Hi, everybody. Thanks for joining us. Today, we're going to be talking about nonconventional ventilator modes in pediatric ARDS. My name is Nader Yahya. These are my disclosures. So, today, we're going to briefly review the mechanics of mechanical ventilation. We're going to review the evidence, and we're going to propose some clinical applications of these nonconventional modes. We're going to talk about four modes today, airway pressure release ventilation, APRV, high-frequency oscillatory ventilation, or HFOV, high-frequency percussive ventilation, HFPV, using the VDR ventilator, and high-frequency jet ventilation. The first mode we'll talk about is APRV. So, APRV is a subset of bi-level in which you set a P-high for a T-high and a P-low for a T-low. Any mode in which you set P-high, T-high, P-low, and T-low is a form of bi-level. And APRV is the name that has been adopted by the version of bi-level where you have extreme IDE inversion on the order of, like, IDE ratios of 10 to 1. What this is trying to accomplish is sustained alveolar recruitment with a prolonged P-high with spontaneous respirations during maximal lung recruitment so that you optimize alveolar ventilation. And ideally, you have better matching of ventilation and perfusion, thereby improving oxygenation. And the hope is that you're able to do this with lower ventilator pressures than you would see on conventional when you're trying to recruit more lung. But that, unfortunately, would end up necessitating the recruitment maneuvers or higher PIPs or plateau pressures. Rather, if you have a sustained prolonged CPAP, which is your P-high, for a sustained long I-time, which is your T-high in this mode of ventilation, that you can maximally recruit that lung at a lower peak pressure and then allow spontaneous ventilation on top of that. The spontaneous ventilation part of it is actually supposed to be key to how this works. Because if you're activating your diaphragm, then you should preferentially be recruiting and opening up lung that's in the more dependent regions of your body, which tend to be more socked in, more diseased, but it's also preferentially where your blood flow goes. So you should, if APRV is working correctly with spontaneous ventilation, you should be able to recruit that lung as much as possible with a prolonged P-high for a longer T-high. And the spontaneous ventilation is allowing you to optimize ventilator perfusion matching in a way that conventional ventilation doesn't let you do. And if this works, you should be seeing improved hemodynamics and less sedation. No neuromuscular blockade is listed as a potential benefit here, but it's actually quite fundamental to how APRV is thought to work, particularly optimally. You can do APRV with neuromuscular blockade, but you're negating a lot of the benefits of APRV if you do it that way. This is what the waveforms look like. On the top is a pressure over time scalar, and the bottom is flow versus time. And in both cases, it's fundamentally not that different than pressure-controlled ventilation. The flow scalar, for example, is a decelerating flow pattern that you would see in any pressure-controlled ventilation mode. And the P-high is a peak pressure that is sustained for a prolonged T-high. And what this is meant to do is to, again, recruit the lung and have spontaneous respirations during this cycle, allowing you to actually breathe during this cycle, during this period of maximum recruitment. The P-low for T-low, in many places by convention, is set to zero. In some places, it's set to any integer less than five or some very low value. And the goal of P-low for T-low is to get rid of all that CO2 that you've been ventilating and producing during this spontaneous ventilation during P-high. And getting rid of the CO2, you need it to leave the system. So it's essentially like a CO2 dump to allow it to go out of the exhalation valve of the ventilator, the exhalation tubing of the ventilator. And so you don't need it to happen that long. And if you look at the flow diagram during expiration, what you're actually doing is terminating the flow between 50% and 75% of peak expiratory flow, and you're not letting it achieve zero. So you're essentially causing a purposeful breath stacking or air trapping with this. You're not letting your flow go back to zero because that's just not how this mode of ventilation works. It's just you have to think about it differently. But that extra bit of pressure, which is left in, is meant to prevent complete alveolar de-recruitment. There have been some animal studies which looked at some of the physiologic characteristics of APRV. There's a pig study, not that many animals, which used ischemia, reperfusion, and peritoneal sepsis model in which APRV was compared to low tidal volume ventilation versus sham. And APRV overall had less risk of development of ARDS, overall less lung edema, better surfactant protein expression, better gas exchange, better lung compliance, and better histology. There has been a non-randomized crossover study in adults in which the administration of APRV resulted in lower peak airway pressures. In more recent adult studies, they haven't found as much difference in ventilator days or length of stay or mortality. So there was some question as to whether or not APRV was, even though it may have led to some potential physiologic benefit, whether or not it was leading to any actual clinical benefit. Part of this is because some of the measurements that we make at airway pressures, for example, don't necessarily account for the pressure that's generated during strong spontaneous ventilation. And so a mode like APRV, which promotes strong spontaneous ventilation, could end up actually having negative intrathoracic pressures, which are not going to be captured by pressures measured at a ventilator. So it's possible that the nominal lower peak airway pressures may not have been much lower than conventional low tidal volume ventilation if you account for the patient effort and their contribution to the transpulmonary pressure, which without an esophageal manometer or some other way to test it, you wouldn't necessarily have a good sense of what the contribution of the patient themselves to their transpulmonary pressure was. In two adult studies, people found longer lengths of stay and longer ventilation and increased complications with APRV. A more recent randomized control trial showed some potential promise in which 138 adults with ARDS were randomized to low tidal volume ventilation versus APRV, and the APRV arm had substantially more ventilator-free days and better mortality. The criticism of this study was that the allocation of the onset of ARDS in a small study by chance, the randomization scheme fell such that the low tidal volume arm got somewhat sicker patients. And so it was unclear whether or not some of the benefit could be attributed to the baseline severity of illness that a small trial is not quite as well able to account for. In pediatrics, there have been some case series which shows that APRV has the potential to improve oxygenation. In a crossover study, children had comparable gas exchange comfort and hemodynamics on conventional mechanical ventilation as they did on APRV. This study, unfortunately, did not include children with severe ARDS. So overall, APRV likely is safe to use in kids. It has the potential for improvement in oxygenation. This should be considered an oxygenation rescue, not a ventilation rescue. It's hard to invoke something with a 10 to 1 IDE ratio as good for blowing off CO2. It does not require a change in the ventilator circuit, which most conventional ventilator machines can be adapted or have already adapted or have software or a package to deliver some form of bilevel or APRV. And there should be less sedation required than in high-frequency modes. This is fundamentally just a different version of decelerating flow ventilation, and so regular machines can do it. And it's good for patients in whom you don't want to paralyze or whom you think can get by without paralytic, in which you're not using paralytic for some other reason or you're not using paralytic as a respiratory adjunct. This mode of ventilation as a oxygenation rescue mode actually would do better in patients who are spontaneously breathing. Hypercarbia, as stated, may be a barrier. This is not really considered a ventilatory rescue. This is considered an oxygenation rescue. Anecdotally, people say it may work better in older children, but some of that is just some of what we're willing to tolerate as providers. Tachypnea in younger children, while we would be willing to tolerate it in certain physiologies like bronchiolitis, looks less tolerable in certain other diagnoses like ARDS or oxygenation failure. And so it's not exactly clear to me that APRV is necessarily better tolerated in older children versus our threshold for tolerating their tachypnea is different for younger versus older children. To that point, I have tried this in different age groups and seen variable success with it. The one consideration I would have is that because this mode relies on synchrony and getting a patient to breathe reasonably comfortably without dyspnea during the P-high, this really does require a fair amount of comfort. And so if you put somebody on APRV and they actually do seem more dyspneic or more uncomfortable by metrics potentially other than just respiratory rate by like heart rate or how their face appears or their grimacing or other sort of clues that like they really don't like this, then you really are also negating a lot of the benefits of APRV. But if it works and they're comfy, then this is a good oxygenation rescue. The next mode we'll talk about is high-frequency oscillatory ventilation. And the oscillator has gotten a lot of bad press. And we're going to talk about some of the data that's surrounding it and just try to figure out like from all the heat rather than from the light that has been shed on things from many of these studies where exactly the oscillator falls in the current pantheon of rescue ventilators. So the oscillator delivers a constant mean airway pressure. And so what you do is you find a higher level of CPAP and you wiggle around it. And that is how the oscillator works. The amount of recruited lung is directly dependent on the mean airway pressure that you choose and the stiffness and resistance characteristics of the lung, primarily the compliance characteristics of the lung that you're using this on. And it oscillates very small tidal volumes, like dead space or sub-dead space tidal volumes via an oscillating pump. And the degree of displacement is determined by a combination of the frequency and the power, which delivers an amplitude. So you sustain a mean airway pressure, which recruits and opens up lungs to whatever degree that pressure will allow that lung to be opened up. And you wiggle or oscillate around that mean airway pressure. And unlike other modes of ventilation, both inhalation and exhalation are active processes on the oscillator. In the 90s, almost 30 years ago now, there was a randomized control trial of the oscillator versus conventional ventilation in 70 children in which the oscillator arm did see some improved oxygenation and they were overall less oxygen dependent at 30 days, but there was no real difference in ventilator days or survival. And this trial has been criticized because it was conducted prior to the advent of low tidal volume ventilation in adult and then eventually in pediatric ARDS. And so it wasn't clear whether the conventional ventilator strategy was comparable to what would be done now. In an observational study of 66 children with ARDS, Brogan et al. found that patients with primary lung disease, that's to say direct ARDS like pneumonia or aspiration, were potentially more likely to respond. And this cohort had a particularly low mortality rate and they found that organ failures were associated with higher mortality. And so the oscillator may be doing what you need it to do with respect to the lung, but you can't necessarily control for all of the other extra pulmonary organ failures which could contribute to outcome. This is a truism of all pediatric ARDS studies and trials. There was some question about whether or not we're conceptualizing the oscillator incorrectly, like we currently use it as a salvage mode for oxygenation and occasionally ventilation failure, but should we be using it up front? Is early in fact better? And there were some observational studies which suggested that earlier application may in fact be better. In a study in 1996, Sarnak et al. found that lower OI at start equated to improved survival. And a different study like Slee-Witchfels et al. found that in 53 children, earlier application of the oscillator may be better than later application of the oscillator. And survivors had a better oxygenation response. And the confounding vindication of these studies is what tempers enthusiasm for using the oscillator up front. In the first study, for example, a lower OI at the start also somewhat implies that patients with lower OIs have better survival to begin with because oxygenation index is a predictor of survival. And so lower OI, a different way of saying that, is milder ARDS. And so what the first study may have just proven is that milder ARDS has better survival than moderate and severe ARDS, which is not particularly surprising but is not easily attributable to the oscillator as having a benefit. Similarly, the second study of 53 children where it suggests that early versus late application of the oscillator is better is confounded by the fact that if you put somebody on the oscillator early and they do well, that's a certain type of patient. And if you put somebody on the oscillator later, then that means that by definition they failed conventional mechanical ventilation or failed whatever else you were doing in that time. And so what you're doing in that kind of a study setup is comparing early successes to eventual failures of something or other. And failures always do worse. Failures always fail for a reason. And so this study also has some intrinsic bias, which makes it difficult to interpret as to whether or not early oscillation is actually causal for improved outcome. More recently, Rowan et al. looked at 85 pediatric hematopoietic stem cell transplant patients and survivors were transitioned to the oscillator earlier in the course of ventilation at zero to two days. And they were also transitioned at lower oxygenation indices than non-survivors. And no patient who transitioned to the oscillator in this stem cell transplant cohort survived if they were transitioned after a week of ventilation. And the authors here concluded that early oscillation was associated with improved survival compared to late oscillation. And just like the previous slide, this form of study carries with it the same sort of concerns about intrinsic bias, where lower OI and earlier transition to the oscillator is fundamentally picking up a different cohort of less sick patients than higher OIs and later transitions to the oscillator. Nevertheless, these studies do establish the premise that it's possible that we are not thinking about the oscillator correctly as a salvage mode and that we should, in fact, be designing trials to test out whether this would be a better upfront mode. It's a hard question to answer with observational studies because of the biases and the confounding by indication that we've talked about. There have been multiple studies which have looked at the oscillator and its association with worse outcomes. In retrospective cohort studies of 104 patients exposed to both the oscillator and APRV, there's more inhaled nitric oxide use, neuromuscoblockade, vasopressor use. Gupta et al. in a registry study looking at 9,000 patients, of whom 900 got the oscillator, found higher mortality and longer ventilation with the oscillator with no difference in outcome attributable to when the oscillator started. Bateman et al. in a post hoc reanalysis from the Restore study found after propensity matching, like the 210 or so oscillator patients, that they found no difference in mortality, but they did find longer ventilation, more sedation, neuromuscoblockade. None of these studies are free from the biases, despite the propensity matching or the large sample sizes available in some of these studies. None of them are free from the confounding by indication bias. The Bateman et al. study, for example, suffered from pretty poor propensity matching and some pretty serious residual confounding. Gupta et al. study lacked a lot of the variables that would be important to include as severity of illness markers, such as oxygenation index or CO2 or PIPs between the oscillator and the non-oscillator groups. And so a lot of these studies cannot overcome their intrinsic biases by both their observational nature and the availability of the data. So overall, consider the oscillator as a salvage mode for oxygenation rescue. It absolutely seems to have a role. I don't think we can apply adult data and the negative adult trials to pediatrics, primarily because the adult trials failed because the adults couldn't tolerate the toxicity of the oscillator. And it's not clear whether the adult protocols for the trials are exactly how we would use the oscillator in pediatrics customarily. So it becomes hard to interpret that population. Pediatrics has a better comorbidity profile in many ways. They tolerate higher mean airway pressures without as much hemodynamic compromise. And they're just different. They're just different. And so it's worth investigating the efficacy of oscillators specifically within pediatrics. In the absence of good data suggesting that we should use it up front, it really should be reserved as a salvage mode of ventilation. It does consistently require more sedation neuromuscular blockade. Like the oscillator is weird. If you've ever put yourself on the oscillator, like through a face mask, it's a weird way to breathe. The PALIC 2015 guidelines suggested that you could consider the oscillator if you have elevated plateau pressures, suggesting that to adequately oxygenate our ventilator patient, you would need higher pressures on conventional ventilation than most people are comfortable with and higher than what PALIC recommends. And some specific populations that may benefit specifically from the oscillator are those with primarily respiratory disease without significant multi-system organ failure, very stiff lungs in which you're using toxic conventional ventilator peak and plateau pressures, very de-recruited lungs that just for a while just need a sustained mean airway pressure with an infinite eye time, which is pretty much what the oscillator is providing you. And patients who you've already tried neuromuscular blockade and yet your ventilator pressures and your gas exchange is still not optimized. There's some controversy as to whether or not early transition may have benefit, but the confounding by indication, confounding by severity of illness of a lot of the observational studies makes this at best a hypothesis in need of testing. The next mode of ventilation we'll talk about is high-frequency percussive ventilation, which is delivered by the BDR. Pressure control elements of conventional pressure control, and it combines some of the high-frequency elements of high-frequency oscillatory ventilation, it uses percussion, not oscillation, but that allows it to do two major things. And one of them is clear CO2, and the second thing is clear secretions. So these pneumatically driven small percussions, the percussive rates, is a form of incremental breath stacking. So if in conventional ventilation, a PIP of 30 over a PIP of 10 is like you getting on an elevator and going from the 10th floor to the 30th floor, then the BDR going from a PIP of 10 to a PIP of 30 is like taking the stairs. As you percuss in the PIP, then you're taking smaller incremental breath stacks of going from like 10 to 15, 15 to 20, 20 to 25, and then 25 to 30. And so you get there in smaller subtitle volume increments that eventually gets you to the PIP that you wanted to set. And then you stay there for a bit, which is your eye time, and you percuss while your lung is fully recruited at that PIP, and then there's a passive exhalation. Most of the data for this is in the burn population, and that is because constant percussion during conventional pressure control allows really good mobilization of secretions, which is why this was essentially developed. And in a study of pediatric burn patients, there were lower peak pressures without a difference in mortality. And in a second burn study, there was improved oxygenation, which was described with lower peak pressures, but again, no difference in mortality. In more recent pediatric ARDS data, people have found improvements in oxygenation index and CO2, with a reduction in peak inspiratory pressures and improvements which were sustained over 48 hours, suggesting that it improves gas exchange. And this does seem to be a reliable, particularly CO2 rescue. This is a good ventilatory rescue. Oxygenation does improve, but that seems to be less so than ventilation. And so this is a good example of a good ventilatory rescue. In a study of ECMO patients, specifically looking at the implementation of high-frequency percussive ventilation while on ECMO, the combination of high-frequency percussive ventilation plus bronchoscopies was a form of aggressive pulmonary toilet, which seemed to increase ECMO-free days. And this utility of high-frequency percussive ventilation, the secretion clearance, has been described in case reports in hydrocarbon aspiration, cystic fibrosis, and tracheal injury on ECMO, all of which are high secretion burden, pulmonary toilet-requiring type of diagnoses. And so this kind of suggests that the places to consider high-frequency percussive ventilation are hypercarbia and secretion-based diseases, which seems to be where its indication is falling out. There's not sufficient data to recommend routine use. And so don't put it on every bronchiolytic who needs an extra treatment or extra suctioning. It may have benefit in excessive secretion burden. So inhalational injury, again, cystic fibrosis, surfactant disorders, and really bad bronchiolitis that still seems to keep plugging despite your traditional respiratory toilet that you would do. During ECMO, there may be some benefit. And you can consider during ECMO during when after stabilization on your ECMO settings and you're willing to be a little bit more aggressive with their lung recruitment and actively try to clean out their lungs, particularly like for primary pneumonia or primary aspiration ARDSs. Like this actually may be a good time to try something like the VDR for recruitment and plus-minus bronchoscopies with the goal of trying to get them off ECMO faster by cleaning out their lungs and increasing the amount of available parenchyma. The last mode of ventilation we'll talk about is the jet ventilator, high-frequency jet. So this was developed in the 60s and 70s to aid with bronchoscopy, which is still a way that it's used now. The upshot of this was that people needed to be able to do bronchoscopies without having to stop or desaturations all the time. But you have a big bronchoscope, which is including your endotracheal tube. And so there needed to be some way of delivering oxygen better than what was currently being provided. So people came up with this high-flow jet adapter, which would jet in in spurts oxygen, primarily to aid with improved bronchoscopy. And this eventually found a role eventually in neonates, but it has been used across a bunch of different populations. The jet inductor tip is placed at the endotracheal tube and it provides short bursts of inspiratory gas. And because it's a short inspiratory thing, while air trapping can occur, like the idea is actually that the eye time is so short that this has found a niche in primarily obstructive disease processes, which need long expiratory times. By using the short burst or short eye time of the inspiratory cycle in this mode of ventilation, you're essentially providing a really long E time. There's a separate peep, and this can be combined with periodic side breaths or conventional ventilator breaths to prevent atelectasis, because like short bursts are not going to provide sufficient lung recruitment. When you look at the different modes of ventilation, the jet ventilator ends up taking on some of the properties of high frequency as well as conventional ventilation. And you can get superimposed side breaths on top of even this jet ventilator waveform. The forced jet stream, much like the oscillator, forces the central air to flow inward. So the oxygenated air that you're delivering is flowing inward, and that allows exhaled air to come out around that. And it can move pretty well in obstructive physiology because of these properties. So you can get the air in and you can exhale because of this high frequency jet bursting down the middle of a tube nature of what you're delivering. This has some utility in secretion removal. The bursting jet down the middle can sometimes break up secretions and that can facilitate their removal. It doesn't work quite as well as the VDR for this, but it can be done. It's been used in meconium aspiration and neonates, plastic bronchitis after Fontan's, and its best description remains as an adjunct to bronchoscopy in which it's actually quite good. In neonates, in a multi-center trial of patients with PIE, and PIE is the most obstructive physiology you can imagine in pediatrics, then the PIE itself actually improved on jet ventilator. And overall, patients were able to use lower peak airway pressures, which resulted in fewer pneumothoraces and air leaks. And this was associated with an improved survival. In adults, in a crossover study, there was some improved oxygenation using jet ventilator and reduced CO2 and peak pressures, as well as improved aeration of the lung on imaging. In pediatrics, the data is a little bit more sparse. In pediatric ARDS, people saw similar sorts of things where they saw improved oxygenation and lower peak pressures with less barotrauma, or at least no evidence of new barotrauma. With six kids, it's hard to make a definitive statement about more or less of anything. But they were able to do this safely. They were able to oxygenate and ventilate safely at lower nominal peak pressures than they were on conventional without apparent evidence of harm, is what these authors concluded, which is probably a fair conclusion. In a somewhat larger study, in patients that already had barotrauma, then there was no additional air leak induced by jet ventilator. And as every other nonconventional ventilator study that we've discussed, with the associated bias of confounding by indication, then survivors transitioned earlier. The overall data in the jet is limited. And it's primarily older data. The VDR has really replaced a lot of the jet ventilator indications in many ICUs around the country. You can consider it in patients with pneumothoraces, pneumometastinas, bronchopleural fistulas. All of these are air leak conditions that would benefit from short I times, lower flows, and longer E times. Intellational injury, hydrocarbons, foreign body aspirations, obstructive secretions are very comparable indications to high frequency percussive ventilation. But they're also there for the jet because those concepts are still there. Whereas the VDR and high frequency percussive ventilation percusses out those secretions and blows off CO2, the jet has high frequency bursts of oxygenation with prolonged E time to blow off the CO2. When you compare the modes, APRB is really for spontaneously breathing patients requiring hypoxemia rescue. High frequency percussive and jet ventilation are primarily for hypercarbia or CO2 rescue and secretion clearance. The oscillator is for both except for secretion clearance. The oscillator is good for oxygenation rescue as well as ventilation rescue. It's billed a little bit more as an oxygenation rescue because the sustained CPAP is really good. But active exhalation and the ability to ventilate using power and frequency somewhat independent of a fully recruited lung, you recruit the lung with mean airway pressure, you get it open to eight or nine ribs, and then you can adjust your ventilation according to your power and your frequency, actually make the oscillator a pretty good ventilatory rescue as well. What the oscillator is absolutely terrible at is pulmonary toilet. So if secretions are your major malfunction, then the oscillator is in fact one of the less desirable modes of ventilation. Sometimes you just need to do it. Sometimes lungs are just so derecruited or so stiff that you just need the sustained recruitment of the oscillator. But if secretions are a significant contributor to your problem, then a sustained mean airway pressure has the unfortunate side effect of pushing all the secretions out to the periphery. And then you get diminishing returns for that mean airway pressure because you've induced a lot of peripheral obstruction by pushing all the secretions out to the periphery. And so there's stuff that will eventually just need to come out or resorb. And so if secretions are a major problem, then the oscillator, that is the one time that the oscillator is not a particularly good mode of rescue. The oscillator of the ventilator modes listed here probably represents the last stand before ECMO for a lot of centers. And so its ability to provide good oxygenation and good CO2 rescue probably does make an attractive last ventilator mode to attempt, or at least to stabilize on as a patient goes on to ECMO. To conclude, all of these modes require higher mean airway pressures than conventional, which is probably why you're doing it. Even the rescue modes of the jet and high frequency percussive ventilation either have comparable mean airway pressures or somewhat elevated mean airway pressures. Even when you're doing them for primarily CO2 rescue, you do need to pay attention to that aspect of it. Part of how these modes work is by improved recruitment. There's theoretical advantages for each of them. They each have their own individual niche. APRV is good for patients in which you can maintain spontaneous breathing in synchrony. The high frequency oscillation is good for oxygenation and CO2 rescue in which you're willing to paralyze and which you require just sustained constant mean airway pressure. It's less good for secretions and it's probably more cardiotoxic than the other modes of ventilation. Sustained mean airway pressure without a high and low pressure is probably the hardest thing for your right ventricle to pump against. High frequency percussive ventilation, high frequency jet ventilation have found a role as CO2 rescue and secretion clearance rescue. And so the benefits are probably for targeted population. There's no evidence that early intervention necessarily improves things, but there is a move to consider whether we are in fact thinking about these modes correctly as salvage versus upfront modes. And that requires specific testing and specific populations.

pediatric ARDS

nonconventional ventilator modes

airway pressure release ventilation

high-frequency oscillatory ventilation

high-frequency percussive ventilation

high-frequency jet ventilation

alveolar recruitment