everybody. Welcome to Hawaii and welcome to our session on severe ARDS, state-of-the-art management. It's really a pleasure to get to speak to you today. I'm Dr. Dan Ouellette. I'm chief of pulmonary at Henry Ford Hospital in Detroit. The topic today is one that is of special importance to me and I hope to use since you're here in the audience. I put together this session last year thinking that it would be an important topic for us to discuss and I invited some friends to come here and talk with me. So you'll be meeting them through the course of this morning. The speakers on the panel are all younger faculty and I think they have exciting viewpoints to give to you today. We're a diverse group but except for one thing, geographically, we're all from the Midwest. So we're happy to be in Hawaii and I hope you enjoy your time with us today. I need to read to you the following. Keep your phone handy. Use the app to answer audience response questions by finding the session in your schedule and clicking on the polling link at the bottom of the session details. There will also be a QR code that you can scan each time a question is asked. So thank you very much for that and I'm going to go ahead and start to talk to you about PEEP. So this is who I am, chief of pulmonary at Henry Ford Hospital in Detroit and we're going to talk about selecting the right level of PEEP in your ARDS patients. I have several disclosures to talk to you about. The only one that is probably relevant is the last one with Dompe Pharmaceutical and that's about a biologic, it's not about mechanical ventilation and we haven't started recruiting yet so I'm not sure how important that is but that's to you to decide. What we're going to do today is to review the evidence that implicates outcomes in ARDS as a result of the application of PEEP. We're going to learn about using PEEP tables and how to use them to determine the amount of PEEP to apply and we're going to discuss certain physiologic determinants that may be important in selecting PEEP. So here is my code so you can click on that. I do have a couple of embedded audience response questions. I will give the disclaimer right away that these are not board questions, they're meant to be provocative and controversial and I certainly hope that we don't all get the same answer. I've selected answers that I like the best but if you disagree with me that's okay because my two questions have elements in them that I think are important to discuss beyond that. I like to start with a case presentation. This is a patient from my ICU, a 68-year-old man who comes in who's been short of breath, he's got a non-productive cough, he's got airspace disease on his x-ray that you can see, he's febrile, he's got a high heart rate, a respiratory rate that's elevated and he's borderline hypotensive. Not a surprise that he has pneumonia. His laboratory studies are here and his initial room air blood gas at the bottom shows that he was hypoxic, although at the time of first look he did not have an acidosis. He had a past history of diabetes, hypertension and what did we do in our ER at Henry Ford Hospital where people started talking about goal-directed therapy 20 years ago, we gave him antibiotics and crystalloid and put him on a non-rebreather. He had the pneumococcus, he worsened over the next couple of hours, he became acidemic and he remained hypoxic despite being on a high level of oxygen. So, according to the best available evidence, which of the following is true? There we go. So there you go. There's four choices there. Patients with moderate to severe ARDS have reduced in-hospital mortality when treated with high levels of PEEP. Patients have improved outcomes when optimal PEEP is determined by metrics related to lung compliance as opposed to PEEP being set using PEEP tables. Patients with high levels of PEEP are more likely to require vasoactive agents and the FIO2 levels for patients treated with high levels of PEEP were lower than compared to those treated with low levels of PEEP after treatment. And again, there may be several choices. My favorite answer here is the last one. I'm going to talk about all of these elements here a little bit in the next few minutes because I think they all have elements that we need to address and talk about and I don't disagree necessarily with anyone who's answered any of these responses and I'll tell you why. I got this, I wanted to get a copy of Dr. Petty's article in 1967 and I asked my library about this and they sent me an article and I took this image and I just reflected when I saw it that, you know, this journal was not in a digital copy. In fact, you can't find a digital image of it and it reminded me when I was a resident of clipping articles out of my journals and putting them in an actual file cabinet and if I didn't have it or couldn't find it, I had to go to the library and look at something like this which is a book. My son, 25 years old, thinks that this is amazing but this is how it was. I also would look, remind, take a look at the chart that I've given you there. This is a patient who on the second line you can see is PO2 or her PO2 improved dramatically with the application of PEEP and it was this article by Ashbaugh and Petty in 67 that was really the first note that PEEP could be effective in improving oxygenation with application of PEEP. This is from the alveoli trial, this is a PEEP table, you guys are all familiar with looking at that. I'm just putting it there to highlight the fact that such things exist and that there are higher and lower levels of PEEP that are used. One way to set PEEP in an ARDS patient is by using the PEEP tables. A second way was described in the EXPRESS trial and in the EXPRESS trial what they did was they used a minimal distention group which is their control and had a total PEEP between 5 and 9 and then they used an increased recruitment group which used a PEEP that was targeted to maintain the plateau pressure between 28 to 30, a second method of setting PEEP and you can see by the mortality graphic to the right that the mortality in the second group, the group where the plateau pressure was targeted between 28 and 30 had improved mortality in this trial compared to the control group with a relatively low level of PEEP. So both the alveoli trial with the tables and the EXPRESS trial with this way of measuring or applying PEEP could be used. I'd like to briefly mention a third method that hasn't been studied as carefully but this is to use something called the stress index and what one does with this is relies on ventilator graphics to give you an assessment of lung compliance to try to determine whether or not the applied level of PEEP is optimal. To do this properly you have to have your ventilator set in an assist control mode with constant flow during inspiration which hence the flow time graphics at the top of the slide and then you look at the curve of the pressure time scaler, the pressure time graphic to help determine what the stress index is and whether or not an optimal level of PEEP is being applied. So when the stress index is less than one, the idea is that you can use more PEEP. When the stress index is one, PEEP may be optimal and when the stress index is greater than one as determined by the configuration, then you may be causing over distention. And so those are three methods, three methods that I commonly talk about. In fact, three methods that I often use in concert when I'm taking care of my ARDS patients. I often look at the table, look at the plateau pressure and then maybe take a look at this graphic also to help me know if I'm using the right level of PEEP. I'd like to turn now to some meta-analyses and that's where I think in 2023 the waters become muddier because I'm going to talk to you about three different meta-analyses largely looking at the same trial, asking the same questions about the use of PEEP and ARDS using different methodology and coming up with different answers. And so I'd like you to reflect on this and learn this information and think about how it applies to your practice of medicine. This is a meta-analysis that was done and published in JAMA in 2010 and it's called an individual patient data meta-analysis. Normally when you do a meta-analysis you take the data that's provided for you in the target articles and you extract them and you use them to come up with a synthesis. In this case they took three trials, four really, three that were included in the primary meta-analysis and they got the individual patient data and then they re-assessed the individual patient data to come up with some conclusions and in this JAMA article they used the alveoli, the LUVs and the EXPRESS trial. What they found out was that in patients with severe ARDS that there was actually a mortality benefit and a time to unassisted benefit in those patients with the most severe disease if you used high levels of PEEP, the high PEEP table or the high measurement. And it was from after I read this article that I was fairly convinced that I should be using high levels of PEEP in most of my patients either as indicated by the PEEP tables or by using the plateau pressure method in the EXPRESS trial. And so looking at their data what we found, what they found was that in the patients with the higher levels of PEEP that the FiO2 required to maintain adequate saturations was lower in the higher levels of PEEP. There were no differences although a trend in pneumothorax rate and no difference in vasopressor use, hence my question about vasopressor use not being changed by the use of PEEP. This is a different kind of a meta-analysis. It was published last year and this is a network meta-analysis and the strength of a network meta-analysis is that it allows you to take multiple interventions and compare them. And so you can see here that they looked at low PEEP, higher PEEP, PEEP with and without recruitment maneuvers and also esophageal pressure guided measurements. And in a network meta-analysis compared these and the results are listed here. And what I would point out in the top box which is the highest level is that it was thought that higher PEEP without lung recruitment maneuvers was probably more effective than lower PEEP when you're determining PEEP. And you can see that there are some other elements there. So with a brief recruitment maneuver, high PEEP is better than low PEEP and may result in little difference, esophageal pressure guided versus lower PEEP. So they went through the various comparisons and again last year when I read this I thought that kind of confirms what I knew from the JAMA meta-analysis that maybe higher levels of PEEP are important. So we have to fast forward now to 2023 when we have a new guideline from the ESICM about the use of PEEP in ARDS patients. And they did a standard meta-analysis. And so I put the plot there for you to look at. They looked at the same data, the same studies but they did a standard meta-analysis and they found no difference between higher and lower levels of PEEP and they therefore made no recommendation about the use of higher levels of PEEP. And that's why I say today maybe the waters are a little bit muddied because three different types of meta-analysis with three different ways of looking at the data gave different answers. And so today I would be less likely to say that I'm certain that higher levels compared to lower levels of PEEP are important. This is the second recommendation and it was a recommendation about PEEP titration guided by respiratory mechanics compared to PEEP titration based on a PEEP table or PEEP to FR2 method. In this meta-analysis they looked at four studies. They looked at the EP vent 1 and EP vent 2 where esophageal pressure measurements were used to guide PEEP level setting compared to in EP1 a low PEEP table and in EP2 a higher PEEP table. And the second two studies involved looking for optimal PEEP by targeting the smallest driving pressure in the patients as they incremented PEEP. One of the studies, the ART study, looked at recruitment maneuvers as part of that process. And what you can see in the meta-analysis that was done in the Griselli clinical practice guideline is that there was no difference in using these different kinds of strategies. To get to a little bit more granularity, on the left you see some data from the ART trial and actually using lung recruitment maneuver and titrated PEEP they found that mortality was actually increased in that particular study compared to low PEEP. In the other three studies that kind of mortality difference didn't bear out. In the right is a plot from the Griselli guideline in their appendix looking at barotrauma in these measures. And you actually see that there's a trend in the compliance metric types of targeting to more barotrauma although it wasn't significant. And so I think that may be important as well. Here's my next question. In severe patients with severe ARDS, which of the following is true concerning PEEP? Regional over-distention does not occur in ARDS when PEEP is optimally applied. The smaller the difference between the plateau pressure and PEEP, the more likely the patient is to suffer ventilator-related lung injury. High levels of PEEP may be more beneficial with a hyper-inflammatory sub-phenotype or esophageal monitoring is not useful because of increased lung compliance. What do you think? I like the answer that most of you are picking. I'm going to go through each of these briefly to talk about this. So clearly regional hyperinflation can occur even when you optimally apply PEEP. This is an animal study and it looks at nuclear medicine graphics of the thorax of these animals during various maneuvers. There's an LPS and hydrochloride group to be models for ARDS. And what you can see is that with lung recruitment maneuvers and so forth and using open lung ventilation, that even with open lung ventilation, there's areas where the lung may be over-distended. So regional over-distention can occur even if you're optimally applying PEEP, so that answer's not true. This is from Amado's study in the New England Journal and what we should know is that a lower driving pressure is actually better for patients and this shows that data. And in terms of esophageal manometry in EP Vent 1, there was some evidence that this measurement worked, but it was really optimal targets were, it's because of decreased lung compliance in ARDS patients and not increased lung compliance. EP Vent 1 using esophageal pressures demonstrated some benefit from using esophageal pressure guided PEEP setting, but EP Vent 2 didn't confirm that. There's some differences between these two. EP Vent 1 is a single study. There was a less, single center. There was a less aggressive comparator PEEP. There was a higher driving pressure in the comparator and extra pulmonary ARDS and with, are predominated. In EP Vent 2, there was multiple centers. The control group had higher PEEP and in that study, there was no difference in using that. One last thing I'd like to put in your mind that in the future, we may be looking at personalizing the therapy for our ARDS patients. This is from an intriguing article by Cathy and they found that they could divide ARDS patients into two different phenotypes, one of which in the red line is called an inflammatory phenotype. A single biomarker does not distinguish between these two, but when you look at a constellation of biomarkers, one can see that there are two different phenotypes that can emerge in our ARDS patients. I'd like you to keep that in mind and then think about the following because in Cathy's article, what they did then is to look at the effects of PEEP, low PEEP versus high PEEP and in the first line, the mortality line, you can see that the higher levels of PEEP in the patients with phenotype 1 led to higher mortality, but in those with an inflammatory type of PEEP or type of ARDS, higher PEEP actually led to slight improvements in mortality. Also, ventilator-free days and organ-free days were outcomes which were improved as well, suggesting that maybe in the future, we may be distinguishing phenotypes of ARDS and applying PEEP in other strategies is different depending on what we learn about those phenotypes. Problems with the current tools are they're empiric. PEEP tables are one-size-fits-all patient. When we get a big person in Detroit, we think about using esophageal measurements because we know that chest wall compliance is a significant contributor and determining optimal PEEP becomes more challenging. If you use a stress index, the accuracy of graphics is of concern. There are potential errors with esophageal manometry, so we don't have a perfect tool. Future directions, I think, is that we should look at recruitability. Should we look at recruitability? What's the hemodynamic cost of using PEEP? What's the role of using esophageal pressure-guided measurements? And I think finally, I think the future may hold a personalization of ARDS and phenotyping of ARDS patients and selecting therapies that may target their physiology better. In terms of COVID-ARDS, the recent guidelines had no recommendations. There are my references that you can look at if you like, and I want to thank you very much for hearing about optimizing PEEP. I'd like to introduce our next speaker, Dr. Alex Garbarino from the University of Cincinnati. Alex? Good morning, everyone. Thanks, Dr. Ullett, for that nice introduction. Happy to have you all here today. My name's Alex Garbarino. I'm one of the associate program directors down at the University of Cincinnati, and it's my pleasure today to discuss the use of prone positioning in severe ARDS. I have nothing to disclose that's relevant to the talk. Today I'll briefly touch on the physiologic advantages of proning, touch the evidence behind its use in ARDS, address the challenges with its regular use in ICUs, and identify some promising areas for future research. I, too, will be packaging a little case-based question, so if you haven't already bar-coded in, now's your time to do so. All right, so let's start things off with a case just to get our minds working. You have a 47-year-old obese patient with a history of hypertension, diabetes, comes to your ED in respiratory distress. His wife notes that a number of coworkers were recently out sick with influenza. He's tachycardic, tachypneic, and hypoxic, despite oxygenation with a non-rebreather. He's intubated in the ED and admitted to your ICU. You start lung protective ventilation. You set your FIO2 to 70%, and his PEEP is at 12. After 12 hours, you have an arterial blood gas that shows a PO2 of 65 with a PF ratio of 93, and so you consider prone positioning to assist with oxygenation in the patient. So which of the following is correct regarding the use of prone positioning in ARDS? Patients who undergo prone positioning must be paralyzed. Eight hours of prone positioning a day was shown to have mortality and ventilator-free benefit in severe ARDS. Prone positioning causes a decrease in chest wall compliance and an increase in lung compliance, or prone positioning is unsafe in pregnancy and should be avoided. We'll circle back to this at the end of the talk. Since the 1970s, there have been studies showing the physiologic benefit of prone positioning with regards to ventilation and perfusion in human lungs. Because the lungs are a cone-shaped structure stored in a cylindrically shaped container, hydrostatic and gravitational forces are distributed unevenly in the pleural space. Thus, the anterior pleural pressure in a supine position is more negative relative to posterior pleural pressure, leading to anterior overdistension and posterior atelectasis. This atelectasis reduces pulmonary compliance, and the resultant higher ventilator pressures are unequally distributed to the already distended alveoli, leading to ventilator-induced lung injury. Placing patients into the prone position leads to a more even distribution of these forces, which allows for a more homogenous lung ventilation and perfusion, improving VQ ratios and oxygenation. Quantitative CT studies have demonstrated a more even gas-to-tissue ratio, which is a surrogate for alveolar distension, in the prone position, irrespective of the presence or absence of lung disease. Gattanoni and colleagues demonstrated in 1997 that proning patients with acute lung injury led to improvements in oxygenation and lung compliance within two hours, although the compliance of the chest wall significantly decreased during the same time frame. Indeed, some studies since then have noted a relative decrease in plateau pressures following proning, likely because the amount of alveolar recruitment has outstripped the limitations placed by chest wall restriction, and these reductions in plateau pressure persisted following supination of patients. The first large-scale trials attempting to demonstrate a benefit from prone positioning in ARDS started in the early 2000s. While no study from 2001 to 2013's landmark PROCEVA trial demonstrated any mortality benefit, there are a number of limitations that are now evident that may have prevented them from a positive result. Some trials did not adhere to lung protective ventilation strategies, while others had much shorter intervals of proning or included patients with milder disease. Some trended toward significance, but may have been underpowered. The PROCEVA trial remains the only randomized controlled trial to demonstrate improvement in mortality and reduction in ventilator days, although subsequent meta-analyses of pooled data to that point did show benefits in the subgroup with moderate to severe disease, as well as the group who underwent proning for more than 12 hours a day. Let's dive a little further into PROCEVA, as it was the only randomized controlled trial to demonstrate mortality benefit, and its criteria are widely considered the gold standard for selecting patients for prone positioning. The study was a multicenter RCT performed at several European hospitals, all of which had some experience with proning patients via a set protocol. The study evaluated 466 patients with severe ARDS, as defined as a PF ratio of less than 150, with a PEEP of at least 5 and an FIO2 of 60%, and randomized them to prone positioning of at least 16 hours a day versus remaining supine. Notably, these patients had to be receiving about 6 cc per kg of ideal body weight in tidal volume, and patients with a MAP of less than 65 were excluded. The study groups were similarly sick, similarly sized, and had roughly the same amount of adjunctive therapies on board at the time of enrollment. Perhaps most importantly, the patients were almost identically hypoxic, with extremely similar respiratory parameters. The study demonstrated not only a survival benefit at 28 and 90 days, but an increase in ventilator-free days and a reduction in ICU length of stay for the prone group, with no significant difference in adverse outcomes, save a higher incidence of cardiac arrest in the supine group. So the PROSIPA trial gave us easy criteria to apply proning to our ARDS patient. If your patient is intubated with ARDS and a PDF of less than 150, PEEP is at least 5 or greater, and the FIO2 is at 60%, consider proning them for 16 hours a day. When you do that, you will swim, which is to turn the head and neck and alternate the arm positions of the patient every few hours. Continue daily prone positioning until either an ABG after 4 hours supine shows a PDF of greater than 150, with your PEEP at 10 or less, or prone positioning simply doesn't help with your oxygenation, as demonstrated by your PF ratio. So with all this robust data, surely we all immediately jumped on this and implemented it in all our ICUs across the globe, right? Unfortunately, change in clinical practice is sometimes easier said than done. The LungSafe study conducted in 2014, which was likely too soon after PROSIPA to truly reflect its data, suggested that only 16% of severe ARDS cases underwent prone positioning. The APRONET study in 2017 showed improvement on this front, with nearly 33% of severe ARDS patients undergoing proning. Interestingly, this prevalent study showed that 10% of patients who did not meet PROSIPA criteria still underwent proning, while only 40% of the patients who met PROSIPA criteria were proned. The most common reasons given for not proning patients were insufficient hypoxemia and hemodynamic instability. However, the only truly absolute contraindication to prone positioning is an unstable spinal fracture. While other relative contraindications exist, two conditions that are commonly cited by clinicians as reason not to prone, pregnancy and obesity, should not be considered barriers to proning. While obese patients may have greater concerns with chest wall compliance while prone, they are often the patients who benefit most from additional lung recruitment in prone positioning. As pregnant women were often excluded for trials on proning, there's a relative paucity of data on its effectiveness. However, recent ACOG guidelines, including visual aids, have been published online illustrating safe proning techniques for patients with gravid uteri, and experts suggest proning both awake and intubated pregnant patients if they would otherwise benefit from it. In the last few years at our institution, we have had multiple pregnant patients survive ICU stays involving proning and carry their infants to full term. Patients with tracheostomies can also be safely proned with the use of special head cushions to allow access to the tracheostomy and inline suctioning. While clinicians, myself included, are hesitant to prone patients who are on high doses of vasopressors, there have been no studies showing worsening incidences of shock, cardiac arrest, or pressure requirements with proning. Additionally, in patients with corpulmonal, recruitment of lung parenchyma can help offload the right ventricle and actually increase cardiac index. However, if the patient's extremely unstable, the challenges of rapidly supinating a patient to perform chest compressions, for example, may be a contraindication. Additional relative contraindications include massive hemoptysis, elevated intracranial pressure, and recent sternotomies. The logistical challenges of prone positioning may be another barrier to widespread adoption. Physicians, nurses, and respiratory therapists must all have training in the process, and a safe proning usually requires between four to six people at the bedside. A number of proning protocols have been published to help streamline the process for ICUs looking to adopt prone positioning as part of their regular practice. Protocols should include thorough checklists for preparation, as well as post-proning checklists for proper application of monitoring equipment and regular swimming intervals to prevent pressure sores from forming on the ventral surfaces of the body. This is just a clip from our protocol down at the University of Cincinnati. Some considerations worth noting that may not be completely obvious until you've proned enough patients. It's helpful to secure all your requisite IV access, central lines, PICC lines, dialysis, catheters, et cetera, prior to proning the patient. Try to batch any ancillary tests, imaging, echocardiography for when the patient is supine so you don't have to flip the patient back early. Because the patients are going to be swum every few hours, they're going to have a hand near the ET tube, and it's important to deeply sedate, if not paralyze, these patients to prevent tube dislodgement, although paralysis is not mandatory. Lastly, while it's appropriate to continue enteral nutrition in a proned patient, it's prudent to hold enteral feeds prior to flipping the patient in either direction. Looking ahead to future directions in prone positioning research. The COVID-19 pandemic brought a massive influx of ARDS patients to hospitals across the globe. Necessity being the mother of invention, reports surfaced early on of the patient's oxygenation improving when self-proning before intubation. Awake-prone positioning has since emerged as a way to improve oxygenation and reduce advanced respiratory care needs for non-intubated patients, including the pregnant and the obese. The emergent nature of the technique deployment meant that studies of the method to date have been wildly variable in their methodology, with disparities in patient location and illness severity, duration of proning, and outcomes investigated. Further research is needed to help clearly define the patient population who might benefit from this technique and its optimal use. As early proning trials had other limitation beyond the clinical severity of their enrolled patients, it may be worth revisiting the possible benefits of extended proning in mild to moderate ARDS patients who have otherwise had lung protective ventilation deployed. Additionally, the optimal duration of prone positioning has not been fully elucidated, nor has when to stop prone positioning. Recent publications have identified the SF ratio as a surrogate value for the PF ratio for hospitals unable to reliably obtain ABG values. If this is widely adopted, further research will need to determine the optimal SF ratios for the initiation of prone positioning as well. Putting it all together, we can revisit our case question and hopefully find our way to the answer that at least I was hoping would be picked as correct, which is C, prone positioning causes a decrease in chest wall compliance and an increase in lung compliance. We don't have to paralyze prone positioning patients, as long as they're sedated enough that they won't pose a danger to their own airway. And the PROSIVA trial suggested that 16 hours of prone positioning per day is beneficial in these patients with ARDS. And I've hopefully convinced you guys that prone positioning should not be avoided in pregnancy. Just a few take-home points, again, we're reducing chest wall compliance and improving lung compliance, ventilation, and perfusion homogeneity. We've been shown to improve outcomes in patients with moderate to severe ARDS when used at 12- to 16-hour intervals. Proning obese and pregnant patients is possible, indicated, and often more beneficial due to increased improvements in respiratory mechanics. Proning protocols implementation in a hospital requires careful planning, checklists, and buy-in from physicians, nurses, and respiratory therapists. And much of our proning protocols are derived from a single study, individual components of which should be studied further, including awake proning protocols. That's all I have for you guys. Thank you. Thank you, Alex. I'd like to introduce Dr. Paige Marty from the Cleveland Clinic, who's going to speak to us on driving pressure and severe ARDS. Thank you, everyone, for being here. This morning, I'll be speaking about driving pressure and severe ARDS. Like Dr. Ouellette introduced me, I'm Paige Marty. I am a new associate staff physician in the Respiratory Institute at Cleveland Clinic in Ohio, and I have no disclosures today. The objectives for today's talk will be to review current recommendations for ventilator management and ARDS, define the concept of driving pressure, and review how it's calculated. Dr. Ouellette spoke a lot about this earlier, but I'll also be reviewing some evidence of PEEP titration and its effect on respiratory mechanics, including driving pressure, and then describe day-to-day implementation of driving pressure in patients with ARDS. So getting into a bit of background, as we all know, patients with ARDS have decreased respiratory system compliance. Traditional recommendations that we all know for lung protective mechanical ventilation include targeting a low plateau pressure, a low tidal volume, and typically a higher PEEP. In severe ARDS, the amount of normally aerated lung tissue is markedly diminished, and this has been termed the baby lung, and this physiology is due to a decrease in lung compliance. Mechanical power is a topic of recent discussion in the hopes to help us better understand how static and dynamic parameters affect ventilator-induced lung injury. Mechanical power aims to describe how the transfer of energy can affect the lung parenchyma itself in ARDS. In an actual patient, this is represented by the transfer of tidal volume per unit time. And the equation to calculate this is shown here, where respiratory rate is multiplied by work, which is the energy delivered to the lungs in a single respiratory cycle, which is then equal to the integral of airway pressure with respect to change in volume. And understandably, this can be quite difficult to calculate clinically, so fortunately, driving pressure is a bit more simple. Ventilator-induced lung injury is an important concept in ARDS management. This is related to alveolar overdistension and instability, resulting in further lung inflammation. This can manifest clinically as pulmonary edema or pneumothorax. There are a few important mechanisms to consider here, one of which is global strain. This is a response to an applied stress, or in other words, tidal volume, which is dynamic, and PEEP, which is static. The inspiratory capacity refers to the volume constraint of the lung itself. And finally, energy load refers to PEEP, as well as the dynamic cyclical nature of driving pressure itself. The successive energy and power can lead to ventilator-induced lung injury. As mentioned earlier, the 2023 ESICM practice guidelines came out earlier this year for intubated patients with ARDS, reflecting the things that we've known about previously, including using low tidal volume ventilation, do not use high-pressure recruitment maneuvers, use of the prone position, not routinely using continuous neuromuscular blockade if not necessary, and referral for ECMO when patients meet criteria. Prone position and low tidal volume ventilation have been shown to improve mortality. Other recommendations, as seen as this figure here, not from these guidelines, but from a different publication, also review a couple of other rules that we can consider in caring for ventilated patients with ARDS. In looking at rule number five, considering personalized PEEP strategies, and rule number three recommends looking at a driving pressure, a delta P, of less than 15 centimeters of water. The authors of these guidelines, however, felt there was not yet enough evidence for specific recommendations on driving pressure in ARDS, but recent studies have shown that it is feasible to study driving pressure in ARDS. So what is driving pressure? Driving pressure, or delta P, describes the relationship of tidal volume to lung compliance in patients with ARDS. This can be calculated by dividing the tidal volume by compliance, or a quick bedside calculation on the ventilator of plateau pressure minus PEEP. These calculations are for patients not making their own respiratory efforts, and is overall felt to represent the stress on the lung. This paper has come up multiple times today, but Amato and colleagues in 2015 performed a multilevel mediation analysis of ventilator variables for over 3,500 patients previously rolled in randomized control trials. They found that among ventilator variables, increase in driving pressure was associated with increased mortality, and best stratified the risk of death at day 60. The authors also tested other ventilator variables, including plateau pressure, PEEP, tidal volume, and while these ventilation variables, in addition to compliance, are related to driving pressure itself, independently they did not provide the predictive information that driving pressure was able to in their survival model seen here. The figure demonstrates the relative risk of death in relation to driving pressure, noting an increased risk of death at values greater than 15, noting where that recommendation earlier came from. The authors adjusted for age, risk of death according to the patient's APACHE score, and P to F ratios, which we know are also prognostic indicators in ARDS, but even adjusting for this driving pressure was still significantly related to mortality at higher levels. So the next few slides will review PEEP titration strategies, which was mentioned earlier. I'll go into a bit of detail about just a few of the studies that were mentioned earlier. Like was mentioned in the guidelines, there is a lack of specific evidence regarding driving pressure itself to make a recommendation, but the overall concept of driving pressure is related to lung compliance, and so titrating PEEP to a lung compliance strategy is something that's been studied throughout the years. So in 2008, the AP VENT1 trial was completed. In this trial, there were 61 patients. This was PEEP titration based on esophageal pressure monitoring to estimate transpulmonary pressures versus the PEEP FIO2 table that we're all familiar with. The patients in the mechanics-driven strategy showed significant improvements in oxygenation and compliance, and there was a possible trend towards improvement in mortality in this group as well, but it was not statistically significant. That can be seen in these figures here, showing in the esophageal pressure group, elevated P to F ratio, respiratory system compliance, noting at 28 days, there was a trend towards mortality but didn't quite meet significance. These two groups here had different driving pressures compared to AP VENT2. In 2013, another study was completed by Pintado and colleagues. In this study, 70 patients had PEEP titrated based on compliance versus, again, the PEEP FIO2 table. In this study, PEEP was titrated to achieve the best compliance, defined as tidal volume divided by the pressure difference at the end of an inspiratory hold. The PEEP was adjusted by two until the best compliance was achieved, and the only statistically significant outcome is the study was an improvement in multi-organ dysfunction at 28 days in the compliance-driven PEEP titration strategy group. There was a trend towards improved oxygenation and compliance in the compliance-driven group, but there was ultimately no specific changes in mortality, similar to the previous study. The ART trial in 2017, which was reviewed earlier, was an RCT of over 1,000 patients, about half of whom were in a compliance-driven PEEP strategy versus the low PEEP FIO2 table strategy. And as Dr. Ouellette mentioned earlier, this study was different from the other ones, as the authors used a recruitment maneuver in the intervention group, starting at 25 of PEEP for one minute, and this was ultimately increased stepwise up to 45 of PEEP for two minutes. Then the PEEP was systematically decreased to find the best compliance. After the optimal PEEP was decided, a second recruitment maneuver was completed at 35. After about half of the patients were enrolled, three patients experienced cardiac arrest felt to be related to these high-pressure recruitment maneuvers, and the maximum PEEP for the recruitment maneuver was reduced from 45 to 35. The results of this study showed an increase in overall mortality. This was a figure that was seen earlier, as well as an increased number of pneumothorax felt to be related to these high-pressure recruitment maneuvers. And finally, the AP VENT2 trial took place in 2019, and this was a study of 200 patients, the Baylor colleagues randomized patients with moderate to severe ARDS to receive esophageal pressure-guided PEEP to maintain end-expiratory pulmonary pressure between 0 to 6. And similar to the other trials, a PEEP-FAO2 table was used to guide PEEP in the control group. The driving pressure in both groups was similar. It was around 13 in both groups, so no significant differences there. And ultimately, the transpulmonary pressures were similar between the two groups, with no differences in mortality at day 28. So how does driving pressure relate to lung stress? This was studied in 2016 with Chumulo and colleagues. 150 patients were studied here, most of them with mild to moderate ARDS, but there was a small percentage that had technically severe ARDS. And these authors studied lung stress and driving pressure at two different values of PEEP, 5 centimeters and 15 centimeters. And these were divided into groups of patients who experienced less than a 15 of driving pressure or greater than or equal to 15 of driving pressure. Lung stress was measured by the difference in airway plateau pressure and esophageal plateau pressure and was considered elevated at values of 24 to 26. As we can see in these figures here, the higher the driving pressure, the higher the lung stress. So we can see this result here in the acute phase of how elevated driving pressure can increase lung stress. But what do we know about long-term outcomes of driving pressure use? This was studied in 2018. Telfin Jr. and colleagues studied patients who had ARDS one and six months after ARDS. This was a small trial of 33 patients. Looking back on their initial admission for ARDS, the patients were divided into high and low driving pressure subsets, defined as 13 or higher in the high group and less than 13 in the low group. At one and six months, patients had a worsened FVC on pulmonary function testing. If they were in the higher driving pressure group at six months, the patients also underwent a high-resolution chest CT, which showed a greater amount of poorly aerated lung parenchyma in those with the higher driving pressure during their admission, suggesting that there could be some possible long-term outcomes related to increased lung stress in the acute setting. So how do we implement driving pressure on a day-to-day basis? In one prospective study, practitioners had to adjust their tidal volume in most of patients over 90% to achieve a driving pressure of 12 to 14. The question has come up, should tidal volume be adjusted to respiratory mechanics as opposed to predicted body weight, as this may provide a better reduction in mechanical power? But there are some limitations to driving pressure, which can be seen with lots of aspects of ARDS. It's uncertain how this applies to patients spontaneously breathing, and it's not a one-size-fits-all as with most things, unfortunately. We have to consider patients' lung sizes, chest wall elastance and body habitus, and overall homogeneity of their ARDS. So to summarize, driving pressure is calculated by plateau pressure subtracted from PEEP, their tidal volume over compliance. There are no current guidelines for specific recommendations for driving pressure value in ARDS, but there is evidence that this is related to mortality at higher levels. Lungs-driven PEEP adjustments are an alternative strategy to the PEEP-FIO2 table, and we should always consider tidal volume in relation to a patient's respiratory mechanics and not just necessarily their predicted body weight. And finally, we know that elevated driving pressure increases lung stress in ARDS, and this may present long-term functional consequences for our patients. Thank you. Thank you very much, Paige. I'd like to introduce Dr. Priya Balakrishnan from Cleveland Clinic, who's going to talk to us about ECMO. Thank you, Dr. Ouellette, for the presentation. So we're going to be switching gears a little bit, talking about ECMO in the last part of our talk today. So I'm Priya Balakrishnan. I'm from Cleveland Clinic, and I have no disclosures relevant to the talk. So this is really just a concise review of the role of ECMO in the management of severe ARDS, primarily focusing on patient selection, evidence, as well as limitations of ECMO therapy. So final reminder about audience response questions. All right. So I think all of us have seen a variation of this case in our daily practice. We'll start with the clinical scenario just to kickstart our discussion today. So this is a young, healthy lady with severe illness due to a common infectious agent. She's on her second day of mechanical ventilation. We've tried our best to maintain oxygenation and ventilation with lung protective strategies, deep sedation, as well as neuromuscular blockade, and restrictive fluid strategies. Despite all these efforts, she's still hypercapnic, hypoxic, and at worrisome limits of mechanical ventilation with a peak pressure of 45 and plateau of 37. Her chest X-ray shows diffused bilateral asthma, so past these, there's no clinical evidence of cardiogenic edema. So should we start ECMO now? This is a polling question. And as you all know, there's no right or wrong answers here. All right. Okay. So let's talk about it a little bit more. So a lot of us have discussed the guidelines by ESICM. It was published earlier this year, or rather last year, and they really tried to give us a solution to the problem that we see on a day-to-day basis. The panel basically recommended the use of ECMO at ECMO centers in patients with severe ARDS meeting the eligibility criteria outlined in the EOLIA trial. So we kind of need to take a deeper dive here and examine the patient selection at EOLIA trial to identify the patient best suited for ECMO therapy in severe ARDS. In general, the patients must fulfill the Berlin ARDS criteria and are early in their course of acute lung injury, which is basically less than seven days. There must be an attempt to optimize ventilator ARDS management strategies, namely lung protective ventilation, be on at least 80% FiO2, and a positive and expiratory pressure of 10. It is noteworthy that patients in the EOLIA trial underwent neuromuscular brocade and proning measures before randomization. Refractory hypoxia in the EOLIA trial was defined as a sustained PF ratio of less than 50 for three hours, or less than 80 for six hours despite maximal therapy. And hypercapnia was defined as acidosis for at least six hours with a respiratory rate of 35 breaths per minute. So the green box here lists out the EOLIA's trial exclusion criteria. Patients with more than one organ failure or poor functionality prior to current acute lung injury were selected out of EOLIA trial. In practice, conditions like BMI more than 55, which is actually becoming more prevalent and with more studies highlighting the obesity paradox of survival in ARDS and ECMO. And finally, duration of mechanical ventilation of more than seven days are becoming more common. So anyone that has a contraindication to anticoagulation and at a higher risk for bleeding should be carefully evaluated prior to ECMO because of the use of anticoagulants to protect circuit integrity. So bearing this information in mind, let's just go back to the case and work through the pros and cons of ECMO initiation on our patient. So does she have hypoxia and hypercapnia? I think all of us would agree based on her blood gas results. Is she less than seven days in mechanical ventilation? Yes, that's true. Does she have a reversible cause of respiratory failure? Yes, because it was just influenza and she didn't have any lung injury prior to this. Is there lack of contraindications? Some might say her BMI, but I think more ECMO centers are getting more used to the idea of managing patients with a high BMI level. So have we optimized medical management for ARDS? Not really, because we haven't tried proning. No one has discussed proning yet for this patient on her second day of mechanical ventilation. So is it futile to actually attempt proning as well as ECMO in this patient? I don't think so because she has really very minimal past medical history in today's setting. So I think then it is important to consider what's the probability of survival for this patient if we were to start ECMO? Will this patient get well enough to leave the hospital alive if ECMO was started now? The REST score was developed using data from the Extracorporeal Life Support Organization, or ELSO, based on an international registry over a 12-year period, including about 2,500 patients. They did a multivariable logistic regression to create the respiratory ECMO survival prediction score, so REST score. The receiver operating curve analysis was a 7.4 with external ventilation exhibiting excellent discrimination. So it's a good score to use at the bedside on a daily basis when we see this kind of patient. Based on our patient that we were discussing, her REST score would be 4 with a risk class of 2. In-hospital mortality for this patient would be 76%. So the higher your REST score, the better the survival. The REST score is not included in any guidelines as of now. However, it can be utilized as a decision tool when considering ECMO as a rescue therapy for patients. So now let's just stop a minute and try to think about whether or not we are actually providing optimal care for ARDS patients who are rapidly worsening in front of us. What truly happens in the real world? So Kadir and colleagues examined this question in 2021, and they looked at 29 U.S. ECMO centers over one year in 2016. They examined about 2,500 patients, and they noted that lung protective ventilation strategies was only used in 31% of cases. However, lung protective ventilation was associated with a lower overall mortality. The pattern of adjunctive therapies used mirrors international cohorts, where about 57% of patients received adjunctive therapies. Notice the higher rate of systemic steroids and neuromuscular blockade use, but lower rates on proning, despite being the most clearly beneficial therapy for moderate to severe ARDS. ECMO is the third most common adjunctive therapy used, and in about 16 of 110 patients, it was the only adjunctive therapy used. In 43% of cases, it was the first therapy utilized, taking precedence over therapies like prone positioning. So is ECMO truly beneficial in patients with severe ARDS? To answer this question, we're reliant on data coming from two randomized control trials, the first being the SESA trial. I think it is crucial to keep in mind here that this study actually looked at the survival of severe ARDS patients managed in ECMO centers versus non-ECMO centers. It made headlines when it came out with a relative risk reduction of 0.69 in patients who were treated in ECMO centers. However, up to 20% of this patient transferred to ECMO centers actually did not get ECMO. Other criticisms of this study include lack of protocolization for ventilator management. Lung protective ventilation was only used in 93% of cases in the interventional group compared to 70% in the control group. There was also a high dropout rate affecting the intention to treat analysis. Switching to the EOLIA trial, so this study was terminated early, as most of us know, because of the predefined futility rule. The study was designed to look at absolute reduction mortality of 20% favoring ECMO. Despite the survival benefit over conventional therapy, it was not numerically high enough. There was about 28 patients that were crossover from the control to the intervention group, again muddying the waters on the intention to treat analysis. I guess this highlights the important question of whether the absolute risk reduction of 20% being overly optimistic. In the SESA trial, it was only 16%. The post hoc Bayesian analysis tried to answer some of the statistical quandaries brought up after the publication of the EOLIA trial. The Bayesian approach is unique because it actually tries to make a direct estimate of survival benefit occurring. It utilizes prior data and beliefs in addition to new data to offer an estimate of a future outcome. Caution is advised if the prior data is subjective. The Bayesian outcome is unlike most studies that we see a binary outcome, whether it's only a positive or negative answer. So the post hoc Bayesian analysis of the EOLIA data revealed there's a 92% chance of absolute risk reduction in mortality with ECMO of more than 2%. So ECMO does offer some mortality benefit in severe ARDS, but there's a substantial variability on the large benefit. So then it begs the question, what is the minimum clinically important difference to establish ECMO's mortality benefit in the management of severe ARDS? So like all therapies, we have to weigh the risk and benefits. ECMO weak spot lies in its limited availability, high operating costs, and high complication rate. So this table lists their commoner ECMO complications, but this is not an exhaustive list. So putting it all together for our patient, I think it's important to facilitate discussion with caregivers as well as family that ECMO may improve outcomes in severe ARDS when conventional treatment fails. Bearing in mind that ECMO benefit is more pronounced earlier, i.e. less than seven days, it is of utmost importance to rapidly implement and sustain conventional therapies that have a proven track record of benefit. This include lung protective ventilation as well as proning. Patient selection is crucial. The EOLA trial eligibility criteria and the utilization of REST score can aid in this determination on a day-to-day basis. So in the interest of time, I'm going to finish off this portion of the presentation by discussing the ESICM 2022 guideline on the utilization of extracorporeal CO2 removal. While in theory it might provide lung REST by implementation of extra low tidal volume ventilation, it is still insufficient to generate a mortality benefit as of now. So this concludes the ECMO in severe ARDS portion of the presentation. Thank you, everyone. Don't forget to evaluate us.

severe ARDS

positive end-expiratory pressure

PEEP

prone positioning

extracorporeal membrane oxygenation

ECMO

lung compliance

stress index

individualized management