All right, it's 8.15, I'm going to go ahead and get started. My name's Todd Rice. I'm the Associate Editor for Critical Care for the journal CHEST, and this is a CHEST journal session. The goal here is to try and have some of the cutting-edge critical care research presented at CHEST, and it doesn't have to be research that's published in CHEST. It can be research that's published anywhere, and you'll see today we have two topics that are not published in CHEST. We actually were supposed to have a third topic, which ended up not getting to fruition and it's not being presented today, so we'll have two topics. But I think this actually works quite well, because I think at the end it will allow us a significant portion of time to actually talk about outcomes and what we hope for and what we look for in critical care outcomes or outcomes from our critical care trials. And so I think it'll lead to some robust discussion. I got a whole list of things I'm supposed to tell you. First of all, the sessions can be evaluated through the mobile app and or the online program. The presenters will verbally disclose any financial interests, conflicts of interest that they have. I think those are the big ones. CME will open on Wednesday at noon, and then will be open for a while after that. So without further ado, let's get started. We have two trials. One is the patch trial, and the second is the device trial. Paul Young is going to present the patch trial for us. Paul joins us from New Zealand. Many of you probably know Paul or his name. He was just telling us that this is a great place to come from New Zealand, because it's 23 hours difference. And so it feels exactly like he's in New, well, maybe not exactly like he's in New Zealand, but from a time standpoint it feels like he's in New Zealand and doesn't have jet lag and that sort of stuff. So Paul's going to walk through patch trial and the results. And then Matt Precker is going to talk about the device trial. And then unless you have a trial-specific question, and we can take those after each one, at the end we'll do open questions and we'll talk a little bit about outcomes and what outcomes do we care about. Thanks, Todd. Hello, everyone. Pleasure to be here. And Todd's asked me to talk today about the patch trial, and in particular about this really interesting issue of what should guide clinical practice change. Should it be mortality or survival with good neurological function? Now I have no conflicts of interest to declare other than being an investigator on the trial, which I suppose is an academic conflict of interest. And so what I'm going to do in this talk is I'm going to give a bit of background about TXA and trauma, I'm going to talk about the patch trial results, and then I'm going to focus on that question of having done the trial and found what we found, what should we actually do? And so the effect of in-hospital TXA and trauma was evaluated in the clinical randomization of an antifibrinolytic in significant hemorrhage crash two trial. And in that trial, they found that TXA administered within three hours of injury reduced 28-day in-hospital mortality. Now the crash two investigators did an excellent job of simplifying their message. And this is it. Systolic 100. Systolic 9. Give me blood. A plot stabilizer given within three hours of injury reduces the risk of bleeding to death by 30%. Check out the crash two trial. Isn't that lovely? So it's probably worth highlighting that the trial compared TXA with placebo given up to within eight hours of injury, but subsequent analysis showed that it was really only within three hours that it was effective. Of course, the reality is that the message is not quite as simple as is made out in the cartoon. The cartoon's really effective at getting across the idea that you just give TXA and then you don't bleed to death. But the thing is that in the crash two trial, most of the patients who were enrolled were enrolled in low and middle income countries. They were enrolled in centres where they didn't have CT scans. They didn't have blood banks. They didn't have really access to operating theatres to arrest bleeding. And they had little or no access to interventional radiology. And so they basically did not have established trauma systems that we see in advanced centres. The primary outcome in the crash two trial was in hospital mortality at 28 days. And patients were not followed beyond hospital discharge. So while we know that TXA reduces early mortality due to bleeding, its effect on survival beyond 28 days is not known. And many patients who suffer severe trauma will survive with severe impairments. And so we were really interested in the question of whether TXA improved survival with a good functional outcome. Another sort of important issue is that although TXA wasn't shown to increase the risk of DVT, PE, and myocardial infarction in the crash two trial, many of the sites involved in the trial really didn't have robust systems to detect these events. And sort of layered on top of this, in the HALT-IT trial, which was a trial of TXA in gastrointestinal bleeding, there was no reduction in bleeding, but an increased risk of DVT and PE. So the sort of concern that in an advanced trauma center, TXA might not actually be that useful and might cause harm, we thought was a valid and important issue. So the patch study was designed to test the hypothesis that TXA initiated pre-hospital in advanced trauma systems would increase survival with a favorable functional outcome at six months. We enrolled patients with severe trauma who were at risk of acute trauma-related coagulopathy. So this is a condition which occurs early after massive trauma, when you have shock, hypoperfusion, and vascular damage. And the mechanisms for this abnormality include activation of protein C, endothelial disruption, and depletion of fibrinogen. So we used this score, the COAST-3 score. And a score of greater than or equal to 3 indicates a group of patients identified pre-hospital who have a high risk of trauma-related coagulopathy. Now, based on previous work, we used this score of greater than 3 because this captured a high-risk population. And we sort of reasoned that by doing so, we would enrich our population with the people who were most likely to benefit from the treatment, because these people would have a high risk of having fibrin breakdown. And in the end, we actually succeeded, I think, in enrolling a group of really very unwell trauma patients. So this injury severity score that you can see here, this represents a very severe degree of injury. More than a third of these patients actually received red blood cells before they got to hospital. It is notable, I think, that there's relatively little penetrating trauma in this trial. So here are some more data from table 1, just focusing on the physiology. So the first thing I just want to point out here is that just over a third of the patients had an initial GCS of less than 9 when they were assessed by the ambulance. And these were generally patients who were showing pretty significant hemodynamic changes, right? So they had a high heart rate, had a low blood pressure. And so this is a group of trauma patients who are bleeding. So we randomly assigned patients to either receive a gram of tranexamic acid as a bolus pre-hospital, followed by an infusion over eight hours after arrival in hospital, or a matched placebo. And here again is the study hypothesis. So TXA initiated in pre-hospital and advanced trauma systems would increase survival with favorable functional outcome at six months. Before I get on to the primary outcome, I'm going to start with some of the secondary outcomes. And as you can see, this is the cumulative mortality plot for TXA versus placebo through to six months. And I think it's pretty obvious that there is very early and wide, and it's also statistically significant separation of the survival curves. So these are the secondary outcomes in tabular form. And what you can see in relation to the mortality is that the mortality effect is most marked quite early on, but it's really sustained throughout the entire period of follow-up. And there's a suggestion here with the exsanguination data that really the mechanism that's responsible for the survival benefit is stopping people from bleeding to death. And so I think that in the end, when you look at the data from PATCH and you look at the data from CRASH-2 and from other sources as well, I think probably we're at the point where you have as high a degree of certainty as you have about anything in medicine that TXA reduces the risk of early bleeding-related death. I don't think there's much doubt about that. But really interestingly, I think, when you look at our primary outcome, survival with a favorable functional outcome at six months occurred in 53.7% of patients who got TXA, and it occurred in 53.5% of patients who were in the placebo group, for a relative risk of 1.00. If you look at the colors up there, you can see that there are differences, right? So in our trial, for every 100 patients who received TXA, there were an extra four patients who were alive at six months. But also in our trial, for every 100 patients who received TXA, there were four extra patients who were severely disabled at six months. They were in the category of lower severe disability, which corresponds to requiring full assistance with all activities of daily living. So our conclusion in the trial is that TXA decreases deaths, but also increases fully dependent survivors who need help with all activities of daily living at six months. There are some important caveats. So the first caveat is that the primary outcome data were missing for 13% of randomized patients, mostly due to losses to follow-up. But analyses that account for such missingness, they basically had similar findings to the primary analysis. Similarly, around 16% of patients in both groups received open-labeled TXA in breach of the protocol. But a per-protocol analysis, again, yielded similar results to the main analyses. Now, it remains possible that TXA will prevent death from bleeding in some patients who will go on to subsequently make a good functional recovery. The problem is that in our trial, we can't identify those groups. So if you look at the subgroup analyses, and remember that this is pre-hospital TXA, right? So trying to determine, for example, if a patient has a brain injury in the pre-hospital setting is actually exceptionally difficult, because people who are in severe hemorrhagic shock without a brain injury are often unconscious. And so I guess if you look down that plot there, there's really no groups where we could identify a favorable functional outcome being increased with TXA. It is notable that TXA, given beyond two hours after injury, there's this sort of signal that maybe it might be harmful. Again, sort of emphasizing the point that came out of the CRASH-2 trial that TXA needs to be given early. I think it's notable that the point estimate favors TXA for penetrating trauma. As I emphasized earlier on, we didn't have a lot of patients with penetrating trauma in the trial. And so I think the risks and benefits of TXA in that group do remain quite uncertain. The other thing that we found was that TXA did not significantly increase vascular occlusive events. And it's really important to emphasize that we were very systematic about screening for these events in the patients in this trial. I think if you look at that DVT point estimate, there still is a suggestion that perhaps TXA does increase the risk of a DVT somewhat, but the effect is not large. So getting back to the sort of crux of the issue, the thing that Todd actually asked me to talk about, well, I mean, what should we do, TXA or no TXA? And I think, in fact, that this is just a heavily value-laden question. So the patch trial provides really informative and useful data, but the question of implementation remains, to my mind at least, very complicated. There's a famous poem by Wallace Stevens called 13 Ways of Looking at a Blackbird. And it's like you look at the blackbird really closely and you see the reflection of the landscape around it and its little eye. And it's sort of a poem about the importance of perspective. And so while I don't have an answer about whether mortality or survival with a good functional outcome matters most, I have tried to consider a number of perspectives. And I think the first perspective is really easy. The first perspective is cost effectiveness. It's likely that many of the patients who are in a highly disabled state at six months after injury have sustained severe traumatic brain injuries. And these are often young patients, and some will improve after six months. But many will actually remain fully dependent on others for many, many decades, and their care will cost many, many millions of dollars. And so when considering the cost per quality adjusted life year perspective, TXA in pre-hospital trauma is actually a very expensive treatment. Now the importance of this perspective is going to vary depending on where you work and things, but I think in publicly funded health care systems in particular, it's quite an important perspective. And I think it's pretty reasonable to say, actually, even in the absence of formal cost effectiveness analyses, which are still to come, that from a cost effectiveness perspective, TXA is going to lose. Probably the most critical consideration in all of this is personal values. So there are likely to be a range of perspectives. And if we take the results at face value, then some people will certainly say that they would rather die than survive in a highly dependent state. Unfortunately, this doesn't help us very much. Because people who are critically injured in situations where TXA might be considered are never going to be in a fit state to have a discussion about their values and wishes. And we also know from other areas that people often say they don't want to survive in a disabled state, but when they do, they're often glad to be alive. And so I think while all this stuff is true, I don't think it's going to help us very much in our decision making. Another perspective that's relevant to consider is the perspective that would come from disability advocacy groups. So I think it's going to be pretty hard to sell a perspective that being disabled is worse than being dead to disability advocates. And that's a valid perspective to consider. The next thing that I think is interesting to consider is that beyond the level of personal values, there are likely to be broadly held cultural norms in different places that affect people's views about whether not being dead is more important than being alive with a good functional outcome. And I guess the other thing that's relevant in terms of the overall equation and weighing these things up is that there is uncertainty. So in relation to the specific data from this trial, whatever decisions we decide we're going to take about pre-hospital administration of TXA, we need to take into account a number of uncertainties. So first, there's a spectrum of outcomes that is encompassed within the need help with all activities of daily living. So that GOSC category will have some people who are extremely disabled and unable to interact or engage with others in any meaningful way. But on the other hand, there will be other people within that category who need a lot of help but have a lot of meaning and joy in their life. And actually, we just don't have the information from the trial about which group of people ended up in that category. So the reality is that the GOSC provides a really imperfect assessment of categorizing outcomes as good versus bad. And this is an uncertainty that we need to consider when we decide whether or not mortality or survival with a good outcome matters more. We also need to deal with the uncertainty about what happens next. So we followed patients out to six months. And we have actually looked further down the track to see what happens to people beyond that six months. And basically, what happens beyond that six months is entirely predictable. So the longer that you go away from the time of randomization, the more loss to follow-up you get. And it's non-random. So the people who do best disappear. And the people who do worst disappear. And so what happens is the further away from randomization you get, the more or less precision you get about what actually happens. So there's no chance that we're going to show that there becomes a benefit in survival with functional outcome with more time. All that's going to happen is we're going to have less certainty than we have now. So we're never going to get a resolution. So, I mean, I guess the other way to decide what to do is just ask Twitter. So if you ask Twitter, then two-thirds of people say that you should give pre-hospital TXA. Thanks very much. Great. Well, thanks, Todd. Good morning, everyone. Thanks, Paul, for an excellent talk on the PATCH trial and a tough act to follow, but I'll do my best. I'm here to present the results of the device trial, review the results of the device trial, and talk a little bit about the outcome we chose as primary, maybe defend it, think more about it. And I think the fun thing about doing emergency innovation research is that it's something we all, as pulmonologists and intensivists, anesthesia, emergency medicine, really share. And it's part of our practice, everyday practice. So as you watch the talk, you'll think in your head, do I prefer one or the other of these learning scopes? What have I used before? Is Matt Freger telling the truth or not? And when we get to the end, maybe we'll have a bit of a discussion about that. The device trial was funded by the Department of Defense. I don't have any personal conflicts of interest besides I was the protocol chair for this trial and first author on the paper. We know that emergency tracheal intubation is important to our patients. It occurs 1 and 1 half million times or more annually in the US. And depending on the setting in which tracheal intubation takes place, the level of risk to the patient is much, much different. So in the ED or the ICU, this is one of the highest risk procedures that emergency physicians, intensivists, anesthesiologists perform. Yet in an elective operating room setting, when the patient's fasted, it's a relatively healthy patient with a pre-op H&P, the risk is 20 times less. The reason that failing to intubate the person in your first attempt is important is because it's been closely associated with severe complications. And failure to intubate occurs in up to 30% of our first attempts, and that leaves us a big opportunity to improve the quality of this procedure. This link between failure to intubate and life-threatening complications is shown in this graph, where taking some pooled data from the studies I list, you can see that the incidence of severe hypoxemia, aspiration, and cardiac arrest is much, much higher in patients that the operator fails to secure the airway on the first attempt. At the same time, there's two main types of laryngoscopes that are used to perform this procedure. A direct laryngoscope, which you all are familiar with, includes a handle, a blade, and a light source. This device and technique hasn't changed much in 70 years. The blade's inserted, and the mandible and tongue are displaced, and the operator's meant to directly visualize the laryngeal inlet. 15, 20 years ago, along came high-quality video laryngoscopes. And this device has the same components of a direct laryngoscope, but adds a camera in the distal half of the laryngoscope blade attached to a screen, so an operator, if they desire, can use that camera view to locate anatomic structures and guide endotracheal tube passage. So what do our society and professional guidelines say about this dilemma about what laryngoscope you should choose on the first attempt? Well, the contemporary guidelines say that either is acceptable. And interpreting prior randomized data on this topic needs to be done with a couple of caveats in mind. One is that previous trials had really modest sample sizes, 40 to 600 patients. All but one of these previous trials was conducted at a single center, so generalizability is limited. The only previous multi-center trial before device of VL versus DL was the MACMAN trial that came from seven French ICUs published in 2017, mainly French internal medicine residents, so inexperienced operators intubating. And that trial randomized a McGrath video laryngoscope versus a direct laryngoscope and found no difference in first attempt success. Since the MACMAN trial, I would argue that clinicians' experience with vitilary laryngoscopy has increased a great deal, in part due to the COVID pandemic. And so these trials from pre-2017 really need to be interpreted with the idea that a lot of operators are relatively inexperienced with this technology. Be that as it may, the INTUBE study, Worldwide Point Prevalence Study of How Emergency Intubation is Performed Outside of the Operating Room, still told us just in the last year or two that a direct laryngoscope is used for over 80% of intubations in the ED and ICU worldwide. So this is current practice, and that includes more than half of first intubation attempts in the US. And in Australia and New Zealand, where Dr. Young works, they are really the only region in the world where perhaps video is used a little more frequently than direct, according to this study. So the hypothesis of the device trial, then, is that use of a video laryngoscope will increase the incidence of successful intubation on the first attempt. So to do the trial, we engaged with the Pragmatic Critical Care Research Group. This is a clinical trial network in the US that uses 20 sites that have EDs and ICUs participating in work. This is a multidisciplinary group, including emergency medicine, anesthesiology, and critical care. And then importantly, the focus of our network is conducting pragmatic trials, looking at comparative effectiveness of emergency interventions. And the network's done over 20 trials since 2013 and enrolled 50,000 patients, so a lot of experience here. Brief review of the methods of the device trial. This was a pragmatic parallel group randomized trial, randomizing patients to VL or DL on the first attempt. We enrolled all adults undergoing endotracheal intubation in our EDs and ICUs unless they were pregnant or known to be a prisoner, unless there was an immediate need for intubation that precluded the randomization procedures, or unless clinicians decided they didn't have equipoise, that the video or direct laryngoscope was absolutely needed or contraindicated for that case. Because this was an emergency intervention with two devices that both included the standard of care, it was conducted with a waiver of informed consent. Here are the 11 medical centers participating in the device trial. These encompass seven EDs and 10 ICUs and are geographically nicely distributed around the U.S. As they conducted the device trial then, it was a very simple practical study in that allocation of randomization assignment was concealed in opaque envelopes that were kept near the area equipment at each site. Randomization was performed in a one-to-one fashion, stratified by trial site, and of course, clinicians and patients and investigators were not blinded to the intervention. When those opaque envelopes were open and randomization was revealed, this is what the operator of that procedure, the clinician performing the procedure, saw. If the patient was randomized to the video laryngoscope group, they were asked to perform the first intubation attempt with the video laryngoscope, defined it as a scope that has a camera and a screen. They were asked to look at the screen for the majority of the first attempt, rather than in the mouth. And all other aspects of that intubation procedure were up to the operator. If they were assigned to the direct laryngoscope group, they used a direct laryngoscope on the first attempt. They could choose the blade shape, and all other aspects of the intubation procedure could proceed as the responsible clinician desired. Here's how data was collected for the device trial. Interestingly, we asked the operators, pre-randomization, before the envelopes opened, for a kind of from-the-door gestalt perspective on the anticipated difficulty of that procedure. They could rate the difficulty anticipated as easy, moderate, or difficult. During the conduct of the procedure, data was collected by a trained observer, not part of the procedure. They recorded data for the primary outcome by recording the number of times a laryngoscope blade, a bougie, if it was used, and an endotracheal tube entered the mouth. They also recorded information on the duration of intubation, as well as the lowest oxygen saturation blood pressure between induction and two minutes after intubation. After the trachea was intubated, the operator reported the best laryngeal view. They reported complications that occurred, including aspiration, esophageal intubation, or dental injury. And then, also importantly, the operators reported their experience performing intubation, in terms of the overall number of intubations they'd done in their career, as well as the proportion of those that were conducted with a video laryngoscope. So compared to previous trials, I think we have pretty granular data describing the human beings that were holding these laryngoscopes and performing the procedure. I wanna highlight the primary outcome of the device trial, because this is what we may chat about a little bit later. It was successful intubation in the first attempt, but defined a little bit differently than prior trials. I'd say with a bit more nuance, recognizing that there are really two halves, or two parts of tracheal intubation. So to be successful in the first attempt, the operator had to intubate with a single insertion of the laryngoscope blade in the mouth, and either a single insertion of the endotracheal tube, or a single insertion of a bougie, followed by a single tube insertion. So it really required perfection to qualify for the primary outcome. The secondary outcome of the trial was severe complications, and they're shown here. Severe hypoxemia, hypotension, new vasopressor, cardiac arrest, or death. We powered the device trial with these assumptions. Looking at our prior work, we understood that successful intubation on the first attempt could happen about 80% of the time in the DL group. So with 90% power and alpha of 0.05, missing data less than 4%. We concluded that a sample size of 2,000 patients would allow us to detect an absolute difference in first attempt success of 5%, which is, I would argue, clinically meaningful. The primary analysis was very simple. It was just a chi-square test on the between-group comparison of first attempt success. And importantly, we pre-specified a single interim analysis after half of patients were enrolled, after 1,000 patients, and we asked the Data Safety Monitoring Board to recommend stopping the trial if the p-value for the comparison of the primary outcome was less than 0.001, according to that chi-square test. And that became very important on November 17th of last year when after eight months of enrollment, the DSMB came back at the interim analysis and recommended stopping the device trial for efficacy. So I wanna just explain briefly the trial results before I get to some comments. Here's the flow of patients through the trial. We assessed nearly 2,000 patients for eligibility at these 17 sites and randomized 1,420 patients. That's 73% of all those screened. So I think we're capturing a wide variety of patients that were intubated here. The trial enrolled for eight months, and so the average number of enrollments per month at these sites was nearly 200. So this was very brisk. And we ended up analyzing 1,417 patients. Let me show you a little bit about exclusions. 488 patients then were excluded. 373 were because one or more exclusion criteria was met. More than half of those, the exclusion criteria was the intubation occurred too urgently to complete trial procedures. And only 4% of patients screened were excluded because of lack of equipoise. Those are the 82 patients there where VL or DL was required. And looking at the patients then that underwent randomization, we assigned 705 patients to the video laryngoscope group and 712 were analyzed in the direct laryngoscope group. Three patients were excluded post-randomization because they were discovered to be prisoners. And at that time, data analysis was stopped. Here's kind of the tables of our manuscript reviewing some of the characteristics of the patients. They were very similar between groups. The median age was about 55. 34% or so were female. Given that 70% of the patients were intubated and enrolled in an emergency department, altered mental status was the most frequent indication for intubation followed by acute respiratory failure. And then importantly to me anyway, the operators pre-randomization assessment of difficulty included difficult anticipated intubation in nearly 10% and moderately difficult in about half of patients. So we were not just enrolling so-called easy to intubate patients but got a good mix. Characteristics of the operators then, I would say this is a fairly inexperienced group. The median number of lifetime intubations an operator in the trial had performed was 50. And of those, the experience with VL and DL was similar in about half of those operators. Just about 90% of operators in this trial were either an emergency medicine resident or a critical care fellow. You can see that pre-auctionation was used uniformly as was an RSI approach to these airways. The 5% or so patients did not receive a sedative or neuromuscular blocker were in cardiac arrest at the time of intubation. As a clinical trialist, this was also something that made us very happy. We had a really excellent adherence to the randomization assignment in the device trial making trial interpretation a lot easier. 100% of patients randomized to the video laryngoscope group received a video laryngoscope on the first attempt. And nearly all patients assigned to DL received a direct laryngoscope on the first attempt. For people that think a lot about intubation like me, here's a breakdown of the laryngoscopes used in the trial. In the direct laryngoscope group, the great majority of intubations were done with a Macintosh-shaped blade. In the video laryngoscope group, the most common video laryngoscope the operator selected at these sites was a Storz C-Mac device with a standard geometry blade in about 60%. And a hyper-angulated blade or the classic glidescope-style blade was selected in 14% of intubations in the video laryngoscope group. And like previous trials, the device trial confirmed that video laryngoscopy improves the view of the glottis for the operator. And that's well-illustrated here with a grade one Cormac Lehane view where most or all of the glottis is seen. That occurred in 75% of patients in the video laryngoscope group and only 46% in the direct group. But this is fine. How does this translate then into what probably is important to the patient? How soon can I get the endotracheal tube in place and begin ventilating and oxygenating that patient again? And that comes to the primary outcome of the device trial, this very strict definition of successful intubation on the first attempt. And when we stopped enrollment for efficacy at the halfway point, analyzed data on just over 1,400 patients, we found that indeed in the direct laryngoscope group, successful intubation on the first attempt occurred in 70.8%. That was compared to 85.1% in the video group, which came to an absolute risk difference of 14.3% and a very significant chi-square test. So practically that effect size, that 14% percentage point difference in the primary outcome translates to a really incidence of failure in the first attempt that's half in the video laryngoscope group compared to the direct laryngoscope group. Here's the primary outcome of the device trial shown as a cumulative incidence curve of successful intubation on the first attempt compared to the number of seconds between the start of laryngoscopy and successful intubation. And you can see video laryngoscope, laryngoscopy was superior and it was actually faster by a median of about eight seconds compared to direct laryngoscope on the first attempt. Just like the patch trial, we took a look at subgroup analysis in this trial and the takeaway from our forest plot here looking at the absolute risk difference between trial groups is that video laryngoscopy was favored in every subgroup we looked at. And I wanna give you a little more detail on two of those because I think they're quite interesting. Looking at overall operator experience, we found evidence for heterogeneity of treatment effect here where the operators that seem to benefit the most from video laryngoscope group were those that were least experienced. It intubated a total of less than 25 times. They had a benefit of over 26 percentage points in first attempt success using video laryngoscopy. But even our operators that were the most experienced that had intubated over 100 times in the past, the point estimate still favored video laryngoscopy by an effect size that was equal to or just a little bit above the one we sought to find in the trial overall. How about experience with video laryngoscopy? We know that probably went up during COVID and we had a group that had a fairly mixed experience. Well, likewise here, the take-home point is that even for operators that were the most experienced with direct laryngoscopy that had used a video laryngoscope less than 25% of the time, still had benefit from using a video laryngoscope with a point estimate there of about 10 percentage points, absolute risk difference. Here's the secondary outcome of the device trial shown in graphical form. No difference in severe complications during intubation. 21.4% in the video laryngoscope group, 20.9 in the direct laryngoscope group and the breakdown of each component of this composite also very similar. This is some additional exploratory outcomes in tabular form. Importantly, the second row there shows first attempt success in the old definition, one laryngoscope blade insertion. You can see that was about 90% in the video laryngoscope group, approximately 80% in the direct laryngoscope group. And so comparing to previous trials, our rate of success using direct laryngoscopy was very similar if not a little bit better. So I don't think the result here is explained by failure to do DL adequately. We also looked at exploratory safety and clinical outcomes and these were very similar between groups. So to summarize then, we conducted a 1,417 patient randomized trial in 17 US EDs and ICUs, found a large effect size suggesting efficacy of video laryngoscopy in terms of first attempt success. And this is important because failure to intubate in the first attempts consistently associated with life threat and complications. And in current clinical care, direct laryngoscopy is still done in the majority of intubations worldwide. So that 14 percentage point difference in the primary outcome favoring video laryngoscopy translates to a number needed to treat to prevent a first attempt failure of seven. And that's illustrated here. So the next seven patients you intubate in your unit, five of them will likely be able to be intubated with either video or direct laryngoscope. One patient will, both devices will fail. And then that seventh patient there shown in blue, you'll be able to, at least the operators in our trial would be able to intubate that patient with the video scope, but not a direct laryngoscope. So why were our results different than the MAC-MAN trial? That previous multicenter trial from several years prior coming out of France. In both trials, we found that video laryngoscopy improves the greater view. That's not a new finding. But unlike MAC-MAN, we found that that better view actually translated into better success, easier and more efficient endotracheal tube passage in the device trial. Could be explained for a couple of reasons. We used a stylet or a bougie 100% of the time in the device trial. It wasn't mandated in the protocol, but that's the practice of these sites. Whereas that use was only 16% in MAC-MAN. And then the issue of operator experience, I think should come to bear here, where most operators in the device trial had performed at least as many intubations with videos with direct laryngoscopy. And so one important question that we received presenting the trial results is how did this apply to operators practically with more or less intubating experience? And the clearest answer from the trial comes in the novice operator, maybe those residents or fellows that you all are supervising your sites if you're at a teaching place. There we could really reduce the failure rate from 50% with a direct laryngoscope to only 20% with a video laryngoscope. So that's important. Moderately experienced operator also benefited from video laryngoscopy use in our trial. And then importantly, the late career or very experienced master of intubation who had maybe done thousands of prior intubations, it's tough to know how the trial results of device apply to that type of operator. So maybe a relevant analogy would be landmark guided central line placement where if you've been doing non-ultrasound guided procedures for your whole career and you're very comfortable with that, can achieve high success, you could continue to do that. But for most clinicians, adding an ultrasound adds a margin of efficacy and safety. A few caveats or limitations of the device trial I should mention. We didn't mandate the video laryngoscope that the operators chose. They got to do that based on local practice and their own preference. So the device trial doesn't inform the best type of laryngoscope to use, but I think that's a question for future study. We had a relatively inexperienced group of operators. 97% had intubated less than 250 times. And then importantly, this was in trial of ED and ICU intubation of critically ill adults. And our findings probably don't extrapolate well to the operating room setting. So one question I was asked to reflect on a little bit was this is all great, but did we choose the right outcome? Can this primary outcome convince peers and colleagues that if you're using a direct laryngoscope, you should switch to a video on the first attempt? So the first attempt success has a few pros and cons. I would argue that it has precedence. This is the most common primary outcome of emergency area management trials in the past. So we stuck with what was used before. This procedure outcome is tied very closely to the intervention. So it makes sense that it could influence it. And then, of course, in observational literature, it's tightly associated with something patients care about, and that's severe complications. I'd also argue that first attempt success is easy to measure at the bedside. It's something feedback the operator and the team get right away, so it's an important quality marker. And I think a lot of our programs should really strive to see greater than 95 or 98% first attempt success in all our intubations. We should measure that outcome to make sure we're continuously improving. The naysayers would invoke arguments around, hey, first attempt success is really a surrogate outcome? I'm not sure it's patient-centered. Things like severe complications, death, cardiac arrest might be. And then after the trial, going back to Twitter, we got a number of comments around, why didn't you use a hybrid or composite outcome like first attempt success without severe complications, right, requiring both to be present? And we did that analysis in our trial report in table three, and it was different with about the same effect size between the groups favoring video laryngoscopy, but that difference in our trial was completely explained by first attempt success, not the rate of complications. So it's a bit misleading in that way, and that's one issue with hybrid complications. So I was gonna go through some of the statistical reasons why we didn't have adequate power to detect a difference in complications in the device trial, but I'm gonna skip this in the interest of getting to our discussion. Suffice it to say that to determine whether video or direct laryngoscopy affected severe complications, we'd have to have a trial seven or 10 times as big as the one we conducted, and that has feasibility limitations at the time we did the trial. So perhaps we can invoke Winston Churchill's favorite quote here about democracy is the worst form of government except for all the others, right, where it could be relevant to emergency intubation. First attempt success might be the worst outcome except for all the others we may consider. So interested in the group's thoughts on that to come. But we concluded that among critically ill adults undergoing tracheal intubation in ED or ICU, use of a video laryngoscope increased successful intubation in the first attempt compared to direct laryngoscope, and we published our trial in print on August 3rd in the New England Journal. And thanks again for the opportunity to present.

Patch Trial

Device Trial

tranexamic acid

survival

functional outcome

video laryngoscopy

tracheal intubation

emergency departments