, you have the conventional capacity, which means patients, the hospital's functioning completely as usual. And we really don't make any changes. We're gonna treat the same patients that we would have placed on ECMO that didn't have COVID or prior to the COVID pandemic. And then there was the Tier 1, where hospital is still functioning as usual, but you've expanded services. You've opened up with COVID ICU, you've dedicated a team to care for these patients. Maybe you've slowed down elective cases a little bit. So that's kind of this Tier 1. And Tier 2 is in a little bit more severe state where you're gonna start actually being very selective and who you put on ECMO, really looking at the contraindications and potentially even... Okay, there we go. So the Tier 2 is we're really picking very select criteria, select patients, so no relative contraindications at all. People that we think that are gonna really benefit from this and not have added burden to the system. And then you have Tier 3, which is crisis capacity. And that's essentially ECMO is probably not going to be offered as a therapy in Tier 3. And the focus now is on trying to do the largest, greatest good for the greatest number of patients. And it's really not appropriate to utilize this highly intensive resource therapy. And we've never made it to Tier 2 or crisis capacity, which was, thank God, we really vacillated between this conventional capacity and Tier 1 in terms of where we fell. So we really, our criteria did change. And then we actually, I'll show you that we went through four iterations of our criteria over the past two years. And I think the most important thing was, we looked at our data, we looked at published data, and we looked at the local experience. And with that, we were able to revise and make our criteria more applicable to the patients that were coming into the hospital. And really make sure that we weren't restricting therapies or that we were putting patients on that we found out didn't do well. I'm just going to have to screen the end of the news and I'm going to go to that. Great, so you didn't see anything I was talking about for the past few minutes, that's okay. So basically, just highlighting that, we took the data that our experience, our local data, published data, and we revised our criteria. And we went through actually four different iterations over the pandemic. And really what we were doing is, during the first wave, we had really good outcomes. So we thought, hey, let's offer this to more patients. That potentially can do better. So we liberated criteria. You know, instead of 65 as our age cutoff, we said 70. Initially, we were going to, initially we were going to use 60 as our cutoff, but my chief basically told me that he needed to be eligible. So we increased it. So we went to 70 as our age cutoff. BMI was a big thing. What we found is that there were a lot of younger patients with high BMI's that were doing poorly. And we thought, well, these patients should not be denied at therapy that may be beneficial for them. So we liberated our BMI criteria. We had it in absolute contraindication. We moved it down to relative, and then we increased it from 35 to 40 and 45. And then essentially 50. The big criteria that really that we changed was adding an exclusion criteria regarding the duration of non-invasive ventilatory support. So I'll get to that. So it's very common for most centers to have an exclusion criteria regarding how long you've been intubated for, with the typical seven days being that cutoff. And the thinking is if you've had severe ARDS for seven days on high mechanical ventilatory settings, you are now entering an irreversible, long-damage situation where the likelihood of ECMO being beneficial is much lower. So a lot of programs have a seven-day mechanical ventilation criteria. Some programs are even more restrictive at five days. Some programs go up to 10 days, but nobody really thought about the duration of non-invasive ventilatory support prior to the COVID pandemic. And in the first wave of COVID, people were being intubated very early. You got put on six liters of oxygen, your next step was to get intubated for a couple of reasons. One reason is we were still scared. We didn't know what the transmission, we didn't know how effective this was. So the thought was, you were not going to do any aerosolizing procedures. If you have escalating oxygen requirements, we're going to intubate you in a very controlled fashion. While in the second kind of wave and later waves, when we got more comfortable, we started using non-invasive techniques. People were on bypass, people were on CPAP, people were on high flow. In fact, the pendulum's throwing so much that the public thought that getting intubated meant you were going to die, so they would actively not want to be intubated. And they would be on high flow for two weeks, three weeks. And then during that time, they could be having injurious breathing because they could be having huge respiratory effort with negative insritory swings, causing a large, transpulmonary pressure gradient. And that term is, kind of came out during COVID, it's called silly, self-induced lung injury. So the idea is that just time that they're spontaneously breathing, we have no way of understanding how injurious their breathing is, and they are probably injuring their lung, even though they're not on a ventilator, because of their increased work of breathing and huge negative respiratory pressure swings, they're actually damaging their lungs. So what we would see is patients on bypass, mostly high flow actually for two weeks, they would then crash, get emerging intubated. And within two days, they were on maximum settings. So within two days of being intubated. So we would evaluate them and say, they've only been intubated for two days. They're a great candidate for ECMO. We put them on ECMO, and we would see these outcomes were very poor. And we started noticing that, but then we didn't think much of it until we went to, we had a local regional conference, and we heard that, you know, cross town, they were having the same issues, and they actually added criteria, looking at duration of non-invasive ventilation and high flow as an exclusion criteria. And so we, you know, we're looking at our experience, we adopt that as well, and then we got on a couple of conference calls around the country and saw that a lot of centers were experiencing this, and a lot of centers were adding non-invasive ventilation and high flow as part of their exclusion criteria. And here's a study kind of looking at this, suggesting that this is, you know, what we were seeing is something that may actually be reported in the literature. So this study looked at, they divided respiratory support prior to intubation out of three-day mark, and they found that patients that had non-invasive or high flow for greater than three days prior to intubating had a less likelihood of being liberated from ECMO or mechanical ventilation. So now that brings us to cannulation strategies. So we figured out, or at least it was a work in progress, which patients we were going to cannulate, and then, you know, how are we going to cannulate these patients? So our standard at the Brigham was to do dual site cannulation, FEMIJ. We would do a 25 multi-stage shuangish cannula via the femoral vein, was its tip sitting in the IVCRA junction, and then a 19-french return cannula in the IJ, sitting at the IJ kind of SVCRA junction. And two reasons why we chose that. Early on, we wanted to keep all the cannulations in the room at the ICU bedside. So we didn't have to move patients around, transmission, et cetera. And so, you know, any of the single cannula dual luminescence require fluoroscopy and RTE. So we wanted to be able to do a simple at the bedside using a bedside ultrasound. So that's one of the big reasons why we chose the strategy. And the second reason is, as we started doing this, we found out that, you know, there is a patient profile of hyperinflammatory state, high cardiac output, and those patients needed high flows. So we're going to be able to achieve higher flows with a dual slight cannulation. Here's kind of a picture of how we did it in the ICU at the bedside. One of the things, you know, kind of halfway through the pandemic, you noticed that, you know, there was this hyperinflammatory profile of patients that were, you know, bigger, males, younger, and they needed a lot more flow. So we actually switched our cannulation strategy to a 27 French trainage cannula, 21 return. You know, they may have needed up to six, six and a half liters of flow to really adequately support their oxygenation needs. And then some centers, actually, their strategy was to do VPA ECMO or OXERVAB, which is essentially Venus pulmonary artery ECMO. So they would put in a single cannula, two lumens. You know, the one that they were using was the Protect Duo, and that would go in through the IJ with its tip sitting in the main PA. And you would train from the RA, and you would return to the PA. So it's essentially in our bed, in addition to ECMO. So you'd bypass the RV completely. And you know, this is beneficial for two reasons. One, the amount of recirculation was minimal, given where the drainage and return ports are. And then two, there were a touch upon later, there was a huge incidence of right ventricular failure in our COVID ECMO patient population. So right off the bat, they've already, you know, they're treating potential RV failure down the line. And the centers that did this, they report phenomenal outcomes, like way, way better than the outcomes of the other centers. You know, it was really hard to understand if they were putting the same patients on, whether there was, you know, a patient bias, but, and they were early excavating their patients as well, something that I'll touch upon that we rarely did for our COVID patients, just because we just couldn't. And so here's, you know, a multi-center retrospective study, looking at the three types of cannulation, you know, dual-site, VV, single-site VV, or the OXIAR-VED VPA configuration, and they found in hospital mortality lower in the VPA. But again, this is retrospective data, unclear for the same patient population. So there may be a signal, I'm not saying that, there definitely is a signal that VPA echoes the way to go for this. So what about the rest of the management? We really stuck to the basics. You know, early on, there's this thought that COVID-RDS is a different type of ARDS, and it's a different phenotype. And then later on, it kind of, you know, we realized that, you know, we're gonna treat this, how we know how to treat patients with a currestory failure with our standard building blocks of ARDS management. And I think the pendulum is really swung to that, you know, COVID-RDS is just another type of ARDS. And the different phenotypes that you saw, were consistent with different phenotypes that you saw in any ARDS patient population. So, you know, it touched upon this foundation a little bit earlier, but this was a nightmare. That's the day I'm in Baruch members this. These patients for some reason, they, you know, the goal with sedation in analgesia in restatory failure ECMO is really just to have them calm, comfortable, and non-injurious, and breathing. We put them on ECMO so we can rest the lungs. Ideally, we'd like to have them awake. In fact, if we could activate them, that'd be fantastic. But they can't be agitated and breathing in an injurious fashion. We found that the minute we lighten sedation on most of these patients, they're heart rate through in the 140s, the rest of the time rate is in the 40s and 50s, which for adults is really bad. And you couldn't get them comfortable, and they were just breathing very injuriously. So we had to, you know, recidate them. They were on three, four IV sedation drips. Three, four, enteral sedation meds. And then, you know, it's not infrequently that even after ECMO, we had to have them paralyzed for a brief portion of their time. We just could not get these patients comfortable. So, and that was an experience that many centers shared. So, you know, I don't know how, why? There is a hypothesis potentially that the inflammation, they saw in their lungs, because we could control their PCO2, we can control their pH, and we can control their oxygen, right, with ECMO. So we should really be able to suppress their drive to breathe, but even, you know, making them alkylotic, making their PCO2 low, you know, having high oxygen levels, they were still very agitated. And there's a thought that maybe there's some inflammatory process in the lungs, activating the stretch receptors, making it very uncomfortable for them. The other thing is it could just be severe delirium because prior to going on ECMO, they were paralyzed on many, many different trips. So we were just seeing them come out of a hyper agitated delirious state. But unfortunately, we never figured out the magic potion to get these patients awake, calm, and exibated while on ECMO. You know, the whole goal of putting a patient on ECMO for respiratory failure is to rest the lungs. So we chose standard rest settings in which we define as an respiratory plateau pressure less than 25, you know, a peep that keeps the lung open, a low respiratory rate and an FIO2, that's as low as possible. Typically, we would just pick pressure control 10, peep 10, respiratory 10, you'd achieve all of those goals with the simple 10, 10 and 10 strategy. And then volume management is paramount for any ARDS patient, but particularly with the COVID ARDS as well. We really kept them as dry as possible. Two reasons for lung recovery. And then this fluid pre-leading to RV, distension, RV overload is real. And we really have to keep a close eye on the volume management of these patients. Once we did that, it was easy, right? We just set back and waited, when we waited, the lungs recovered, we excavated them and we said, we'd great job. Absolutely not. These patients had several recurring issues that we had to deal with. That most ECMO patients have elements of these, but we saw them more intensely in the COVID ECMO patient population. The first one is refractory hypoxemia. So there are several reasons why we saw this. As soon as we went from cannulation to rest settings and prior to going to rest settings, these patients run very high ventilatory settings to keep their lungs expanded. And what we saw is that the next day or the day after their X-ray got much worse. They basically consolidated their lungs, as you can see here. So this is cannulation, and then this is two days after cannulation. You can see that the X-ray looks much worse. Same thing in this patient, cannulation, and then post cannulation two days later. And with lungs like that, there's no gas exchange occurring. We had tidal lions in some patients less than 10. It's a single digit tidal lions. When you have lungs like that, there's no gas exchange occurring at all. So oftentimes, I mean, put patients on ECMO for ARDS, they have some native lung function. So they're able to have some native gas exchange and then a majority of it's being performed by the ECMO. But when you have zero lung function, the ECMO is doing all of the gas exchange. And in order for the ECMO to give you an oxygen saturation of 90%, if your lungs don't do anything, you have to capture at least two-thirds of your cardiac output. So I mentioned that you have lungs like that. And if you're the hyperinflammatory profile, where you have a high cardiac output, for example, let's say that your cardiac output is 10, you have no lung function. Your ECMO flows four liters. That means you have six liters of blood flow going through the native circulation, native lungs that are having no gas exchange. So even on high flow ECMO, you're not able to have an oxygen saturation greater than 70%, greater than 75%. So we saw this quite frequently. I would say probably a quarter of our patients that we placed on ECMO had this profile. So what did we do? Obviously, the first thing we did was make sure that the ECMO circuit and oxygenator and pump and everything were working fine. We monitored, made sure that we didn't have any recirculation. And then we cranked up the flow as high as we could go until we started to see complications like chatter, he molasses. And if that wasn't enough, then we had to really figure out, was there anything else we could do to optimize the lung? Meaning could we squeeze out a couple of more LV OLY to have a native lung function? You know, oftentimes we would put a soft agio balloons in to determine what the optimal people is. Maybe 10 wasn't a good people to this patient. Maybe they needed 15 to keep some lung open with some gas exchange. We rarely did this, but we even prone patients on ECMO to see if we could get a little bit more gas exchange to have their SAT greater than 75% or 80%. The other thing that we tried and experimented with is kind of naturally suppressing the patient's cardiac output with Esmolol. To maybe now we suppress their cardiac output, we have a higher shunt fraction going to the ECMO circuit. That sometimes made the number look better, but did it really make the patient better? It was unclear. And the other thing that's to keep it close on was they were hyperinflammatory and high cardiac output, but in two days, they could be RV failure. Esmolol is not a great drug for RV failure. It's just to be very careful when we were doing things like that manipulating the cardiac output. We routinely, we would heavily sedate them, paralyze them to decrease the patient's metabolic demand. And then we would shoot for a higher hemoglobin. And children are not little adults. We tolerate hemoglobin in the seven, eight. That's our standard for critical care. But in these patient population, where the sat's were not greater than 80%, we would increase hemoglobin above 10, to increase the oxygen delivery. If we can't saturate, well, at least we have more hemoglobin to deliver the oxygen. And if the lactate was normal, the end organ was perfusing, we were okay with keeping them paralyzed, sedated, hemoglobin 10 sat 75% until they got better. The end of the line was to consider another drainage cannula to increase flow. We never actually had to have got there. And then this is kind of as in line with what the ELSO guidelines are regarding oxygen saturation. But this was a paradigm shift for the ICU team's care and feed patients. It's very uncomfortable to see somebody on ECMO saturating 75% within oxygen saturation. And we had to kind of re-educate. This is verbatim from the ELSO respiratory failure guidelines. And I would print this out, highlight the arterial saturation is usually 80 to 85%, but maybe 75 to 80%. This is ample oxygen hemoglobin saturation for normal systemic oxygen delivery. And then bolded this second to last statement and then highlighted, avoid the temptation to turn up the ventilator settings or FIO2 above rest settings. And the idea is that we put patients on ECMO so we can rest their lungs. Now if you have them cranking out the ventilator on 100% FIO2, we're not resting the lungs and that's going to delay their recovery. If they're maintaining oxygen delivery with a status 75%, it's okay. And that's how we treated these patients. Here's an example of a patient for, this is about four weeks worth of blood gases here. And not once is the PIO2 greater than 55 and dips as low as 40. And this patient ultimately recovered completely neurologic intake and left the hospital to go get married after he left the hospital. So that was the way we dealt with reflector hypoxia men, which is not the way that every center dealt with it. So here's a, this is the group out in Vanderbilt. And what they did, they also saw about 25% of their patients had refractory hypoxemia and they were a lot more aggressive and they put a second ECMO circuit, whole ECMO circuit. So they have a typical FEMIJ cannulation strategy that we had on the right side and then to increase flow, they would put a whole nother circuit on the left side. And so if they were flowing 5.5 liters, now they could flow seven, seven and a half liters and they would fix the hypoxemia. And they had a remarkable 72% survival rate in that subpopulation of the reflector hypoxemia. And I saw case reports of this and never actually thought about doing this prior until actually two weeks ago, this is a non-COVID patient, but we had a 18 year old six foot two who either has RSV or vaping induced lung injury. And we had one circuit flowing 5.5, six liters. We could not keep up with this cardiac output on ultrasound, it looked like the estimated cardiac output around 12 to 13 liters per minute. So we trial this dual circuit option and it worked very well. We were able to flow seven and a half liters, we were able to unparalyze him, get the sedation off and ultimately extubated him. So we'll see, I'm a big proponent of this news in the end of one, right? That's all you need. But that's how some centers dealt with the reflector hypoxemia they saw during COVID. I mentioned this before, but be mindful of RB failure. We saw this and about a quarter of the patients had significant RV failure and about 10% had refractory RV failure. And RV failure on ECMO for ARDS really just increases your mortality significantly. So we had a very aggressive strategy of monitoring serial echoes. Any new hypertension, get an echo, make sure the RV's not failing. It treated ionotropes, vasopressors, I know dilators inhaled the repostinal and then heavy dioresis. And if we were unable to achieve it with medications, we would put them on CRRT to really drive them out. And then, you know, like I mentioned, 10% of our patients we actually had to convert to VAV ECMO. So here's one of the, here's the case. One of the first patients we placed on VAV ECMO. So it was a 46 year old, no other medical problems. Get's calculated standard VV ECMO fashion. On day 12, yes, escalating shock. They're trying to diorecem because the RV looks like it's not doing well. They're unable to. And then in the 20, in a 48 hour period, it becomes nine meters positive. So, you know, I come around to them and you look at it, is echo and this is what we see. So this is a parastronal short axis. This is the RV, this is the RV. And the RV is, you know, three times the size of the RV and it's squishing on the septum. So here's an apical foretam review. Again, looking at here's the RV. So the RV is small, underfilled. The RV is huge and very hypocontractile. So we converted from VV ECMO to VAV ECMO. And, you know, what that means is, you had your standard thermal drainage cannula. You have your return cannula and circuit tubing post-oxidinator here, you split it. You put a Y connector in there and now you have two return cannulas. One going to the IJ to supply its, you know, VV ECMO portion. And then one going to the thermal artery, which bypasses the heart. And that allows us basically to unload the RV. And with this strategy, you actually don't need tons of flow through the arterial limb. Once you get a liter, a liter and half off, you know, the RV, our research will look a lot happier. And then you're able to, you know, support the patient. So we did this, you know, five times out of the 50-some-odd patients that we did. And again, you know, if they need this, they're in a very bad situation. So here's the pre-ecco I showed you and then here's the immediate post-VAB ECMO. And you can already see if the RV is not twice the size of the RV. It's, you know, the same size of the RV. The RV is now able to actually fill and contract. And ultimately, this patient, you know, did well, decannulated from all ECMO. Unfortunately, five days after decannulation, you actually developed hypercarbic respiratory failure. And we end up having kind of a limitation on how many ECMO circuits we had. We placed them on an extra-corporal carbon dioxide removal device that we just obtained for emergency utilization and to use it on him for the first time. And ultimately, you know, he went to the whole spectrum. He went VV, VA, VV, and then E-core. And then he was discharged and did well. That's his initial X-ray and that's his discharge X-ray. So the other issue was this debate that COVID was a pro thrombotic hypercargable state versus, you know, a bleeding state. And different centers had different experiences. I don't know if it's patient population, it's what's in the diet. But there are centers that all they saw were thrombotic complications and they used higher intensity anti-quagulation. We saw the reverse or the opposite. We saw the bleeding side. We didn't really have many thrombotic complications, but we saw bleeding from everywhere. We had several severe ENT bleeds that required packing by ENT. We had, it was routine have, you know, tracheostomy bleeding, hemoptosis bleeding from cannula sites. So we actually used a lower intensity heparin algorithm. And in fact, many of our patients, we withheld systemic anti-quagulation throughout until they didn't have a bleeding risk. And we did not see an increase in oxygenator or clot kind of burden as opposed to our standard anti-quagulation strategy. The other thing that we saw a lot of were, when we tried to manage the plural space, they would just hemorrhage hemothorax. So we became very conservative in putting chest tubes and pigtails unless they absolutely needed it. We said, nope, don't touch them. Don't, you know, they're gonna bleed out. So that, you know, that was our experience in terms of the bleeding versus thrombosis. In this study here, done in Europe, they found that there were more bleeding events than thrombotic events. But you talked to other centers. They had the exact reverse where they were doing pTTs of 100 above 100 because they were having thrombotic complications. And then the last thing that was very different from the COVID population from non-COVID to RDS was the duration of how long they were on ECMO. So pre-COVID, the median time for RDS ECMO was about two weeks. And that was our experience. That was the experience in the major trials. And that was the experience in the ELSA registry data. So two weeks. So it was very rare for us to have somebody on ECMO for RDS for more than 30 days. That was, you know, I could count that prior to COVID, you know, once or twice we did that. I'm not talking about the bridge to lung transplant population where they're on it until they get a lung transplant, but the bridge to recovery RDS population very rare that we had patients on for more than 30 days. And, you know, what we saw in COVID, the first wave was very similar, like we talked about traditional RDS, but the later waves, we saw durations of ECMO that were way longer than anything we would experience before. The median duration was 24 days with, you know, it falling between 11 and 39 days. If you took out the first wave, the median duration was closer to 30 days. And then our longest ECMO run that recovered was 142 days. And in the adult world, you know, we don't typically keep patients on that long and wait while waiting for recovery, but what COVID showed us is that there was remarkable potential for the lung to recover if it remained single organ failure. And that, you know, these cutoffs of 30 days or two weeks, really, were, didn't make sense. And we're, you know, we had to kind of tailor the situation to the patient. So here's an example of a patient that was on ECMO day 44 as long as, you know, still look pretty bad. And then he has complete recovery and he's got a little fibiotic changes, but as long as, you know, essentially go back to normal. Here's another patient, day 34 on ECMO, his lung still don't look great. And then, you know, this is prior to discharge, we got a repeat CT scan. So, you know, just within a couple of, you know, weeks, his lungs go back. And then here's the patient. And this is day 39 of ECMO, complete opacification of lung still. We are telling the family no chance, you know, she's, she's, you know, over a month. There's no way she's going to make it. We should really start thinking about comfort measures, palliative, changing the goals of care from treatment to comfort. And, you know, they said, no, we said, okay. And then 20 days later, lungs clear up, get her off ECMO and she walks out of here. We had so many stories like that, especially during the first wave, where at day 30, you know, we thought for some reason, we had 30 days as our, you know, our number of, should they recover or not? And then many patients recovered after 30 days. That really changed the way we thought about single primary, just one organ down non-recovering. So that brings us to the next questions. When, you know, it's taking so long for them to get better, when you consider transplant. And the answer is, we don't know. But this is a published in Lancet Respiratory. And now, Unkid Peraud over in Northwestern, you know, he really is leading the way, you know, lung transplant, COVID lung transplant. So this is him with a couple of other centers internationally. They published their, you know, their case series of COVID ECMO patients that had lung transplants. And their short-term outcomes are the same as patient that were standard lung transplants. But the median length of time on ECMO before being listed was 49 days, but the range of 38 to 80. Again, that may be too early because we've seen recovery after 80 days, after 90 days. But it still remains. And then this is a small study done out at UT Southwestern. And they kind of just looked at, look at our patients that had ECMO for more than 30 days that were, you know, for COVID. And they had 10 patients. And they found that, you know, the median duration of support was 85 days. Six patients survived to decannulation. Three patients died. And only one patient received the lung transplant. So I think, you know, the answer is we don't know. There were six patients that we, you know, they weren't listed, but that we actively consulted the lung transplant team. And then these are their outcomes. Three died and three recovered. You know, one recovered them day 36, day 46, and day 142 while we had, you know, the other patients, you know, died during that time period. So it's still a long fear when, you know, it's considered lung transplant. You know, ideally, if they're otherwise young, they don't have recovery and are otherwise good candidates. You know, centers are thinking to just start the process somewhere between six to eight weeks. You know, once you get listed, you don't get along automatically. And then, you know, see if they recover. Some centers are waiting 12 weeks. Some centers were initiating at six weeks. But, you know, I think the future is, you know, are there CT scan changes? Are there radiographic changes that we can look at that can predict non-recovery? Versus recovery. Right now, we don't know. But I think that may be where the future is for prognostication of these types of patients. And then lastly, just a little bit about the data. So the interesting thing that was experienced throughout the world was worse mortality in the later waves. So that was universal. First waves, great outcomes, very similar to traditional ADS. Later waves, worse outcomes. And that was our experience. And that was the experience everywhere else. Now, why did that happen? Still not entirely sure. These are some of the proposed reasons. You know, the big thing that I think is definitely played a role is the later waves were a different phenotype. They were patients that failed COVID directed treatments. The first wave, there were no COVID directed treatments. They were intubated very early. So they were just very inflammatory standard ARDS patients. The second wave, they weren't intubated early. They were on high flow. They already had steroids. So they had a full dose of treatment. And they may or may not have had some other immune modulators, et cetera. So those are the patients that we were seeing. And those patients, very different from the first wave patient. So maybe there's something there. The COVID variants. And then lastly, we saw this with doing so well with the first wave, we liberalized our inclusion criteria. And then we deliborized them as we saw some of the outcomes. So that may changing in criteria. Another thing that we saw is centers that weren't ECMO centers became ECMO centers. So maybe they didn't have the experience in the volume and then those outcomes weren't as good. And then lastly, we don't have any randomized control trials looking at COVID ECMO for severe COVID RDS versus conventional management without ECMO for severe COVID RDS. But we do have several observational studies using fancies statistical methods to make the two groups as similar as possible. And this was a large study that was published in intensive care medicine during after the first wave. And they found that ECMO was beneficial in these patients. And then this was a larger study that they published in BMJ or a couple of months ago. And they also found using their comparative effect in this method that ECMO did confer a mortality benefit. And they found that ECMO was most effective in younger patients that had severe hypoxemia and that were intubated for less than 10 days. And then I just want to touch briefly upon kind of the institutional things that we have to do to be able to care for these patients. So once the pandemic started and it looked like this is going to be a big problem here. We had to start figuring out how can we accommodate these increases in flexes? You should be able to flex up and be able to flex down and take care of the patients until we reach that severe crisis capacity where ECMO is not even on the table anymore. And what we found is equipment was obviously one of the big limitations. And in fact, children's hospitals helped us out a lot. And in the fact that they weren't hit as hard from the ECMO standpoint. So you guys were actually able to help us with some pumps and circuits to borrow while we were in our capacity. We also had to think about using non-traditional methods of ECMO. So using our percutaneous, our center mag that we use as our bivod, well, we can displace in an oxygenator and now we have ECMO. The other thing that we were able to change was our staffing models. Prior to the pandemic, we were one ECMO specialist per pump. Regardless of how sick that patient was. Every patient on ECMO had an ECMO specialist. Well, we found out that we could actually share responsibilities with the nurse. And now we're able to have one ECMO specialist cover three. So changing staffing models was an important aspect. In fact, we're training some of the cardiac surgery nurses to be ECMO specialists now as a result of this. And then regional collaboration was the really important. We had this consortium that we formed very early on with the Neckmo consortium that we met in the first wave. We met very regularly and we kind of, not necessarily standardized, but we agreed upon common indications. We talked to each other about our experiences. That was the first time I heard about the high flow non-invasive problem. We talked about how they're dealing with the high cardiac output, hypoxemia issues. And it was all via Zoom and it was very helpful. As capacity became more of an issue, we used that group and coordinated with the Region 4 Emergency Management Area System to figure out how, with Incordination Boston Med Flight, how can we make sure that the outlining hospitals know which hospital has ECMO capacity because it was very frequently that we were at capacity, MGH for the capacity, BI with the capacity. But oftentimes there was one hospital in the area that had capacity. So we created this kind of system that they would be able to at least refer the patient to the hospital that has the ECMO capacity. So the summary, we learned a lot during COVID with regards of how to care for patients at ECMO and how to use ECMO within the settings of a capacity crisis situation. And I think there's a lot more that we need to, in terms of candidacy, cannulation strategies, lung transplant that were still, you know, just at the beginning, figuring out where to go from there. That's it. Wow. As terrific and appreciate it. I, we don't put up a question since that, and I regret that we could see that we have Q and A on the Zoom and there is nobody in this auditorium that actually has post-access to read those questions so we apologize. Oh, maybe it may back into the video. You can't take it. I would say that, you know, for those of us who've been taking care of infants and babies on ECMO for decades, we use, you know, see the dismal results with adult ECMO, but that's really changed. And this, you know, what is still relining from this horrible pandemic that the adult world is now really, really experienced in dealing with that moment, you feel good out the ways to do this. Are you able to get access to those questions? So first, is anybody in the room who has a question? A comment? Dr. Loha. Yeah, this indeed was spectacular. And thank you for walking us through just your thought process and the heads you made those adjustments along the way. There were two specific things I wanted to ask about is, and that is one, how often were you, I mean, and did you, I think you alluded to at least one, convert from VV to VA. And number two is, can you tell me more about those, VA can't, the VA can't even, isn't how you directed those into place? Yes. So we, we did, we put 51 patients on ECMO and five of them we converted from VV, and not to VA, but to VAV, because they still needed, you know, oxygenated blood to go through their own, kind of, circuitry system. Otherwise, we'd get the North South syndrome that we see in adults. We don't have the, the, the, the luxury of doing carotid artery cannulations, which would send oxygenated blood forward. So we did that about 10% of the times, we had to go to the VAV. The, protect duo cannula, it's essentially, you know, we, it's done in the, that's the cannula, that's the, the, IJ to the PA cannula. So it's a, you know, dual-loom cannula 31-French, that essentially, you, you, you need fluoroscopy. So, you know, you place, you know, kind of a PA, a blue type of a PA line that has a balloon that goes, you know, into the PA, put a guide wire through that, and then you direct the cannula through there. So it's all, it needs fluoroscopy. Some centers have tried to do it with TE alone, but I think you need, it's best for fluoroscopy. And you don't necessarily need that particular cannula. What we've figured out, if we want to do that strategy of, you know, venous pulmonary artery, is to still cannulate from the femoral vein as your drainage cannula, and then put a single-stage long cannula into the PA. And the reason that's more, that we think that's better is that you can achieve higher flows because you have the single large cannula draining here, and now your return. It's also much cheaper. Great, thanks. Maybe we'll let you choose from the most instructive questions on the Zoom since another rest of us can see them. Yeah, so the first question here, from Sally, of Italy, is, how has the COVID-19 experience changed like most practice for all patients? I think the big thing is waiting for long recovery in single organ failure. So if the patient does not go into multi-organ failure, and it's just that their lungs have not recovered, we will not say that they are non-recoverable unless they have another organ injury organ failure. So really just given time for that long to recover. That's, I think, one of the biggest changes that I've noticed in our ACCO program. And then other changes I think are being more aggressive about dealing with hypoxemia. Like I mentioned with that, young kid is not letting them sat 80% and paralyzed and sedated, but seeing if we can actually do more advanced accangulation techniques to support them. And then by Jeff Burns, how do we get past observational studies and selection bias that comes to them for the next steps? That's a great question. Now unfortunately, as everybody here knows, doing critical care trials and ECMO trials is very difficult. Oftentimes we don't achieve the preset out and that we want. We go over long periods of time, you don't enroll. So I think that's a great question. I think we need to do more randomized control trials in this area. I think that's it. Yeah. You got one quick question. So I, we're one of the fellows, at least in our CDH population where we run a lot of ECMO here. We tend to be on the side that has more of the thrombotic complications. And I was just wondering if there's any differences that you see, I mean, because you guys are just next door and you mentioned that a lot of times you're using the same pumps and circuits that even we have. So any other hypothesis for why there are such dramatic different profiles? And you're seeing thrombotic complications in the circuit. In the patient. In the circuit. Yeah. And how much do you guys flow usually in a small? We're pretty low. We're getting close to 200 MCCs. Sorry, we're done. You guys are too much. I'm not going to talk about it. We'll close 100 per kilo and a 3 kilo wave is 100 CCs. So that's. And I mean, do you guys, what pump do you use for that? We still use mostly roller pumps. So we're in the transition towards the typical pumps. Sort of, but, but still traditionally using roller pumps. Yeah, so I think it's probably that. And when we do run them with no anti-coagulation, we do keep the flows above. We try to keep them above three meters per minute to avoid thrombotic complications. So I think that's the answer. Right. Well, thank you so much. It's highly instructive. And, and, you know, during this, this part of the, you know, pandemic, we were living in two different worlds right across the bridge from each other. Yeah. You had the adult hospitals were full of really sick patients where you were doing all the surge capacity and borrowing machines from us. And we had pretty much an empty hospital because nobody's coming in for anything elective and people are socially distancing and not getting sick. We're actually living in the opposite now. We're all the children's hospitals. This is our COVID where we're full of. These horrible respiratory problems and we're changing for RSV and, and, and, so not so much COVID and influenza. And so I think we're having that to learn, to learn from you. And, and, and, and, and, and, and, thanks for, for educating us in this and, for, actually, all you and all your colleagues did for the community during that horrible time. Thanks for much. Thank you.