Okay, so I'm going to get started. I'm Bob Heise from the University of Michigan. Thanks for coming out to this session. It's called, I already asked in the time of COVID-19, and it's good to be live with everyone. And of course, it's been a long haul for everyone. And it wasn't that long ago how we were all suffering through the first wave of the pandemic. It's not gone for good. And I say that because unfortunately, our fourth speaker from Johns Hopkins, Serena Sahita, is got COVID and not going to be able to be with us. She does promise that her slides will be available for the digital content after the session. Her session was on PEEP. Should we do PEEP differently between COVID-19 and non-COVID-19 patients? You probably are aware of this whole controversy particularly going on in the first wave that related to, is this a different population? And so when I devised this session, and thank you to our speakers for accepting, this is kind of a closeted subphenotype session because you may remember well in the first wave, everyone thought, well, this is completely different. So everything you know is wrong. We should approach this in manner that's different from how we've approached ARDS for years, particularly with regard to ventilator management, but some other things as well. You might remember the first wave, we didn't give steroids. So things change. We've learned a lot and it's not gone for good. So it's still here. And so now it's time to kind of juxtapose the COVID patient against the patients at large. From my own particular perspective, I'm gonna give Dr. Colfey my bias here. We know that ARDS is a gamish of things when we look at autopsies and biopsies, and at least with COVID, generally speaking, unless you get super infected, if in India with aspergillus or United States, maybe with staph, you have a fairly pure population of viral pneumonia that you're dealing with. You don't have all the other confounds that we historically see when we assess patients at autopsy or biopsy. So perhaps this is one end of the spectrum that's integrated into a whole heterogeneous mix of things. So I'm gonna talk about non-invasive ventilation in COVID-19 and otherwise. Otherwise, I mean the other non-COVID-19 conditions. And again, juxtapose what we've learned about COVID-19 in the context of what we've also learned about other causes of ARDS. Here are my disclosures and you can take a look at those. And so we'll talk a little about obviously heated high flow and non-invasive ventilation. Most particularly, those are the modalities. And of course, you have to mention this phenomenon of weight-prone positioning, particularly because what we've come to learn about it in meta-analysis is more on that little foreshadowing for you. It's obviously not a form of respiratory support in terms of a device, but the prone positioning in the non-COVID patient is something that was described in very small series prior to the pandemic and now have had many, many large trials. And so how to interpret that becomes relevant. So the first thing I'd like to consider is should heat high flow patients be included in our definition of ARDS? You see the Leprolin definition, which I'm sure you're all familiar with. It's more than 10 years old now. And if you look for the fudge factor here, it does include non-invasive ventilation. But of course, in 2012, our pediatric colleagues were using heated high flow had really not hit the adult ICU big time in the fashion it has subsequent. So the issue is, should that be? And there was a position paper in Lancet by some of our pedal colleagues a few years ago says, yes, yes, it should. And I think that makes sense too. I mean, particularly when you have the COVID-19 pandemic staring you in the face and these people are incredibly ill and as opposed to some of the early indicators, you know, when you hit six liters, you're doing just two of them. We can keep a tube out in a lot of people with heated high flow. And what do you think they have? I mean, they're pretty hypoxemic, aren't they? And they have bilateral infiltrates and goodness knows it's not heart failure. So don't they qualify? So in this position paper, it said, sure, they should. And using a traditional kind of PDF ratio under 300, which is fairly generous as far as a PF ratio goes. And I do have a few lingering concerns. So initially I'd say, yes, absolutely. We need to include this. Berlin needs to be reconfigured. But one of the interesting things here is that, and this is a COVID population of 148 patients is, if you include heated high flow, what happens when you intubate them? In other words, we're using this PDF ratio, Berlin mild, mild, severe, fudge factor for a positive pressure. But when you actually take that heated high flow, which yes, you get some positive pressure, we all know that's what makes it so great at higher liter flows. But when you transition over from, in this case on the left, you see from heated high flow non-invasive, particularly if you're in the milder form, right? That two to 300 PDF ratio, the PDF ratio kind of jumps up pretty good. You say, well, do you really qualify then by ARDS criteria? Now you can say, it is what it is, right? They're sick, they have COVID, we need to treat them as we treat them. But when it comes to trying to study them, should they've included things like trials when they jump up on their PDF ratio so dramatically? So yes, a little less so, the sicker you get, you see one to 200 here and less 100. Oh, and by the way, we have a non-invasive and there's a phenomenon not necessarily, perhaps not as dramatic, but a similar phenomenon when you transition from non-invasive to invasive ventilation. It makes sense, right? I mean, in other words, the effectiveness of delivery and the pressures generated to the long driving pressure, what have you, are gonna be more robust when you have an ET tube than even with a helmet or a heated high flow cannula. And maybe that's less so in patients on non-invasive, the separate series from the Blue Journal. Last year, you see non-invasive below and you see heated high flow here and you see the left below. Before intubation to the right in each of the two diagrams. After intubation, you are getting a bump. This other series, perhaps it wasn't as great, but these were sicker populations. So as I showed you in the previous slide, those PDEFs around 100, the real super sick ones probably still stayed sick and they didn't really have their PDEF jump out of a normal ARDS range. So this is probably a phenomenon here of a sicker patient cohort in the separate article in the Blue Journal. But again, a similar phenomenon, perhaps less so with non-invasive. So upshot of the matter is ultimately, clearly these people have bad lungs, and DAD, I don't know, is it just viral pneumonia? Well, that's the old paradigm here when it comes to biopsies and autopsies, but they qualify, but do they really qualify when you put them on invasive? In other words, the PDEF ratio does dramatically improve. Which leads me to keeping the two about with things like awake prone positioning. As I said, they had a couple, like two physiologic studies, maybe 10 patients each pre-COVID. It kind of was an interesting notion, but of course, when the pandemic hit, as you all know, we were all overwhelmed. And so people started doing stuff like this. Many trials were done, including some at my place. Put them all together in a meta-trial, and you have efficacy. More on the GEM article last year in just a second, but this was an interesting notion, a meta-trial. Not a meta-analysis, but putting trials together and saying, look, we can probably keep your tube out, but you can't improve mortality. Maybe its usual notion is that an endpoint that's a surrogate for mortality. You might think with a meta-trial of six trials that if there's a mortality signal, you'd see it, but it was not seen in this meta-trial. And the notion here of how many hours a day seems to be somewhere around the eight hour mark that efficacy was best. Something less than that may be less effective. And you see the ROCS index, more on that in just one second, but the notion here of what happens to your breathing rate and your hypoxemia when you institute a prone positioning, as we all know what prone positioning is, and matching a V-cube being better, taking the heart off the lungs and the belly off the lungs. So clearly that's improved. So we do think that maybe that works. Well, again, that's not a respiratory therapy modality, but when you look at the meta-analyses, including these trials, one signal that appears to be true is that the decreased rate of intubation appears to be really isolated in patients who are on what is dubbed advanced respiratory support, which is defined as high flow or non-invasive, and or getting ICU management. So the sicker people seem to be the ones that you keep the tube out in the best. Again, no more effective mortality. So it does relate to the notion of using heated high flow. In other words, we don't have this study at all. What are the RCTs of non-COVID patients with weight prone positioning? Zero, they don't exist. Now, you know, I'm kind of a hypocrite here because on the one hand, I think this is lung injury like we've always known it. But on the other hand, I'm here to say, well, we've never studied a non-COVID population. It's only been the COVID population. But when you do have someone who's sick on advanced support for COVID in the ICU or non-invasive heat high flow, they should probably get a weight prone positioning, I think. And I say I think because not integrated into that analysis was our last study that came out, most recent study that came out in JAMA last year, which actually didn't show efficacy to certainly a mortality signal, not seen ever. But in prone positioning, you see missed the P value. Now, there's a trend here, but it missed the P value. So we will undoubtedly see, and I'm aware of one meta-analysis that will incorporate these data into the larger picture such as that meta-trial. And we'll see when that comes out what the signal is, but it likely will represent efficacy for preventing intubation but no mortality. The point being in the context of having patients who are receiving advanced support and or in the ICU. So what about the ICU? You know, I did a show of hands a few years ago. I'd hate to do it again because it's an embarrassment to my own institution, but how many people do heated high flow on the floor at their institution? All right, see, I need to take a picture and take that home to my hospital. Because certainly we did. You know, we had moderate care units all over the joint during COVID, which are now shut down because we don't need it. But there have been limited, even though you're all doing it, there've been limited data with regard to it. And here is it, here are data from two years ago respiratory care that show you can safely do it on the floor. And two thirds of these patients receive their entire course of heated high flow outside the non-COVID population and more than half avoid the ICU for the entire hospitalization. Not that many people crumped. So it can be done. So now you're justified. And I have, I guess I don't wanna leave myself of my administrative burden on you this morning, but certainly my work cut out for me in order to convince people that this is probably an acceptable way to proceed. So anyway, you're well justified on that. Even though we did it left and right in COVID, there were a couple of articles, you know, about makeshift ICUs that were in the literature. Results not necessarily all consistent, kind of depended on staffing model, but clearly that notion of COVID for heated high flow did have it elsewhere. And in non-COVID, it can be safely done elsewhere as well. Which brings me to heated high flow mortality benefit. We're all familiar with Florale, which is coming to us from eight years ago now, which had the surprise mortality finding for heated high flow. This is actually survival. So the opposite, obviously mortality, when it was a study actually powered for rate of intubation. And the question really here relates to when is it safe to do the heated high flow? And we have something called the ROCKS index, which just emerged, was shuttle ready for COVID. The publication on this was just prior to COVID. And you see these isoplasts that really relate to this notion, if you will, as we adapted to COVID, a happy hypoxemia. In other words, we had some pretty awful, sick COVID patients that were looking at you, talking in full sentences, breathing at 20 or whatever, the proverbial 20 breaths a minute, despite being on 80% heated high flow. It turns out that when you study it in the COVID population, it probably works even better. So this is something we use. I'm sure something you use, but it doesn't obviously take the place of sensitivity and specificity on this ROC curve. Nothing's perfect, right? But at least it gives you some heuristic. It certainly helped our hospitalists for knowing when not to freak out when we did have our floor patients for COVID. So it is something we use. But is the failure rate higher in COVID-19 patients? So here you go back to Florale with the overall population, you're talking about 30%. Now there's probability intubation here, not survival, right? So 30% getting intubated here with a PDF under 200, roughly around the same. And I think that the literature as I've seen it has shown, and I think you can probably, this does comport with, I think, reality, right? That the first wave was grim. I mean, the intubation rate was about 50%, and we're just not seeing it. I'm not sure what, you know, is it because we're giving steroids, other things that we're doing differently? You know, the TOSI, all the things that we've learned quickly, anticoagulation, what have you, on the floor, that we didn't do during that first wave. While all bets are off, we're just panicking, trying to keep up with the volume of patients we were bombarded with. So probably true the intubation rate was worse, and probably now it does seem to comport with the overall ARDS, Florale, pre-COVID population, right, of acute hypoxemic respiratory failure. So this is kind of a qualitative meta-analysis, best seen from the back of the room, because this is just a review piece that looked at series without actually, you know, sort of mathematically meta-analyzing them, if you'll accept that verbification. And you see here in the upper left, nasal heated high flow, non-invasive to the right, CPAP failure rate, and awake prone positioning rate. You kind of get a subjective sense as to what works, what doesn't. And prone positioning seems to be a fairly decent winner. We don't really have a breakdown here, a non-invasive for how it was administered, but I can tell you in COVID-19, what did we tend to do? We tended to give them what? The helmet. So let's talk a little bit about the helmet, and I'll wrap up. Again, I'm speaking more slowly for me, so I'm kind of enjoying slowing down for a change, because we don't have a last speaker, it's quite a, that's not my gift as patients. But anyway, so helmet non-invasive ventilation. This is our friend Bhakti Patel, a JP Cress article from JAMA going back pre-COVID. And again, a modality that was shovel ready. This is Lombardy, Italy during the first wave, where they're lining them up all with the helmet. And, you know, and I think a lot of places use this. We didn't. How many use the helmet for COVID? So that's interesting. So not quite the implementation of, you know, Northern Italy, at least. But, so what do we know about the helmet? Again, now I'm here to tell you, I think COVID is ARDS, like we know it. It's just one end of the spectrum. It's fairly pure without all the other mixture of stuff thrown in. And Carolyn, I'm gonna hear, tell me how wrong I am in just a few minutes. But we do have some notion here that the helmet is probably better than the face mask for non-invasive ventilation. Remember some literature about face mask, non-invasive, giving large tidal volumes, maybe being hurtful. And if you look at LungSafe, PDEF under 150, non-invasive ventilation in that database, observational, had a higher mortality. So face mask non-invasive ventilation was kind of getting a bad name for ARDS until Bhakti Patel and JP had their paper. But also when we implemented the heck out of it, at least many places did with COVID-19. In fact, if you look at meta-analysis here, this is a non-COVID population. Helmet seems to fare better than face mask. And then when you look actually at a network meta-analysis, which is either higher form of reality or complete BS, which probably this is higher form reality. But what I get a kick out of is that, what is a network meta-analysis? It's a syllogism. If A equals B and B equals C, A equals C, right? So we don't have a head-to-head. What you'd really like, what I'd like is a head-to-head in non-COVID, all comers of heated high flow versus helmet. And you can see in this network meta-analysis, we don't have that, okay? But we can see in the forest plot that we do have efficacy suggesting that helmet would be the winner. So you could argue this is, you know, kind of a meta-Bayesian kind of issue here. So we don't have that, both for mortality or intubation. We don't have that, but when you try to statistically model it, again, through syllogism of not having the actual trial, the suggestion here is that would be the winner, okay? And so there, what you have zero and zero, but you at least have a forest plot suggesting that if they were to go head-to-head, helmet might come out the winner. We do have that in COVID-19. So again, my hypocrisy at work for you is to suggest that even though I think it's all the same, I said, yeah, helmet's better, but that's a COVID patient study. When are you going to show me some real data in the big population? So, I mean, I'm a hypocrite, I freely admit it, but we do have one head-to-head in JAMA two years ago now where helmet was just as perhaps was indicated in the network meta-analysis, superior to heated high flow. So, and here are the data here from a media meta-analysis, if you will, if you accept two trials is acceptable for that. So I do think that helmet is probably here to stay. A few hands went up. I'm trying to get this implemented into my place. Again, sometimes things are very challenging to do because people don't always do what I say, right, Jason? So, and we'll hopefully work that out in the future. So what have I told you? And then we'll move on to the next speaker. Heated high flow patients should be included in the airtest definition, but you have to kind of have a fudge factor here for the fact that on the less acute side of the equation, they may be at positive pressure and put them out of the PDF range where we would normally consider to be diagnostic or sufficient to make a diagnosis of the syndromic ARDS. Secondly, awake prone position is probably most beneficial in the sicker COVID-19 patients who are heated high flow, non-invasive and or in the ICU. So if you do that, you probably should flip them on their belly. Should you do that for all ARDS patients? Well, you could say, maybe you should, we don't have a study, but if it's the same stuff, maybe you could and maybe you should, at least for eight hours a day, at least that's what the meta-trial suggests. A heated high flow has enjoyed widespread use, including outside the ICU, but direct comparisons of non-invasive are lacking patients not having COVID-19. But if it is all the same, then helmets appears to be the winner. So it is superior to face mask in the non-COVID patient population. And we'll have to see what the future holds. So what have I told you? We've learned a lot in the COVID era. I think the jury's still out. This notion of ARDS is a gemish. There's no question about it. I think this is a relatively pure population. Obviously, it's been a privilege, really, to work in these networks of networks, to learn so much so quickly and do things differently in a few short months. I would argue in the context of ARDS overall, these patients qualify and are the same as always. They're not unique. I don't think the immune system is a brand new immune system under these circumstances. And we've learned a lot of lessons. And I think many are adaptable and applicable to our non-COVID ARDS population. And the jury's still out as to see in terms of implementation, what the future holds. So thank you for that, and I'll move on.

ARDS management

COVID-19 patients

non-invasive ventilation

awake prone positioning

helmet ventilation