Hello and welcome to the CHESS Board Review Course session on the Acute Respiratory Distress Syndrome. My name is Jason Poston. I'm an Associate Professor at the University of Chicago. I'll start by saying I have no conflict of interest to disclose relevant to this particular disease process. Our learning objectives for this session are to review how to diagnose ARDS using the Berlin definition, to recognize the risk factors for patients to develop ARDS, understand the physiologic principles that guide the respiratory support strategies, including listing those strategies for mechanical ventilation and for non-invasive respiratory support in ARDS. And then finally, in our final part, we will identify and review the literature for ancillary or non-ventilatory strategies that we use to improve ARDS outcomes. As I said, our first segment here, we'll talk about the definition and risk factors for ARDS. So it's important to recognize, and as it's highlighted on this slide, that the acute respiratory distress syndrome is indeed a syndrome, and there are many insults that result in this final common pathway that leads to the physiology and symptom complex that we know as ARDS. So if you'll remember, they sort of flow in and flow out of fluid into the alveolar space. ARDS is often a result of a injury to the capillaries, which leads to capillary leak, which allows protein to leave the vascular space and interstitial and alveolar flooding. Which then subsequently leads to what we see clinically, gas exchange abnormalities, decreased compliance or stiff lungs, and oftentimes an elevated pulmonary vascular resistance and perhaps pulmonary hypertension due to hypoxic vasoconstriction from the hypoxemic units. And these are the clinical entities that we struggle with when we're caring for patients with ARDS. ARDS is incredibly common, and this article published in 2016 really highlighted just how common, not popular, common ARDS is. And it really sort of underscored that we probably miss a lot of ARDS because it's much more common than we think. So this unrecognized ARDS is really highlighted by the fact that only about 51.3% of patients who met the criteria for ARDS that was mild, it was recognized by their clinical team. And even people with severe ARDS, that word did not appear anywhere in their chart, and the team did not recognize in 21.5% of cases, a patient that they were caring for who had severe ARDS. So it's really, when we talk about in this setting, ARDS, we know it's a severe respiratory illness. We know it's attributable morbidity and mortality. And we might say to ourselves, well, how could anyone ever miss it? But the reality is it's missed all the time. And so a vigilance and being deliberate about looking for ARDS is critically important because there are therapeutic changes that you can make to either treat ARDS, or as we'll talk about in a little bit, to recognize people who are at risk for being one of these 10.4% of admissions that have ARDS so that we can prevent it. So what are those risk factors? Well, as we know, ARDS as a syndrome can be from a list of things, as long as my arm, as they say. But we often will divide that into respiratory and non-respiratory factors, just in thinking through your differential. Pneumonia is one of the most common. Obviously, that would be a respiratory cause. But then all of these things that sort of affect the whole body can cause ARDS. Sepsis, a very common clinical entity often complicated by ARDS. Aspiration into the lungs. Trauma patients are at high risk for ARDS, sometimes because of a direct thoracic injury, such as in a lung contusion, but also just sort of long bone trauma can put people at risk for ARDS. There's an inflammatory response to transfusion of blood products that can cause ARDS. People who undergo organ or hematologic transplantation. There are lots of medications, both ones that we give and ones that patients acquire by their own means, and many other sort of associations that are more rare, but are still critically important to think about when you're assessing a patient for the risk for ARDS. So what is the definition of ARDS? So I'm going to give you a couple of different scenarios, and I want you to sort of think as we review the definition, sort of patient one, patient two, do they have ARDS or not? So this is a 55-year-old woman with endometrial cancer. She's admitted for fever, tenderness over her tunneled central venous catheter for the last three days. She is febrile. She has a heart rate of 110. Her blood pressure is 80 over 40, and her respiratory rate is 18. She is intubated for reduced mental status, believed to be due to her shock state. Her chest auscultation reveals bilateral crackles, but while she's on the ventilator, her saturation is 90% on only 30% oxygen with a positive end expiratory pressure of 5 centimeters of water. So I want you to reflect on that patient, the sort of clinical history, the degree of perturbation of her oxygen exchange, and ask yourself, does this patient have ARDS? Moving to patient two, this 25-year-old man is admitted for fever, cough, and increasing dyspnea over two days. He has a fever. His heart rate is 125. His blood pressure is 125 over 64. His respiratory rate is 26. He's intubated for respiratory distress. His exam reveals right-sided crackles and agaphony, and his PAO2 is 65 millimeters on an FIO2 of 50% on a higher peak than our last patient at 7 centimeters of water. So the question is, which of these patients, or do both of these patients, have ARDS? So it's worthwhile reviewing this definition. It is, the A stands for acute, and so the timing suggests that this needs to be less than one week from a known clinical insult to the new or worsening symptoms. It's a diffuse process, and so on chest imaging, we need to have bilateral, a bilateral process that shows up in both lungs that's not explained by non-ARDS etiology, such as infusion or atelectasis or sort of micronodular disease. The origin of the edema in the chest cannot be fully explained by cardiac or volume overloaded, and so sometimes if there's a risk for cardiac or fluid overload, you may need to adjudicate that through diagnostic testing, but if there's no ARDS risk factor, you don't actually need an objective assessment of beyond your history and physical. And then obviously this is a disease of hypoxemia. The oxygenation severity is often graded by the PAO2 to FIO2 or the P to F ratio of these patients, and one of the parts of this criteria that I think we're grappling with as we move to more non-invasive modes of supporting people is that strictly speaking, the definition requires that a patient is on positive pressure ventilation with a PEEP greater than or equal to five centimeters of water, and if the P to F ratio is between two to three hundred, we would characterize it if all of their criteria met. It's mild ARDS, moderate ARDS for 100 to 200, and less than 100 is severe ARDS. So I've highlighted now on the screen a couple of important things that are honestly quite different from the sort of prior definition of ARDS. One is that we can sort of adjudicate the cardiac fluid overload just with our history and physical. Previously a pulmonary capillary wedge pressure was thought to be necessary to prove that you didn't have high left ventricular and diastolic filling pressures leading to high pressure edema. That's no longer necessary for this diagnosis. The second is the inclusion of PEEP and that someone's on positive pressure ventilation to make sure they've had a trial to overcome, you know, what might be low-level anelectasis. And the final is that all of these P to F ratios, anything less than 300, is deemed ARDS. It used to be ARDS in the old definition, required P to F ratios in the moderate to severe range, and the mild was called acute lung injury. We now just call all of these things different grades of ARDS from mild to moderate to severe. So as we revisit our patients and sort of apply these criteria, patient number one has an acute onset of a bilateral process, is on positive pressure ventilation, has a P to F ratio, despite the fact they're not on very much oxygen, very mild oxygen supplementation. But when you look at their P to F ratio, it's about 200. So this particular patient has ARDS, not terribly hypoxemic. And I think I want to underscore that point. People often think to have ARDS, you have to be someone who's on high FiO2, high PEEP. Really, to meet this definition, you really only need sort of mild to moderate oxygen exchange abnormalities. And even someone who's on FiO2 of 0.3 can certainly meet criteria with an acceptable saturation. We look at patient number two. Well, it's acute. The patient's on positive pressure and has PEEP. Their P to F ratio certainly meets criteria with a P to F ratio of about 130. But as you can see on the x-ray, this is not a bilateral process. This is a focal mnemonic process leading to issues with ventilation perfusion ratio and a shunt lesion. This is probably a community acquired pneumonia. This person does not at this point have ARDS, but I think it's important to recognize that while they don't meet criteria right now, they're at tremendous risk for ARDS. And so we'll talk about this in a bit, but when we think about the ventilator strategies that we use to treat ARDS, since this person got intubated, it's probably worthwhile to give them those same ventilator strategies to prevent the development of ARDS, given the high risk from this person's community acquired. So part of the criteria is this P to F ratio. Well, that requires that you have a blood gas. I think in many places we're recognizing that less is more in critical care. So if we can get away without introducing a needle into the arterial system or having an indwelling arterial catheter, shouldn't we do that? Can we still sort of diagnose and treat ARDS? And this is a nice study from Todd Rice, who basically made that point that we can use the SPI2 to FiO2 ratio, and seeing this sort of translation from the P to F ratio is a 300 and 200, using the S to F ratio is roughly 315 to 232, or 235 rather, we can make this diagnosis and characterize someone's hypoxemia using this currency exchange, so to speak, so that we can provide less invasive care. So what's our differential diagnosis? If someone sort of meets these criteria, it still remains a diagnosis of exclusion, so what do we need to rule out? Well, part of it is right there in the diagnostic criteria. We want to make sure that this isn't just high pressure cardiogenic edema, and that is far and away the thing that's going to be the most difficult to differentiate from ARDS, particularly in someone with cardiovascular risk or known cardiovascular disease. It's important to also recognize that you can have ARDS and heart failure at the same time, and so, you know, similar to someone has community-acquired pneumonia and risk for ARDS, someone who has heart failure, if they need to have positive pressure ventilation, it probably behooves you to give them, as we'll talk about, some of these ARDS-directed and protective strategies as well. These are some of the rarer things that can look like ARDS, so exacerbations of interstitial lung disease, acute interstitial processes, lung hemorrhage, eosinophilic lung disease, and organizing pneumonia, much less common just epidemiologically, and again, their radiographic pattern sometimes helps you differentiate, and so they're less commonly confused for ARDS than cardiogenic edema, but we certainly see these present in very similar ways to ARDS. So as we turn our attention to sort of the next phase, we're going to talk about the respiratory support and positive pressure support for people that have ARDS. So when I think about ARDS, I like to think about it from a physiologic standpoint. It's a really, the critical illness comes from a high shunt fraction, which inhibits the ability to oxygenate the hemoglobin. When you can't oxygenate your hemoglobin, you're at risk for not delivering enough oxygen to individual tissues, and so making sure that you have adequate oxygen delivery, a compensatory mechanism for hypoxemia, is often to increase your cardiac output, and so, you know, having an adequate cardiac filling and emptying and adequate hemoglobin is an important part of the physiologic compensatory response to ARDS. Obviously, oxygen consumption will determine the balance or imbalance of oxygen supply and delivery, and so some of the things that we do, you see there, decrease patient oxygen consumption and help restore that balance, and then, you know, a lot of things that we're doing, they're not fixing, you know, sort of the state we're in, but they're preventing it from getting worse, and so we'll talk about lung protective strategy, judicious transfusion to prevent furthering of that injury, and potentially the prone positioning, which might help with some of the shunt fraction reduction, will also potentially avoid further injury. So what is our basic principle of ventilating someone who has ARDS? Well, we want to recruit alveoli that have been washed out with this edema and are largely atelectatic and collapsed, but we want to do that in a fashion that prevents sort of passing the upper deflection point to the volume pressure curve of the lung, where alveoli are becoming overly distended, and remember, this is a regional disease, and so even normal tidal volumes, if there are densely sort of atelectatic and involved regions, you're going to be ventilating the parts of the lung that are actually in the best shape, but if you ventilate them with high tidal volumes, you're likely to sort of overstretch them past that upper deflection point, and that can potentially lead to further perpetuation of the lung injury, and so to prevent collapse and reopening, which can further injure the lung and also just lead to worsening hypoxemia, on the low end of the tidal volume, we need PEEP. We need to maintain that pressure to open those alveoli, but we know we're ventilating sometimes, you know, only a handful of the alveoli. We want to prevent overdistension of those alveoli, so as we're recruiting more lung with PEEP, we're sort of preventing injury with low tidal volumes, and this was really operationalized in, you know, probably one of the more important studies in pulmonary and critical care. This is the original ARDSNet trial published in 2000, which sort of looked at people in volume assist control ventilation who met criteria for ARDS and said, let's give them what were at the time very low tidal volumes, four to six milliliters per kilogram of ideal body weight, versus what were relatively common tidal volumes at the time, 12 milliliters per kilogram, and that's sort of a starting guess that we thought people needed low tidal volumes, but we want to confirm that we're not overdistending those alveoli. The pressure that is going to drive overdistension is going to be best measured by the plateau pressure on the ventilator, and so in addition to these sort of ventilator strategies, we wanted to, in the low tidal volume strategy, keep the plateau pressure less than 30 centimeters of water, recognizing these low tidal volumes are going to lead to low minute ventilations, are going to lead to hypercapnia, and we're going to allow that, and I think you all know how this story ends. There was a significant decrease in mortality from 40 percent in the control group to 31 percent in the intervention group, and in addition, a secondary outcome, more ventilator-free days, and this has really become, you know, there's lots of name for this, low tidal volume ventilation, lung protective ventilation. This has really become the mainstay and the sine qua non of our ventilator strategy in ARDS. Everything else has been sort of additions or modifications of, you know, this basic principle here, giving PEEP to recruit but making sure that the tidal volumes are low and that the plateau pressures are not prohibitively high. Unfortunately, when we look at how are we doing, this article was published in 2000, the original article was published in 2000 for lung protective strategy, but as you see here in this 2006 sort of survey of critical illness, not only do we not recognize ARDS, even when we recognize ARDS, we're pretty bad at adhering to lung protective strategy, and so this is basically looking at people with mild, moderate, and severe ARDS and plotting their ventilator settings and pressures, sort of plotting them on their plateau pressure on the y-axis and their tidal volume on the x-axis. So it's important to recognize that most people should be down in that bottom left quadrant, right, with tidal volumes that are less than six cc's per kilo and plateau pressures that are less than 30, and even if you give an allowance for people to go up to eight milliliters per kilogram, there's still almost a majority of patients that are not getting lung protective strategy, even 16 years after the seminal paper that we'd like to talk about so much, so we probably need to do better to adhere to and apply the knowledge that we know about ARDS. What about the level of oxygen? Well, we want to avoid hyperoxia, so we tend to try and keep people's oxygen saturation only where it needs to be, so you can look here and see that, you know, percentage point increase in SpO2 as we're sort of driving up the oxygen. Oftentimes that can just lead to an increased relative risk of mortality, and so we want to give enough but not too much oxygen, and, you know, that sort of leads us to, well, if we're not titrating oxygen, what are we using? And this is sort of the first study on the higher versus lower PEEP saga, and unfortunately I don't have an answer that I can give you declaratively. That's the bad news. The good news is it's kind of hard to ask a board question when there's not consensus, but what is the data? So after the low tidal volume keeps the plateau pressure less than 30 study, the next ARDNET study sort of took a look at lower versus higher PEEP. There were some protocol modifications through this, but the punch line here is that when you use a higher PEEP strategy, so this is the step up in between turning the FiO2 up and turning the PEEP up, when you use a higher PEEP strategy, there's no difference in mortality, ventilator-free days, or ICU-free days. There is an improvement in the P to F ratio in the higher PEEP group, and there's an improvement in the respiratory system compliance. So at least in this study, in the early 2000s, there was no difference. This is not the end of the story, and people continue to sort of explore what is the right level of PEEP, and I think that's a fun and active discussion. Unfortunately, we don't have sort of a prescriptive thing for this, and I think you're within acceptable practice to use either the lower or the higher PEEP strategies. Why wouldn't you give higher PEEP? Well, recognize when you increase the PEEP, you increase the intrathoracic pressure, and blood has to enter the thorax to get back to the heart to make sure that you have adequate forward flow. ARDS is often complicated with cardiovascular compromise, either from sepsis or just positive pressure ventilation in the setting of hypovolemia, and so PEEP can potentially complicate matters, and so I think that this is a discussion that we continue to have. So where are we? Well, we know that low tidal volume with PEEP reduces mortality if the plateau pressures are less than 30. The higher PEEP improves oxygenation and compliance, but doesn't necessarily improve mortality, and so if we draw that sort of, you know, graphically here, and we think about the plateau and the PEEP, you know, we were trying to sort of figure out what to do with the PEEP, but we know the plateau pressure is important. What lives between them is called the delta P, sometimes referred to as the driving pressure, and we know that that driving pressure is the product of the compliance of the respiratory system, and so maybe if we, with the ventilator, optimize the compliance by giving the right amount of PEEP and giving the right tidal volume, if we optimize, and by optimize I mean minimize the driving pressure, perhaps that's a ventilator strategy that's going to be associated with the best outcomes, and looking backwards, we have data that suggests that's exactly where we are, and so this paper published in 2015 demonstrated that that delta P, so the tidal volume over the compliance of the respiratory system, were simply calculated as the plateau pressure minus the PEEP. When we sort of use that as a clinical variable and did a multi-level mediation looking at over 3,500 patients from the prior trials, we would then able to isolate and evaluate that particular clinical marker as an independent predictor of survival, and what's more is we will we would be able to sort of say, well, what helps people survive? Is it that they're on high PEEP? Is it that we sort of manage their plateau pressure really well and kept it below 30, or is it that we really optimize the driving pressure here? And when you look at this data, this is sort of complicated, and I would encourage you potentially, if you're watching this recording, to hit pause and think through this a little bit. What this demonstrates is when you take all patients and you put them in sort of categories, but you sort them by, in resampling A, matching their PEEP so that all of the groups, despite the fact that they have different survival, are matched on PEEP, or in resampling B, you match them on their driving pressure, or in resampling C, you match them on their plateau pressure. What you see is that in sample A, when you match people on their PEEP, if people have the same PEEP but different driving pressure, the higher the driving pressure, the higher the relative risk of mortality. So you can sort of compare the top panel to the lower panel in that resampling. If you go all the way to the right, resampling C, and you sort of match people and say, well, what about, you know, everybody who has the same plateau pressure should obviously have the same outcome. Well, it's just not so, because we know if the plateau pressures are the same between groups, but the PEEP is rising from sort of the lower to the higher, you're going to have an increased mortality with groups that have a higher driving pressure. So if you're able to, if you have a lower PEEP, but your plateau pressure is high, it means that your lungs are very, very stiff, and perhaps we could do more with the ventilator, or perhaps it's a measure of the intrinsic lung, but as you can see there, as you go left to right, the mortality goes down, because that's a more PEEP, less driving pressure group. But when you match people on their driving pressure, as you do in sample B, you see that across this group, despite the fact that their plateau pressure goes from 20 to even greater than 30, sort of violating lung protective strategy, as long as their driving pressure is the same, their mortality doesn't change. And so some have argued when you sort of look at this data and showing that an increase in the driving pressure of seven centimeters of water was associated with a mortality relative risk of 1.4, that maybe we can do stuff with the ventilator, maybe we can titrate the PEEP for instance, and answer that PEEP question by titrating the PEEP to optimize the driving pressure. And so I think a lot of people have adopted this strategy, I don't know that it's yet been sort of published, that this sort of prospective management leads to mortality improvements. And honestly, we've done so well with survivorship with ARDS, you know, it's actually hard to demonstrate that. But I think a lot of people like the promise of driving pressure and are using it in that fashion to help with their PEEP titration, while also trying to maintain the goals of obviously lung protective strategy with low tidal volumes and keeping plateau pressures less than 30. What are some things that we don't want to do? Well, we talked about maybe using a little bit higher PEEP. Well, what if we use super high PEEP? What if we use what's called a recruitment maneuver, where we increase the PEEP and with reckless abandon for short intervals, 30 seconds, 60 seconds, we turn the PEEP up very high and recruit alveoli, and then we try to keep those alveoli open with high PEEP. This is a study of about 1,000 patients with moderate to severe ARDS, and we compared the traditional ARDSnet ventilation versus something that involved this protocol that you can see of an acute recruitment maneuver, turning the PEEP up to 25, 30, 35, and maintaining it for some number of minutes, and then resetting the PEEP at a higher level. How would those patients do? With the recruitment in titrated PEEP group, there was an increased mortality, an increased rate of pneumothorax, and a reduced number of ventilator-free days. So the upshot for me here is don't use this protocol. It was sort of a double protocol. It used both recruitment and high PEEP, and so it's unclear which part drives the harm, but generally, this maneuver that I think people often would reach for prior to this study, I think people's enthusiasm for it, given the harm that was demonstrated in this trial, has gone down considerably. What if we don't intubate the patient? And I know many of us will use high-flow nasal cannula devices for acute hypoxemic respiratory failure. So again, this is a study that probably applies to ARDS because it was patients with acute hypoxemic respiratory failure, and as we showed, there's a lot of people with ARDS. P to F ratio, less than 300 here. They were randomized in this particular trial and published in 2015. So high-flow nasal cannula versus face mask versus non-invasive, and then they looked at re-intubation, or intubation rates for all patients, which were unchanged. However, high-flow nasal cannula was associated with more ventilator-free days and a reduced risk of death at 90 days with a hazard ratio of two versus a face mask, a two and a half for non-invasive ventilation. When you look at the need for intubation, when patients were more severely ill, high-flow nasal cannula reduced the rate of intubation, and we know that intubation is a bad thing for so many reasons that you'll hear about in this course material, neuromuscular weakness, neurocognitive perturbations, and so keeping people off the ventilator is really clinically important for patients, and this particular management strategy was shown to be beneficial. Now again, this is not just ARDS patients. This is hypoxemia from all causes, so you have to take this with a grain of salt, but I think people with mild oxygen perturbations, certainly during the COVID pandemic, but even in other patients, many people have adopted the use of high-flow nasal cannula systems and heat-humidified high flows of oxygen to manage this pulse. And then finally, all these things we do with the ventilator. What happens if someone has risk for ARDS but doesn't in fact have ARDS, and there's emerging evidence that we should probably be using these ventilator strategies for ARDS patients and using low tidal volumes, even when people are intubated for upper airway lesions or for mental status changes sort of coming out of a procedure, or honestly, even just going to the operating room for an elective procedure, and you can see a nice meta-analysis looking at half a dozen or so studies here showing that even patients that don't have ARDS, there's a risk for injurious ventilation, low tidal volumes as the rule rather than the exception only to be used for patients with ARDS. So we're going to leave the ventilators and the knobs on the bedside positive pressure device, and we're going to talk about the other things that we can do to help people with ARDS. So obviously the ventilator strategy is critically important but there are a lot of other things, and there's been a lot of investigation that help us manage the patient more globally. So if we sort of review our overall physiologic strategies, we've talked about what do we do with the ventilator? Well, the tidal volumes of PEEP, the mechanical ventilation to decrease oxygen consumption probably helps, and the lung protective strategy avoids further injury. But we're going to go on a quick whirlwind through thinking about fluid balance, positional therapy, inhaled vasodilators, the role of neuromuscular blockade, which could potentially decrease oxygen consumption and make people more synchronous with the ventilator, and then we're going to talk about should we just tank people up with hemoglobin. So first and foremost, we'll review fluids, probably one of the most important things we do in the ICU. We spend a lot of time, are they wet? Are they dry? Do we give them fluids? Do we give them diuretics? And so in this particular protocol, again by the ARDS NET group, after they did tidal volume and plateau pressures, they went on to high PEEP versus low PEEP, and then they sort of looked at fluid status. And I always say when I'm working with learners, the only thing worse than low-pressure pulmonary edema, i.e. ARDS, is low-pressure edema being driven by high pressure. And so you have that sort of alveolar leak, but if there's high hydrostatic pressures in your pulmonary vasculature, obviously you're going to have more lung water, more hypoxemia. And that was the sort of thinking behind this, where they took patients that had ARDS and randomized them to receive what was referred to as a conservative or sort of fluid-sparing therapy or a liberal fluid strategy. And it's important to recognize that people with ARDS, because they often have sepsis, they often have an infection, and honestly, just because they're in the ICU, they tend to accumulate fluid. And so this was a very strict protocol where basically as soon as someone was out of shock, you would start diuresis. What did they actually achieve? Well, the people that you really slammed with diuretics were very aggressive. The intervention arm here, they were only 136 milliliters negative at the end of this sort of intervention. However, the liberal fluid strategy, the routine care group was up seven liters. So there's an appreciable difference in whole body volume, both salt and water, given the fluids that we're typically talking about in these two groups. What were the results? Well, if you gave people aggressive diuretics, you did not save lives. The mortality in 60 days was unchanged. However, you did do some very important things for functional recovery. You improved oxygenation, which often led to people being able to come off the ventilator. And on average, they were off the ventilator three days earlier than the patients who were allowed to have a liberal accumulation of fluid. And their ICU days were similarly decreased by, as you can see there, just shy of 20%. Again, we would love to see a mortality benefit, but we also know that the functional status of someone who's leaving the ICU is critically important for their long-term recovery. That's become the gold standard of what's a good intervention in the ICU. And I think by that metric, putting people in negative fluid balance as soon as they're out of shock to improve their oxygenation, get them off the ventilator, have them sort of move around and leave the ICU is a really important patient-focused end point. You can see those here. Well, sometimes people always say, well, I know the physiology, and I know that you carry oxygen on hemoglobin. Let's just give people more hemoglobin. That seems like a bad idea. We've seen in a prospective cohort study that the more packed red blood cells people get, the more likely they are to develop ARDS, and there's a dose-dependent increased mortality. And then we've actually randomized people to that. So this landmark trial in 1999, randomized controlled trial of liberal versus restrictive transfusion strategy, where we were sort of targeting 10 versus seven. The restrictive strategy was at least effective, but it improved hospital mortality and improved mortality, particularly in less ill patients because it allowed them to stay less ill. And so we generally have more restrictive transfusion policies because the sort of biology, biochemistry, and immunology of blood transfusion seems to win the day above and beyond just the oxygen-carrying capacity part of having hemoglobin. And so we don't tank people up with red blood cells despite their hypoxemia to increase their oxygen-delivered tissues because of these particular data. What about neuromuscular blockade? How did this slide get in here? This is an old slide. And so this was the thinking for a while that neuromuscular blockade improved survival in patients based on a 2010 study of 340 patients with ARDS that showed an improved 90-day survival. However, there was probably a bigger and more contemporaneous study, and this was not able to be validated. And so we looked at this study published in 2019 of 1,000 patients with ARDS that had severe ARDS, so a P to F ratio less than 150, on higher levels of PEEP, so the kind of patients that typically might reach for neuromuscular blockade. We compared usual care with light sedation to paralysis, which in turn required deep sedation. And as you can see here from this funnel plot, when you look at survivorship, when you look at discharge, there's sort of no change between these two groups. And so we tend not to reach for neuromuscular blockade in patients who have ARDS just for the purpose of sort of promoting survival because it wasn't seen in this very large, very well-done brand-device-controlled trial. What about prone positioning? This has been shown to improve survival in some initial trials and in a validation trial. So this is the sort of best trial that we have, 466 patients, again, with P to F ratio less than 150. To be eligible for this study, the patients were actually managed not in the prone position and allowed to sort of stabilize and sort of settle out with regard to their hypoxemia for between 12 to 24 hours. And I think that actually removed a lot of the patients that, for instance, were gonna get better anyway because they had a component of high-pressure edema or something was sort of transient. And so these were people with pretty convincing real deal, severe ARDS. And then the protocol, I think importantly, didn't just sort of prone people and then as soon as they weren't hypoxemic, move them supine again. The prone positioning was not just rescued from hypoxemia, it was therapeutic. It helped apply lung protective ventilation and prevent injurious ventilation. So the proning was maintained for at least 16 hours. Importantly, this was done with regular hospital beds, with regular nurses, with a protocol that can be reproduced in ICUs around the world. And the results were pretty remarkable for people with this low P to F ratio with ARDS. The reduced 28-day mortality from 33% to 16% has really made this a mainstay to be considered in patients who have ARDS with very low P to F ratios. I think if ever I could sort of introduce a controversial topic with a non-committal title, it would be this, which is steroids may improve survival in some ARDS. Steroids were sort of a dealer's choice prior to a couple of years ago with some data that showed in not very good studies that they helped in people with ARDS and other data that said, particularly in some subsets of people with ARDS, including influenza, that they could be harmful. And we've learned a lot about this over the last handful of years. So the recovery trial in 2021 during the early days of the COVID pandemic demonstrated the utility of corticosteroids in COVID patients who had ARDS. And you can see the data here showing meaningful improvements in the rates of progression to invasive mechanical ventilation in patients who are invasive mechanically ventilated in terms of their mortality and in people that were receiving oxygen. Notably, there was no improvement in people who weren't requiring oxygen upon admission to the hospital. So we see that mortality benefit there. And you can see, you know, there've been several studies, meta-analysis, and again, the World Health Organization published this in JAMA demonstrating improvement. And so we tend to give corticosteroids in people with this, you know, relatively recently, the most common cause of ARDS that we've seen in our ICUs. But steroids may improve survival in other causes of ARDS as well. And so this was a group of patients who had community-acquired pneumonia. They had a P to F ratio that was very low. It's unclear whether these people had, you know, sort of bilateral, sort of diffuse disease, meeting criteria for ARDS. So again, a little bit of a grain of salt, and you'll probably talk a little bit more about this in this session when we talk about community-acquired pneumonia. But with a recent study published in 2023 for community-acquired pneumonia, continuous infusion of hydrocortisone for at least four and up to eight days with a taper thereafter, improved survivorship in the steroid group compared to usual care. Had a pretty impressive clip there from 11.1% in the control group to 6.2 in the intervention group. And so again, this may be a subset of patients, a pretty epidemiologically, a large portion of patients who have ARDS historically have had community-acquired pneumonia, and they seem to benefit from this. And it doesn't seem to necessarily harm them if they have ARDS as well. And so this would be probably another piece of data that would push us towards giving more routine steroids in patients who have ARDS from sort of global causes. So I think steroids are likely to benefit in early, based on the data, in early moderate to severe ARDS from COVID-19, certainly in the setting of community-acquired pneumonia. If there's a known steroid responsive or sensitive process, so it's a pneumonia with ARDS, but you think there's a component of organizing pneumonia, obviously you're gonna give steroids. And it's okay to give steroids in ARDS, and it may be beneficial, even if you're giving them for another reason, like refractory shock. When do we get nervous? And where is there data that says we probably should try to avoid steroids in ARDS if there's no non-ARDS indications that sort of win the day? One is in influenza, where there's some data that says it might be harmful. And the other is once someone's progressed to sort of fiber proliferative ARDS. We really don't wanna give steroids because there are some data that demonstrate that that can potentially do harm. And so again, we wanna give it early for these indications, avoid it late, and in the setting of an influenza infection. Although I'm sure the next time we have a big influenza season, we will try to revisit the influenza question, given what we've learned about COVID and some of the other things. For now, the best data says best to avoid in influenza. The other group that I think there's less data is in fungal infections. If you think that there's an inflammatory response due to a fungal pneumonia, again, it comes down to sort of clinical decision-making at the bedside, but there's not nearly the strong data that we have for COVID and bacterial community-acquired infections. What are some other things to consider? Well, sometimes we do things because our back's against the wall, even though it might not, in a data-based way, lead to survivorship. And so inhaled pulmonary vasodilators, you shouldn't, I don't think, necessarily reach for them before you prone people, before you do good ventilator strategy, before you give steroids in the right clinical setting. They will improve oxygenation in the subset of patients, so it's not really a patient-centered outcome. And they've failed to show a meaningful benefit from a patient-centered clinical outcome, like mortality, like duration of ventilation. But we will often reach for them, again, when our back's against the wall and we just can't oxygenate a particular patient. Sometimes people will also reach for them when you get the impression that cardiac forward flow is inhibited by pulmonary hypertension due to the ARDS and the hypoxic vasoconstriction, because obviously they'll improve oxygenation, but they may also unburden the right heart. That's a right heart failure from ARDSs, again, when people might reach for it, although they will be doing it without a robust evidence base that shows that it improves meaningful clinical outcomes. What do we not do that over the course of your career, you may have read articles about, and there's an association there, and so they may trick you on a board exam? Well, we tend not to give surfactant. There's a trend to better oxygenation, but no clinical benefit, and it's very messy. There's mixed data, but nothing compelling. Dietary antioxidants never really caught on. Albuterol was thought to perhaps be helpful, but there was no efficacy. And of course, if you're administering that through the IV, there's potential cardiovascular harm. High-frequency osculatory ventilation, long used in pediatrics. There's some data based on the primary analysis of a large trial that it may increase mortality. People continue to debate that, and people continue to advocate for finding a utility in that therapy if it's delivered in the right patient in the right way. But as a blanket statement right now, it's not generally something that is common practice or recommended practice in a typical ARDS patients. And then it's such a big disease process. We've tried everything. We've tried statins, NAC, NSAIDs, activated protein C, glutamine, NEA, ketoconazole. All of these things have not shown benefit. They just really take your eye off the ball of sound ventilator strategy, a reduction in the circulating volume, a careful titration of PEEP, proning if necessary. Those are the things that are really gonna improve our outcomes, and honestly have improved our outcomes and decreased morbidity and mortality from ARDS significantly over the last 20 years. So what about ECMO? So here's a story, and I know this is, if your shop is any like mine, it's a debate that we have over and over and over again. So what do we know about sort of rescuing someone, oxygenating them with a machine, allowing lung recovery with an ECMO circuit? So in the early years, there were a couple of studies presented here where we transferred patients who had severe ARDS to an ECMO center versus leaving them in place. In both of these studies, ECMO improved patient outcomes, improving survivor, and they flipped it and called it decreased mortality in the other study. But in both of these studies, only a subset of patients that actually got transferred got ECMO. So the intervention was probably going to a place that can do ECMO, which tends to be high volume, high performing academic centers that are gonna do a lot of things with more resources and probably be able to pull them off, proning, you know, lung protective strategy with aggressive respiratory therapy, physicians and nurses to make sure that those things are implemented at the bedside. And so for that reason, I think people were unsatisfied with this as a sort of guiding a couple of studies for ECMO therapy. And so we did a large randomized control trial, and this was the AOLIA trial published in 2018. These patients were ones that I think it made a lot of sense. These are the people I would think about ECMO with. These are patients who had a P to F ratio that you just could not get above 30 for three hours or above 80 for six hours, or had really prohibitively low pH and high CO2 with the lung protective strategy. These patients were all sort of encouraged and almost universally had neuromuscular blockade and prone positioning. Other adjunctive therapies were allowed and were intermittently sort of utilized, but these are the patients that in my practice, I would exhaust all of the other things and I would say, well, maybe I should do ECMO. And that's where this study was done. The intervention was percutaneous VV ECMO with heparin administration, and then an ultra lung protective ventilation to keep the plateau pressure less than 25 versus the control group where it was a typical ARDS strategy with higher P targeting a plateau pressure of 28. And importantly, there was some crossover allowed. So again, I think we probably everybody here has at least got wind of how this story ends. This is a great trial. These patient populations look very, very similar as you can see here, and they were very, very, very sick. You'd look at those P to F ratios and they're incredibly low. And the results are that this trial was stopped early for meeting the futility criteria by the Data Safety Monitoring Board. So they had about 125 patients in each group. There was a crossover two ECMO and 35 patients. When you looked at the mortality curve, it was 35% in the ECMO group versus 46 in the traditional group for a P value of 0.07. But that was deemed to be with the remaining enrollments in the power calculations, an insufficient sort of progress towards a potentially positive result. And so this was deemed a negative study. So there was some treatment failure and some death. And so some people have then done sort of calculations and demonstrated that the ECMO hazard ratio for death was 0.62, but obviously wasn't nearly as satiating as everyone thought it would be when we're doing a large randomized control trial. So a lot of people like to say, oh, a trial, there's something for everyone. And if you go to the Church of ECMO, you can say, look at these trends and weren't they silly to stop the study early? And you can make some very reasoned arguments about why that's true. And if you don't go to the Church of ECMO, you can say, well, as I suspected, there was a little bit of improvement, but it's probably not worth the fuss and the trouble. And I think there's some data here to say that it's not the panacea that some people want it to be. There have been subsequent meta-analysis that have taken these patients, the ones that actually got ECMO in the earlier trials that have shown a benefit. And I think people have advocated for that. But I think particularly when we ran up against resource limitation in the COVID pandemic, I think people's appetite for ECMO for ARDS was tempered with the reality of what it takes to give this therapy and how resource intense it is and sort of what it, in an individual patient, but also on a system-wide level, what it distracts and robs you from. We've paired with these data. I think ECMO is still used. I personally believe that it has a place, but I am not someone who would advocate for early aggressive ECMO in people that are sort of dipping their toe in the criteria. I tend to reserve it for people that look very similar to the ones in the study where having exhausted everything else, this is what I have left to offer them. And also recognize that with ECMO comes a lot of morbidity. And so people that are less likely to have the morbidity associated with the potential immobilization and bleeding and other complications from ECMO, they're gonna be better candidates and sort of having an informed consent conversation or a discontinuation of ECMO consent conversation or transition to lung transplantation, which we've done several times in our institution, our necessary conversations to have before cannulation. So that brings us to the conclusion. I will sort of highlight for intended redundancy some of the things that I think are particularly important here as pearls. Remember with the Berlin definition, hypoxemia is graded by the P to F ratio. It's acute respiratory distress in less than one week of bilateral opacities that are not fully explained by a cardiac etiology. And you can see the severity characterization there. The risk is from most things that make you sick, but particularly aspiration, pneumonia, mechanical ventilation, sepsis, trauma, and transfusions. Our ventilator strategy needs to be what we do primarily and that's low tidal volume ventilation, 46 cc's per kilo of ideal body weight with PEEP, titrated to adequate saturation but keeping the plateau less than 30 and possibly using the PEEP to titrate to a delta P or a plateau to PEEP difference that is minimized, i.e. optimized. The non-ventilator strategy, prone positioning more than 16 hours improves survival. Neuromuscular blockade has not been shown in validation to improve survival. Corticosteroids, yes, give them in COVID and community acquired pneumonia, hold them back in people who have flu and late ARDS and everybody in between, it's kind of a dealer's choice. A conservative fluid strategy will increase your ventilator free days which matters to patients. In ECMO, we've all had great outcomes. We've all had bad outcomes. We have data that doesn't solve this. And so that debate is likely to continue for the foreseeable future. So I thank you for your attention. I wish you well in your preparation and study. And thank you.

acute respiratory distress syndrome

ARDS

management strategies

Berlin definition

ventilator strategies

lung protective ventilation

prone positioning

corticosteroids

extracorporeal membrane oxygenation