This lecture is entitled Noninvasive Oxygenation and Ventilation Strategy, and my name is Dr. Bob Heisey from the University of Michigan. And here are our objectives, to review the indications and use of high-flow nasal cannula and NIV in the critically ill, also to discuss the optimal use of NIV, including choice of interface and preferred modes, and number three, review the clinical studies of the use of heated high-flow nasal cannula and NIV to prevent intubation and as an adjunct to liberation from mechanical ventilation. First of all, let's talk about noninvasive ventilation. This use in the critical care units of enroncement is about 1990. It is a way to provide partial ventilatory support to a patient respiratory failure. Here you see a typical example of a patient with a face mask and as an interface and the straps applied to the head. The indications for NIV, and this is a number of indications, the gold standards of which I have included in gold, acute ichronic hypercapnic COPD exacerbations. The question about nonhypercapnic is not super clear, but clearly the acute ichronic hypercapnic COPD patient. Cardiogenic pulmonary edema, non-cardiogenic shock or requiring revascularization, but this use has actually been around for even longer than the acute use in COPD. Immunocompromised patients, we'll talk a little bit about that with the negative trial six years ago. Post-extubation, we'll talk about that as well later on. And palliative, you can take a patient who's do not intubate with COPD, for example, put them on noninvasive with the understanding that they may well be a failure to wean of noninvasive ventilation just as they might invasive. So if you're going to use it in a DNI patient, you have to go in with your eyes open. What are contraindications? Well, basically the contraindications are you're not alert enough or able to protect your airway. So if you're in a rest, if you are failing and need intubation, if you're encephalopathic, if you're bleeding from the upper abdominal source, hemodynamic instability, facial trauma, upper airway obstruction, can't clear secretions, high-risk respiration, and can't cooperate. Those are all contraindications. More with obesity is a relative contraindication. It can be done, but you don't want to wait too long. I put the ARDS patient in gray, more on that issue later, as to whether noninvasive ventilation is the right answer, at least by face mask, to the ARDS patient in the Lung-Safe trial with a PDEF ratio less than 150. The subset of patients who received NIV had a higher mortality, suggesting you should not use it. But more on that later. The advantages are, obviously, to avoid endotracheal intubation. With COPD, the number needed to treat to keep a tube out is three, and the number needed to treat to decrease the life was 11. But like having patients without an endotracheal tube, they tend to get out of the ICU more quickly and therefore have fewer ICU infections, not only pneumonia. And then, of course, less barotrauma because you're not forcing the same kind of pressures and volumes into the lung as you do with conventional mechanical ventilation. So mortality benefit in COPD, it decreases intubation mortality, as I mentioned, but I think the key point here, and a good board's review point, is it's clear that you shouldn't continue the mortality and double down, if you will, when it's clear the patient's failing and needs to be intubated. You can pick your numbers, certainly within four hours, but usually generally in one to two hours. If the pH and restoratory rate are not improved, the patient is not comfortable, then you need to just intubate the patient and not wait too long, in which case you could be putting the patient at risk. I mentioned already that cardiogenic pulmonary disease has been around a while. This is able to decrease the rate of intubation, but not really able to discern which is the best modality. Some studies felt that CPAP may be just as good as bi-level non-invasive. There were some earlier data, which I think has been subsequently invalidated, that some issue with regard to bi-level was less beneficial. The number you treat here, 30. The larger study is called 3CPO, rather, New England Journal, 2008, had no mortality or intubation benefit, but many patients did cross over to the non-invasive group. What about asthma? We always think about COPD, and we have a lot of asthmatic patients. The plain fact of the matter is that the kinds of prospective randomized trials in asthma have not been done. This is a data analysis of 53,000-plus patients who were admitted for asthma and a subset of whom received non-invasive ventilation. It was found in that subset, the non-invasive ventilation use had lower odds of receiving invasive ventilation and lower in-hospital mortality. They also found, and you might resonate with this, that an asthmatic patient who also happened to have pneumonia and severe sepsis was more likely to fail, so a pure asthmatic might potentially benefit. Again, this is a post-hoc data mining, if you will, but not a prospective study. It is clearly used in some asthmatic patients. You can consider some thoughts, though, with regard to air trapping and whether that might be potentially made worse, but so it's not really a strong recommendation, certainly not true in the guidelines, but it can be considered as well. Technical aspects, early on, there were simple BiPAP machines, now we use ICU ventilators. The most common mode is to use bi-level non-invasive ventilation, which essentially is pushing air to the lungs via inspiratory pressure, which is analogous to pressure support ventilation and expiratory wind in your face, if you will, which is CPAP effect, expiratory pressure, which is analogous to CPAP or PEEP. You tend to start low and work your way up, say five to 10 centimeters of water of the positive pressure, inspiratory pressure, and maybe five centimeters of expiratory pressure, and then decide what your goals are, and if your goal subsequently is more ventilation to blow the patient's PCO2 down, if you will, then going up on the PSP part, the inspiratory pressure would be what you'd want to do, and if your goal is to perhaps recruit some edelactatic alveoli more with, say, cardiogenic edema, maybe going up on the expiratory pressure would be more important. Now, you do see in the corner here that conventionally, for many years, we used a full face mask, and clearly we use BiPAP in our patients with nasal pillows and things of that sort, but if you have a patient with acute respiratory failure, they're generally speaking going to be tachypneic and a mouth breather, so covering the nose and the mouth is important. Now, we'll talk about helmets in just a minute, but conventionally, you see we use either a face mask that covers the nose and mouth or maybe even the eyes, but again, important issue about recognizing treatment failure. I think that will be a board's question for sure. We'll talk about that again, 62-year-old patient with COPD who was breathing 32 times a minute but using accessory muscles, his chest x-ray shows hyperinflation, and pH is 7.32, PCO2 65, PO2 78 on six liters, he's on steroids, he's on an antibiotic, and he has started on noninvasive ventilation at 10 inspiratory pressure, 5 expiratory pressure. An hour and a half later, though, he's breathing 35 times, or 34 times rather, per minute, he's a little bit drowsy, he's gotten some Ativan, but he's arousable, he's still got a lot of rockeye, he's still using accessory muscles, and he's leaking a little bit about the mask. Here's his blood gas now, 7.28, and PCO2 now 73, with PO2 of 68 on 10 over 5. What would you do? Would you say, let's crank up the pressure to blow down your CO2, let's change to a full face mask, emergently intubate the patient, or attempt to wake them up with Flumazenil because they did get some Ativan, emergently intubate the patient, and this is a little bit ambiguous there, but the guy's drowsy, and he's more hypercapnic, it's been a couple hours, it doesn't seem like it's working, and you just need to know when it's time to move on. Now, let's also talk about the newer kid on the block, the heated high flow nasal cannula, which we have taken over from the pediatric realm, and I'm fond of saying this is a picture of a guy with heated high flow, he's a happy camper, and why is he a happy camper? Well, you know, to get someone started who's an inspiratory failure with a full face mask can be challenging, and to get that interface tolerated can be challenging, but clearly, a larger bore cannula under the nose is something that's well tolerated by patients, it washes out dead space, but one of the keys to the use of heated high flow is a CPAP effect. As you go up on liter flow, and we're talking about 30, 40, 50 liters per minute, you can see over here you get more CPAP, and CPAP can be beneficial to recruit at a lactetic alveoli in the setting, say, of lung injury and ARDS, or other causes of hypoxemic respiratory failure. I want to tell you about one very important trial called Florelli, where heated high flow was directly compared to noninvasive ventilation and O2 mask in patients without heart failure because CPAP is known to work, without hypercapnia because, of course, noninvasive ventilation is known to work, and we're not neutropenic. The primary endpoint was day 28 intubation, which was not achieved, except it was seen in post hoc analysis with the PDF ratio under 200, which is to say heated high flow in this hypoxemic respiratory failure group had a decreased rate of intubation in the more hypoxemic patients. However, even more important was a secondary endpoint of a lower mortality than face mask or noninvasive ventilation in the hypoxemic respiratory failure patients. So clearly, in this population, there appears to be superiority for keeping an endotracheal tube out in the hypoxemic respiratory failure group. Now, how do you know? We talked a little bit about failure in noninvasive ventilation. Failure in heated high flow can also occur. You may need to intubate a patient. There's been something developed called the ROCKS index, which isn't rocket science. It's SAT over FiO2 divided by respiratory rate. And here's some calculations that you can see, the higher, the worse. And most intubations with heated high flow do occur within 12 to 24 hours, which makes sense too. But just as with noninvasive, you better know when the patient's failing. What are we really saying? I find this graph here to the right to be instructive. What you're really saying is, even if you are on a higher FiO2, say 80%, where you might come running at a patient with an endotracheal tube, if you have rapid channel breathing and you're lacking that, you might be okay. Now, this has also been validated in the COVID patient population. We had a whole lot of patients with the heated high flow on our floors, and they wanted to know that the hospitals care for these patients about failure. And ROCKS index can be instructive as well, especially because we seem to have a lot of patients who were this so-called happy hypoxia, who were quite hypoxemic, but easily conversant, and not to get me again speaking in full sentences, who could be managed just by the high FiO2. This is the mortality benefits seen in florality. So just to harken back on that again, the key point here being that for the hypoxemic respiratory failure, I would argue COVID or not, heated high flow is probably the better way to go. And again, you have to recognize when someone is failing and the ROCKS index might be of some value to you there, but also is looking at the patient. Now, there has been some subsequent attempts to refine our understanding and particularly, what about the immunocompromised patients, which are frequently not enrolled in these trials. And it appeared that in this JAMA study from six years ago now, that non-invasive inhalation did not seem to benefit the immunocompromised patients who, as a rule, have hypoxemic respiratory failure. An analysis from Lancet Respiratory Medicine seemed to suggest that it might even be harmful in these patient populations, even though the original non-invasive data that looked way back when we didn't have heated high flow looked at only 40 in 52 patients. We do have issues also now with regard to whether or not heated high flow is a way to go in immunocompromised. It may not necessarily be better for our patients who are immunocompromised than a full face mask with intubation rates, as you see, and mortality, as you see. But clearly, no harm was accrued. And for my money, I'll still go with heated high flow in this patient population. But now that we've had enough experience with heated high flow, what's true is that meta-analysis tend to show that there's no reduction in mortality, but there is a decreased need for invasive mechanical ventilation, which is a little bit odd. If you think back to the non-invasive ventilation data with face masks, the notion of keeping it someone without an endotracheal tube, having benefits of less pneumonias and barotrauma, you might think this would track out to mortality benefit, but apparently not, by meta-analysis, at least. So I did mention to you already the LoveSafe data that suggested that, with face masks anyway, that non-invasive ventilation had a higher mortality. What about the helmet? And we do have one single center RCT by Bhakti Patel and J.P. Kress at Chicago, obviously not blinded, but suggested with ARDS that a helmet might be able to provide non-invasive ventilation in the ARDS patient population and superior to the face mask alone. And we do have some direct physiologic comparisons. Sorry, I don't think you see the reference here. This is from the Blue Journal that went head-to-head, if you will, heated high flow versus helmet non-invasive ventilation, and found that there was perhaps less work of breathing, which is to say pressure time chronic in non-invasive ventilation, but no change, no difference in transponder pressure, comfort, or PCO2 when you directly compare the two by crossover design using a 50 liter heated high flow versus NIV with these settings here. We do have a network meta-analysis though. And again, network meta-analysis, as I showed in my other lecture, is a way to compare things that haven't been compared before. And there are no RCTs directly comparing heated high flow to a helmet in hypoxemic respiratory failure, but this network analysis suggests that if that were to be done, that the helmet would come out superior. You can see down here, and there are no patients and no trials, and you can't see, unfortunately, the line of identity here, but it did find superiority. You can see the numbers of this. I think that that needs to be vetted. I'm not against the helmet, and it is interesting observation and hypothesis generating via this network meta-analysis. This is the one studied by Bhakti Patel I mentioned to you before, but that was a comparison of helmet versus face mask. So again, we don't have a direct comparison, and whether that helmet might be superior is an intriguing possibility. I do wanna just have a brief direction because if we're talking about ways to keep endotracheal tubes out with either heated high flow or NIV, so there is another wrinkle in that obviously this is not a form of ventilation, it's a form of treatment, and that is prone positioning in the non-intubated patient, and there were at least four trials I'm aware of that looked at this in COVID-19 ARDS. You can see the two examples here. Now, a patient's lying on their belly. We talked about in my other lecture the issue of the compressive adlectosis in independent lung zones, and the weight of the heart and the abdomen on the posterior lung zones when you're lying on your back are relieved when you're in a prone position whether you're on a vent or not. So the question is, does that help? There were some physiologic trials in awake proning even before the COVID pandemic started that show perhaps that was a way to better oxygenate a patient without really any outcomes. We have some physiologic data. We don't have the clinical trials now released, but again, taking the weight of the heart and the belly off the posterior lung zones by proning a non-intubated patient will improve PO2. However, that's not the same as an outcome. We know, for example, in the ARDS patient, improving PO2 with nitric oxide does not translate to mortality benefit. The question is, does awake proning translate into a clinical outcome such as decreased intubation? And we'll have to wait for those trials to come to completion. So I'm not against awake proning. I'm just not for it, pending further data. All right, so that's keeping it to about, what about post-extubation support? We have the same modalities, the same non-invasive modalities to choose from, face mask, heated high flow, and non-invasive ventilation. So we have a 62-year-old male, COPD again, the patient's analysis control, and he's on a total volume of 450, FIO2, 40% PEEP of five. He's breathing about eight liters a minute. His PCO2 now is 54, the pH 7.36. His rapid shallow breathing index is about 110. He's tried on a spontaneous breathing trial, 40% oxygen after 30 minutes. His PCO2 is 56 instead of 54. He's breathing 35 times a minute, a total volume of 250. He's not really retracting, and he's hooked up back to the ventilator. What would you do now? Extubate him to bi-level non-invasive at 10 over five, try another SBT later in the day, rest him for the remainder of the day, and reassess him in the morning, or attempt another SBT later as well. What would you do? I consider extubating the guy to 10 over five. Why? Well, we do have some data that goes back to 2004 in an unselected patient population. They're trying everybody on non-invasive ventilation post-extubation. It didn't help, but there was a possible signal in the hypercapnic subgroup, which makes sense if you think about it. If we know that QNI chronic hypercapnic COPD benefits from non-invasive ventilation in terms of keeping an ET tube out and mortality on the front end, perhaps it's no surprise that extubation via non-invasive ventilation might benefit a patient on the back end of their clinical course. And in fact, that observation was vetted with post-extubation support via non-invasive ventilation having decreased rate of intubation and decreased 90-day mortality in Lancet 2009. Admittedly, other studies were less compelling, but clearly seem to establish this modality, non-invasive ventilation, in this particular setting, hypercapnic COPD. Not the eucapnic patient population, by the way, only the hypercapnic population. The issue about non-invasive ventilation to prevent weaning failure is less clear. In other words, we have some different populations. What do the guidelines tell us? Well, the guidelines are a tad outdated from 2017 when we didn't have as much information, particularly two very important trials, eugenic heat and high flow as an adjunct post-extubation, but it did seem to suggest that non-invasive ventilation for the high-risk patient post-extubation, that is to say with a strong recommendation of least moderate quality evidence. Some of the other aspects I shared with you in my other lecture. We do have heat and high flow. It's been around for a few years now, and it was first examined in the post-cardiac patient population and was found to be non-inferior to non-invasive. So certainly an easier modality to accept. And if it's for cardiac patients, it works just as well, why not use it? But if you do look at what's high risk, we do have two very important trials that came out. One is a direct head-to-head comparison of heat and high flow versus non-invasive ventilation post-extubation in a high-risk population and in a low-risk population. This is the high-risk population where high risk is defined by age and comorbidities such as COPD, it's heart failure, being obese, being on the bed a long time, and it was non-inferior. Now this was not a pure population. There was a subset of COPD patients. I would argue that COPD probably is better treated with non-invasive. When you look at Alzheimer's though, heat and high flow, post-extubation, pre-failure, mind you, again, no one's saying you should do this if you're failing, but you do this the second the tube comes out, is non-inferior to non-invasive in this high-risk population. So consider non-invasive in the hypercapnic and other high-risk patients, but certainly you can use this in post-op cardiac surgical patients, and it really is equivalent in high-risk patient population, understanding that COPD subgroup was small. What about low-risk patients? This is a study that compared heat and high flow, not to non-invasive, but to just good old-fashioned face mask. They excluded COPD, CHF, patients that were too old or on the vent for a long time. This should be greater than 65 here. And the re-intubation rate at 96 hours was lower in the heat and high flow group than the face mask group. This suggests that everyone should get 24 hours of heat and high flow, and I'm not sure I necessarily can sign off on that, but it kind of depends on what your re-intubation rate is overall. But heat and high flow can be a valuable tool post-extubation in the high-risk patient population and in the low-risk population, and I would argue the latter, perhaps on a more selective basis, but this was taken all comers. Then the last wrinkle here was, and I guess you can't see the reference here, a JAMA paper from two years ago now, that said, well, why not use both? I mean, what's a better clue to using both? So this is a group of high-risk patients who had high flow nasal cannula alone versus high flow nasal cannula plus non-invasive, and they flipped back and forth, as you see, between modalities for several days, and in fact, found that the combination was actually better than the high flow nasal cannula alone. Sort of begs the question, in retrospect, looking back on that trial before, why would heat and high flow in that high-risk population be non-inferior and non-invasive, and yet, if you combine the two, be superior to high flow nasal cannula, but these are what the data show, and here's what it shows again. Decreased re-intubation rate at day seven from 18 to 11, but no difference in mortality. We're all talking about, as an end point here, re-intubation. There is a European Society guideline regarding high flow nasal cannula, a strong recommendation in hypoxemic respiratory failure compared to conventional oxygen therapy with moderate certainty, and a conditional recommendation following extubation with moderate certainty. So what's our non-invasive oxygenation ventilation strategy? Prevent intubation for COPD and cardiogenic pulmonary edema. Possibly consider the helmet in ARDS, but otherwise, if you don't have the helmet, go with a heat and high flow nasal cannula. Reassess, and if the patient's failing, intubate them. Now, you can actually cheat a little bit, provide CO2 narcosis as due to failing from the underlying disease, and possibly wake them up with non-invasive, and then give them the non-invasive for acute and chronic hypercapnic COPD. You're living a little bit on the edge there, but I would argue heat and high flow nasal cannula for hypoxemic respiratory failure to keep a tube out, understanding there may be no mortality benefit. And in the high risk patient population, high flow nasal cannula may be just as good, provided you recognize that the COPD group might still be targeted best with non-invasive, and the combination of non-invasive plus heat and high flow is actually better than heat and high flow alone. Thank you.

noninvasive oxygenation

ventilation strategies

high-flow nasal cannula

noninvasive ventilation

HHFNC

post-extubation support