Hello, and welcome to today's Journal Club Critical Care Medicine webcast. This webcast hosted and supported by the Society of Critical Care Medicine as part of the Journal Club Critical Care Medicine series. This webcast features two articles that appear in the October 2022 issue of Critical Care Medicine. This webcast is being recorded. The recordings will be available to registrants on demand within five business days. All you need to do is simply log into mysccm.org or mysccm.org and navigate to the My Learning tab. Hello, and my name is Tony Gerlach. I'm a clinical pharmacist at The Ohio State University Medical Center here in Columbus, Ohio, and I will be serving as today's moderator. Thank you for joining us. Just a few housekeeping items before we get started. There will be a question and answer session at the conclusion of both presentations. At any time during the presentation to submit a question throughout, just type it into the question box located on your control panel. If you have a comment to share during the presentations, you also may use the question box for that as well. And finally, everyone joining us for today's webcast will receive a follow-up email that will include an evaluation. Please take five minutes to complete the evaluation as your feedback is greatly appreciated. Please note that the disclaimer stating that the content to follow is for educational purposes only. And now I would like to introduce today's presenters. First, we have Alex Conroy, who completed his medical degree from the University of Montreal in 2012. He graduated from a specialty in emergency medicine in 2017 and completed a clinician scientist fellowship in 2018. And finally, he obtained his PhD in clinical research in 2021 and has a particular interest in pre-hospital resuscitation and out-of-hospital cardiac arrest prognostication. Next, we have Paul Syed, who received his medical degree in 2013 from the Medical University of Graz. In 2020, he became a specialist in anesthesiology and intensive care medicine. He received the European Diploma in Anesthesia and Intensive Care, along with a doctorate in medical sciences from the Medical University of Graz. And later this year, he received his European Diploma in Intensive Care Medicine. Thank you both for joining us today. Now I'll turn the presentation over to Alex. All right. So thank you very much for having me today. And I'm really glad to present to you this article, which wanted to describe the accuracy of the initial rhythm to predict a short no-flow time for patients with out-of-hospital cardiac arrest. I have got to disclose that I'm receiving financial support from my university department of emergency medicine, and that our local group has participated with in-kind support. This does not cause any conflict of interest regarding these results. I also have to disclose that the fact that the data that we used for this article came from the NHLBI, and precisely the ROC Epistry 1, 2, and 3, and from also two specific trials. The NHLBI nor the person that performed these initial studies had any involvement in the one that I'm presenting today. I also have to disclose that I just got married and that I didn't have much time as such to prepare these slides. So there are less pictures and animation than I would have liked, but I still hope that you're going to enjoy the presentation. So out-of-hospital cardiac arrest is an important topic. It's difficult to argue against. Over 400,000 cases of out-of-hospital cardiac arrest happens every year in North America only. Amongst these patients, only around 10% will survive to hospital discharge, the majority of which, despite that, will have a good neurologic outcome. For a selected patient, extracorporeal resuscitation or eCPR has been proposed to improve the outcomes but it must be used in selected patients. And eCPR eligibility mostly depends on two things. The initial electrical rhythm is used by most jurisdictions to select patients, so patients with an initial shockable rhythm are preferred over those who have an initial PEA, despite them being sometimes considered. And almost all jurisdictions will not perform eCPR on patients with an initial asystole. The other most common variable that is used to select patients for eCPR is the no-flow time, which is the period from the cardiac arrest to the initiation of chest compression by either a bystander or a paramedic. Obviously, patients who have a non-witnessed out-of-hospital cardiac arrest have a non-known no-flow time. This leads patients with a non-witnessed cardiac arrest to almost never be considered for eCPR. However, we could make the hypothesis that patients with a shockable rhythm, when observed, necessarily have a short no-flow time, as we know that ventricular fibrillation degenerates relatively rapidly. Knowing this would necessarily impact prognostication and eCPR selection, because if we could show that the presence of a shockable rhythm has a good sensitivity and specificity to identify patients with a short no-flow time, we could simply ignore the no-flow time for patients with an initial shockable rhythm, and this would allow patients with a non-witnessed cardiac arrest to be considered for eCPR. However, there is conflicting evidence on that matter. Tanguay-Rieux published in Resuscitation in 2018 a study, and they observed that the presence of an initial shockable rhythm had a positive predictive value of 94% for a no-flow time under 10 minutes, which is relatively long, but still a positive predictive value of 94% is great. And they concluded that, yes, patients with an initial shockable rhythm should be considered for eCPR, even though they have a non-witnessed cardiac arrest. Our group published, using data from the Montreal area in 2021, a similar study. And in that study, we did not look at the positive predictive value. We looked at the likelihood ratios. And we basically observed that there was a minimal prognostic value of having an initial shockable rhythm. It was not helpful at all to differentiate patients for a short no-flow time. Patients for a shorter no-flow time that time. We aim to identify patients with a no-flow time of less than five minutes. So this might explain the difference. The statistical analysis explained that difference. Despite that, this evidence did not go at all in the same direction. And this is why we decided to perform a larger study, in which we aim to determine the sensitivity, specificity, and likelihood ratios of an initial shockable rhythm to identify patients with a short no-flow time. So of five minutes or less. Using the large data from the Canrock registry, we included out-of-hospital cardiac arrest patients and excluded patients younger than 18 years old. Patients who had obviously dead criteria or were not treated by EMS. Those who received bystander CPR, because we wanted to evaluate the natural evolution of rhythm, which would be the same in patients with an unwitnessed arrest. And we also excluded unwitnessed arrests because we could not calculate the no-flow time in that population. We also excluded patients who had missing data regarding the specific timings that we were interested in, notably the no-flow time. And in these patients who had a witnessed arrest without cardiac compressions by bystander, we add a proxy value for no-flow time. So the time from the dispatch call to the first EMS electrical rhythm assessment. So this aimed to describe the natural evolution of rhythm, which we could then use for patients with a non-witnessed arrest, because for these patients, there would also be a natural evolution of rhythm. So to make clear, we use the initial rhythm as a diagnosis test, as a diagnostic test to determine if the no-flow time was shorter or longer than five minutes. So the main outcome was a no-flow time of less than five minutes. And we use the statistics that I spoke of earlier. So sensitivity and specificity notably. We also perform supplemental analyses using specific no-flow times of shorter and longer no-flow times, so that clinicians can have access to a broader range for that variable. And we also performed different analyses, including and excluding patients with an EMS witnessed out-of-hospital cardiac arrest. We performed these analyses differently for this population, because we noticed that they do not have the same evolution of rhythm as patients who have bystander or witnessed cardiac arrest. It doesn't seem to be the same pathology and same drive for their electrical rhythm. And we wanted to analyze them differently. So when I speak about sensitivity, I want to speak about the number of patients with an initial shockable rhythm, so A, that would have a no-flow time. So sensitivity would be A over A plus C. I'm sorry for these statistics, but it's going to make the things much more easier when going forward. And specificity would be the number of patients with an initial non-shockable rhythm with a no-flow time over five minutes, over the amount of patients with a no-flow time over five minutes. Amongst the registries, in the registries, there were over 200,000 patients that we all screened for inclusion. And we were able to include over 28,000. 11,000 of them had an EMS witnessed out-of-hospital cardiac arrest, and therefore a no-flow time that was set as zero minutes. 695 had a bystander witnessed out-of-hospital cardiac arrest and a no-flow time of less than five minutes. And over 16,000 had a bystander witnessed out-of-hospital cardiac arrest and a no-flow time of over five minutes. So as we can see, these are the groups that we described earlier. We did not pool the EMS witnessed out-of-hospital cardiac arrest, the EMS witnessed OHCA with the bystander witnessed one because their statistics generally differed, notably in relation to the initial rhythm. Regarding no-flow time, so we can see that EMS were set at zero. And the vast majority of patients who had a no-flow time of less than five minutes had relatively high no-flow time. So it was not dispersed perfectly in that category. And amongst patients with a no-flow time of over five minutes, the average was at around 10 minutes. As I said earlier, there was a lot of differences between EMS witnessed out-of-hospital cardiac arrest and the other patients in relation to the initial rhythm. Indeed, the rate of initial shockable rhythm was much lower in that population, in that group, which influenced our main results. The proportion of shockable rhythm was also higher in the patients who had shorter no-flow time than in patients who had higher no-flow time, which was reassuring regarding our hypotheses. The rate of initial pulseless electrical activity was also much higher in patients who had an initial, in an EMS witnessed cardiac arrest. Regarding outcomes, we could see that the rate of pre-hospital ROSC was very high amongst all group, and that patients who had a shorter no-flow time were more likely to survive than those who had no-flow time. Were more likely to survive than those who had a longer no-flow time, which was also expected. For our main results in these slides, I'm going to present to you the sensitivity, specificity, and likelihood ratios in that order for our analysis. I'm first going to talk to you about the analysis which excluded EMS witnessed cardiac arrest patients, so only patients who had bystander witnessed cardiac arrest. I'm speaking to you about these first because this is the population in which we will apply our results in the real life. They're the one that we're going to try to, we're going to use our results to determine if they or not have a short no-flow time. So, regarding our main outcome, which was for a determination of a no-flow time of less than five minutes, adding a shockable rhythm at a sensitivity of 48% and a specificity of 67% to identify patients with a no-flow time of less than five minutes, which gave a significant but not impressive likelihood ratios of 1.45 and 0.77. We can see that if we try to identify patients with longer no-flow times, the sensitivity tends to lower while the specificity tends to increase, which is to be expected as the rate of shockable rhythms tend to decrease in time. However, this does not influence much the likelihood ratios, either positive or negative. When we included EMS-witnessed cardiac arrest patients, we observed completely different results. The sensitivity was at only 25% and specificity was still at 67%. And the positive and negative likelihood ratios were even inverted in that subcategory. This is explained by the really high proportions of patients with an EMS-witnessed arrest who had an initial PEA, and this tended to flip the results in the other direction. From a clinical perspective, if you treat a patient who has an out-of-hospital cardiac arrest, the witness story is unclear. One paramedic says something, the other has the impression that the first responders heard the family tell them something different. These stories always tend to go in a lot of directions. They might have heard the patient fall to the ground, but were not in the same room or didn't check right away. And with all that information, you estimate that the pre-test probability of a no-flow time of less than five is at 60%. If you observe an initial shockable rhythm, that would make the post-test probability at only 60%. This could or not be adequate for you, but if you wanted to ensure that the patient had a short no-flow time before considering more advanced resuscitative procedure, this might not make it for you. On the contrary, if you observed an initial non-shockable rhythm in that same situation, the post-test probability would be at 44%, which is also not a huge change from the 50%, and you'd still be uncertain. So the initial shockable rhythm has a poor sensitivity, but an average specificity to identify patients with a short no-flow time. This, the nature of the initial rhythm as such can be included in the multimodal prognostication approach that is used by clinicians to stop or continue resuscitation or to consider patients for more advanced procedures. But it must not be considered in silo and must really be contrasted to the other clinical variables that are available. And we also observed that EMS-witnessed cardiac arrest patients have a really high frequency of initial PEA, and as such are not at all the same populations as those who have a bystander witnessed arrest. And the inclusion of these patients for rhythm determination and sensitivity and specificity and determination should not be done as they're not the same population as the ones we're going to apply our results in real life. I would like to thank all of my co-authors. And there's a polling question. I'm, again, a bit sorry about my French accent, but I'm working my best. So, yeah, so the correct answer was a low sensitivity and an average specificity. I'm glad that I had the chance to present this article to you today. All right, so after this great presentation by Alex, I'm honored and pleasured to follow up on a study that colleagues and I conducted regarding the incidence and outcomes of cardiopulmonary resuscitation intensive care units. I think it's great that this has been chosen in this series together with Alex's study because it will highlight what the differences between these patient cohorts between ICU cardiac arrest and out-of-hospital cardiac arrest really are. So let's get things started. I don't have to declare any conflicts of interest regarding this topic and the presentation and study at hand, but I probably do have to make some definitions up front. Even before then, we heard about out-of-hospital cardiac arrest, so someone dropping dead in the street who obviously, as Alex pointed out, undoubtedly has to receive immediate help to even have a chance to survive. That's pretty obvious, but sick people tend to be treated medically, tend to be admitted to hospitals, so there is a baseline chance that patients may suffer from cardiac arrest within hospitals. That's what we call in-hospital cardiac arrest, and that's sometimes something that is a bit forgotten or shed less light upon, although as ICU clinicians, we're usually tasked, especially in Central Europe, to take care of both patient groups, that is, patients who suffer from out-of-hospital cardiac arrest. I myself am honoured to serve as a pre-hospital care provider just as much, so to go out in the streets and take care of patients who might be suffering from cardiac arrest, or as a member of a medical emergency team, tend to patients who suffer from cardiac arrest in the respective wards or clinics. But obviously, as critical care physicians, we treat the sickest of the sick patients, so they are at risk of suffering from cardiac arrest as well, and if so, you could argue they suffer from in-ICU cardiac arrest. The thing is, at least from my experience, we use the same tools to address all of these entities mentioned before. So out-of-hospital cardiac arrest, in-hospital cardiac arrest in general, and intensive care unit cardiac arrest in general are all addressed by what's called cardiopulmonary resuscitation. From my perspective, we perform cardiopulmonary resuscitation based on guidelines issued by European Resuscitation Council, the ERC, which all in all are in line with what the American Heart Association issues, as they are all based on the International Liaisons Committee on Resuscitation's consensus on strength of evidence and treatment recommendations. So although there may be a notion between us, we more or less do the same things. But obviously, we don't do it in the same place. So when Alex apologised for his hardly notable French accent, I do have to apologise even more for my German accent, as I'm from Austria. Austria isn't that country with the kangaroos, but it is a small country within Central Europe with German, again, as the main language. But as I said, it's a comparably small country with a population of 9 million people only, which puts us hardly within the top 100 of the world. But I'm honoured to be part of the populace of a comparably rich and wealthy country with a per capita GDP that actually puts us 14th worldwide. This is to put things into perspective. So when we talk about critical care, intensive care units in Austria, they're relatively numerous. We do support a relatively expensive healthcare system in general. So when we talk about incidences and approaches to patient care, we always have to keep in mind what's there in the first place. And this is a comparably high income country. So be aware that these results that I'm presenting to you are not necessarily translatable to your country or region of practice one by one. And what my colleagues and I were interested in is manifold. First off, as clinicians tasked with running ICUs, taking care of our critically ill patients, we obviously mean to prepare for their eventual demise, that is, do training and education in cardiopulmonary resuscitation, for example. But it's important to find out how often we actually provide cardiopulmonary resuscitation as experts in the field, hopefully, to just give us a feeling and test of how often that's actually done, how much routine we would have. And then obviously, as someone who not only has to prepare their team and their units, but has to do conversations with patients and their relatives and or caregivers about their chances, about their wishes, about their hopes, we also have to debate about what would happen if they went into cardiac arrest, give them an idea, give them an informed idea. And so we meant to inform them about possible outcomes, realistic outcomes following cardiac arrest. We try to prepare, prepare and give well-informed information to our patients, to their relatives and to caregivers. And this is why we felt there was a lack of knowledge in the very field, because ICU cardiac arrest might be a very rare occasion, actually, and thus knowledge was or is limited. But let's get on here. I hope you can still hear me. If not, just throw it into the chat. With regards to the patients we treat, just to give you an idea, again, Austria is a central European country. And this is why we, just as many European countries, have an ever-aging population. I just demonstrate here the different distribution of ages from 2001 to 2022. And you can see that the amount or the proportion of individuals aged 65 years and above has gradually increased and will still increase. So this is an ever-aging population, which affects us in critical care units. Our patients are getting older, are getting sicker, are getting more chronically sick, especially. So these are the patients we care for. These are the patients we collected, again, in a registry-based study. It's about 500,000 of them. And as is typical for patients in the hospital and in the intensive care unit, especially, they're older, on average, than those patients in the general population. So the median age was 68 years as compared to those 42 years in the general population. And there is a preponderance of male patients. That's, again, something we can see worldwide, that male individuals are more likely to require or to receive critical care medicine. As for the study duration, we chose a timeframe from 2005 to 2019. That's for reasons of actuality and for reasons of the aforementioned guidelines that are issued, as you know, every five years. So we could actually subdivide this timeframe into blocks of five years and thus compare whether there are changes in outcomes according to different approaches for CPR in that very timeframe. And as clinicians in the ICU, this is what we have to expect. So approximately per every 1,000 admissions of patients to the intensive care units, be these primary admissions, so patients admitted from the emergency departments, from the operating theatres, from the normal wards to the ICU, or be these readmissions, that is patients who once were in the intensive care unit and have been transferred to other step-down units or normal wards, but do require ICU admission once more for whatever reason. Throughout about 1,000 admissions, we see 40 events of CPR in that. So one in every 25 admissions will actually at least be one resuscitation event. I feel that's reassuring. So with regards to routine and training, both of our trainees and retraining of our specialists, ICU CPR seems to be a relatively common procedure, but there are caveats when it comes to preparation because the vast majority of these ICU CPR events actually occur on the very first day of the patient's admission to the intensive care units, whereas the frank and sudden cardiac arrest of a patient who is monitored, who is already in critical care treatment is vastly lowered. So while about 40 per 1,000 admissions is the number I would give you to ask me directly, the number of cardiopulmonary resuscitation events after the first day of admission is actually only about 10 per 1,000 admissions. So that's getting gradually less and less likely that you, as someone on call in the ICU, will suddenly be woken up to attend to a patient who suddenly suffered from cardiac arrest. This was on admission. So when we plan ICUs, no matter what, it's about filling beds, it's about availability, so we count admissions. But when we talk about outcomes, obviously the patient's interest is what he or she can do or can't do. So when we talk about outcomes, the number is a bit different. This is about patients now. These are fewer than the admissions as readmissions and first admissions were synthesized to represent the respective patients. And again, what you see here is that patients that do receive CPR in the ICU and those that don't actually match up nicely somewhat. So the age is pretty much the same, gender distribution is similar as well, but there are vast differences when it comes to the etiology of their hospital admission in the first place to the ICU admission especially. We run specialized ICUs according to disciplines, so medical ICUs, surgical ICUs, or even further specialized ICUs in Austria. So each and every ICU will treat different patients from baseline. But if you take a look at these graphs that simply depict who will have or will receive CPR in the ICU, you can see that the most important or most abundant group is actually patients who suffer from some kind of cardiovascular condition in the first place. So be these patients who are admitted to the ICU in shock or even following cardiac arrest, these are the patients that are most likely to receive cardiopulmonary resuscitation again. And mostly on the first day. Again, if we look at sudden cardiac arrest and thus receiving ICU CPR, you can see that the distribution of etiologies is much more similar to the overall population than for CPR at any time. So in actual fact, while there is a predominance of cardiovascular diseases shock that leads to CPR on day one, events later down the line may actually occur in all ICU patients apparently. Similarly, if you take a look at the group distribution, like I said, we were to divide patient groups into patients admitted for medical reasons or for surgical reasons, be these elective surgery or non-elective surgery, you can see that for any CPR, the most common, this is in patients who received no surgery, so are admitted due to medical reasons. But if it's a surprise event, it's much more likely to occur in patients who are operated on as well. So again, this is a difference between patients who were admitted for cardiovascular reasons, for shock, for cardiac arrest, and those who are not. So when we talk to patients, talk to their relatives, talk to representatives, what can we tell them? If they had cardiac arrest, if we were to resuscitate them, what was to be expected? So the raw number here is that you can expect a survival of about 34.5% in total, and that's something that hasn't changed dramatically over time in all essence. There are minimal changes from the beginning of the observation period towards 2010 and the slight degrees over then. But surprisingly, there's no real difference between the outcomes from patients who were resuscitated right from the start or who were resuscitated later down for cardiac arrest later during the ICU stay. So wherever the patient comes crashing into the ICU or already is under cardiopulmonary resuscitation or requires CPR later down the line, the numbers are equal. So although the etiologies may differ, although there may be differences in patient characteristics, the achievable outcomes are relatively similar. Now, obviously, mere survival, survival to hospital discharge, that is, is a very crude metric we know full well, and the registry this study is based on isn't focused on cardiac arrest and CPI, especially. So when we talk about outcomes in CPR research, we usually mean functional outcomes, which I really can't give definitive answers on in this study, although we try the very best and use surrogate parameters for functional outcomes. For example, we do document the airway device, if any, in place per day and the ventilation mode, if any, per day. So we can say that following CPR in the ICU, the vast majority, about two-thirds at least, up to three out of four patients after CPR will not require any ventilatory aid or will not require a tracheostomy, especially at hospital discharge. So these are good functional outcomes that are implied. We try to measure conscious state, which obviously isn't represented adequately in this registry, but we use the Rikers sedation and agitation scale as a kind of surrogate parameter. And again, the median Rikers sedation and agitation scale for these patients who did receive CPR was pinpointed at four, so these would be calm and cooperative patients. This again is apparently a good thing, and you could say that ICU CPR, from all the entities I mentioned at the beginning, is probably one of the most promising approaches, although we as clinicians may perceive things differently, as these are patients who are sick and dying in the first place, that's what if they are in the intensive care unit, and the event of cardiac arrest may actually further decrease their chances of survival. However, here's some food for thought, I hope, because when I was talking about ICU CPR, what we counted were events of CPR. Again, the registry has limitations when it comes to sensitivity. We can only record events of CPR, but not necessarily of heart arrest, cardiac death, or cardiac arrest. So if we take a look at all patients who were admitted to the intensive care unit, and who died during their respective hospital stays, only about 18% ever received CPR in the ICU. So you can see that already there is much selection going on here. So in essence, the takeaway message is, if there is physician, clinician, and patient, and relative decision making in the first place, if patients are well selected, just like Alex said in his talk on out-of-hospital cardiac arrest, comparably good outcomes can be achieved, but only then, because the vast majority of patients who will die in the ICU will do so without ever receiving cardiopulmonary resuscitation. With that said, I thank you all for your brilliant attention, and I hope that one or the other question pops up so we can get into a bit of a debate here. Well, thank you to both Paul and Alex for those great presentations. I, for one, have certainly learned a lot. And yet again, to those in the audience, feel free to use the questions box and type any questions into the comments, and I'll get to them as soon as I can. My question is really for both of these people, and I think you kind of highlighted some interesting things, both in out-of-hospital cardiac arrest and in the more controlled environment of the ICU. And really, what do you guys think is the most important thing we can do for our patients to get good outcomes after cardiac arrest? It really depends on the setting, I think. If I could go first, is access to care and access to defibrillation would be the best. Sometimes we always think it's what we do in the hospital, but really what happens in for out-of-hospital cardiac arrest patients, it's mostly what happens even before the patient is transported to the hospital. This is going to make a big difference even before the first responders get there. So rapid access to defibrillation has been shown to be the most important thing for the patients that have a significant chance to survive. So for that specific population, that would be that. But the more we go for in-hospital cardiac arrest, it's not necessarily the same answer that's going to happen, and especially not for patients who have a cardiac arrest in the ICU. Absolutely. So for out-of-hospital cardiac arrest, just like Alex said, we know it's about the chain of survival that is what cannot guarantee but allow patients to survive cardiac arrest. Only if all the chain links actually come together, survival is possible. Now that's also true for in-hospital cardiac arrest and certainly is true for in-ICU cardiac arrest. Although in, like you said, this controllable environment, this is much more easily achieved. The patient's already connected to monitor. Staff is all around. The defibrillator, like Alex said, is right next to the patient. So this becomes less of an issue, but it becomes more and more about A, preventing the cardiac arrest in the first place, and B, if it were ever to materialize, to predefine and ascertain whether the tools of CPR we know and use are actually still adequate for this patient. Because whenever you use a tool in the wrong indication on the wrong workplace, it will become blunted no matter what. Well, thank you very much. And kind of a follow-up question that at least I see in my clinical practice, which is more in a surgical ICU, is it seems like there may be a difference in cardiac arrest due to respiratory versus a true cardiac arrest, and especially like potentially PEA that you might have because you're hypoxic from a drug, for example, since I'm the pharmacist. And I also know at least here in Ohio, we do have, and through the years, an opiate epidemic. So how do you think that that plays in cardiac versus respiratory failure in your results that you see both out of hospital and in the ICU? In the study that we performed, we wanted to really focus on the patients who have a cardiac aetiology and as such excluded patients who had traumatic or obvious intoxication. So our results are really for patients who have no obvious aetiology, which we suppose it's something, a coronary event, because the intoxications and hypoxia in patients who have respiratory symptoms before collapse, they do not necessarily have the same initial rhythms, which would have influenced our results. But obviously, as the treatments differ and the recognition of the aetiology of these patients differs, their outcomes are going to be a bit different, a bit different. And surprisingly, when we look at other literature, it's the patient that have an obvious cardiac aetiology that have the best outcomes. It's not necessarily those that have a reversible cause, even though it's hypoxia or something like that. They do not tend to do as well on a long-term basis. It does have some influence. With regards to ICU CPR, I think, Tony, what you pointed out there is totally represented in the data I presented earlier on, because the patients that are resuscitated on the very day of their ICU admissions, these are the ones who are in cardiogenic shock, who are in shock of any aetiology, actually, who might have suffered from out-of-hospital or in-hospital cardiopress before. So, this is a distinct population where the aetiology, like Alex said, is mostly cardiac. But if you go further down the line, remove yourself a bit from day one, but later during their ICU stay, the aetiology or the patient characteristics become much more similar to overall ICU patients. So, it's all about treating reversible causes and addressing the issue at hand. Like you said, an opioid overdose in a non-intubated patient may turn him or her into an anoxic patient who might require CPI even. Well, thank you very much. I very much appreciate that. And this is also for both of you guys. Do you think hospital mortality is really the best outcome to study, or is there something else that we should really be looking at in patients who are having cardiac arrest, both in the hospital and outside the hospital? Paul can start this one if you want. I kind of have the feeling that we'll give the same answer anyway. So, I'll use the same wording as before. Hospital mortality is a very blunt tool. It does serve the purpose because any meaningful intervention is a very blunt tool. So, it's a very blunt tool. So, it's a very blunt tool. Because any meaningful intervention should increase and improve hospital mortality. But from patients' and relatives' perspective and point of view, hospital mortality isn't just all of it. It's just a part of good functional outcomes. So, this is why I feel that in CPR research, good functional outcome, although, again, debatable in many places, is a much better outcome to aim for. In our study, that's why we tried to circumvent these limitations of our registry to pinpoint what patients are like when they are discharged from the hospital alive. As Paul saw coming, I don't have actually anything to add regarding that. We often study hospital mortality because it's easy to measure patient is dead or not, and it does matter. But if we had all the tools and all the resources that we would want, we would most often prefer to assess functional or neurologic outcomes for patients as these are the ones that matter the most. Yeah, I completely agree with that. I think, especially when you get into big databases like you guys use, I think it's the best thing you can do. And oftentimes, people get lost to follow-up. So, I think this is a good first step. And further down the line is something that we really, really should be looking at with those people. But I completely agree with both of you. And this is for you, Alex. Did you guys try to study or study a time from getting the patient who had out-of-hospital cardiac arrest to the hospital? I know it was kind of in defined areas. And were there any differences if you found them, or was it pretty uniform? So, the timings that we were interested in was the period of time that they didn't have cardiac massage. So, our proxy measure was from the initial 911 call, which was the time of the cardiac arrest, because we only selected patients who had a witnessed cardiac arrest. So, there are sometimes seconds and delays, but when the heiress is witnessed, usually the bystander calls relatively rapidly. And we stopped our timing when EMS measured the initial rhythm. And at that moment, they started CPR. So, we really wanted to focus on the natural degeneration of rhythm when there were no CPR. If we had prolonged the observation until they arrived in the hospital, there would have been a considerable length of CPR, which would have affected our results. Patients in these, all patients in that group, almost all patients in that group, were transported to the hospital. And we have some outcomes for them. And they vary a little bit depending on the category, as we described. But because we selected patients with witnessed arrest and EMS witnessed arrest, the outcomes were generally good as compared to other out-of-hospital cardiac arrest populations. I don't know if that answers your questions. No, no, I think that does. So, thank you very much. I appreciate it. And one last question, and this one is really for you, Paul. And going through your data, I think it was interesting that over the time period you studied that there wasn't really any significant changes in mortality. And I just didn't know, as with the European Resuscitation Council CPR guidelines, did anything drastically change during that time period that you studied? And if so, what was it? And that, again, that's a very keen observation. The at least unadjusted mortality rates hardly ever differed apart from a meager change between the first block and the last block. And yes, that's actually pretty much in parallel to the Resuscitation Council's guidance, at least for in-hospital cardiac arrest. The most notable change over time within the study period was probably the drop, if you will, of atropine for asystole and pulseless electric activity that at the beginning in 2005 still was recommended, but then was left with 2010 and was no longer recommended and used during CPR, even in the ICU. Well, thank you very much. And with that, I would like to thank the audience for participating. Another congratulations. Congratulations to you, Alex, on your recent marriage. That's very awesome. And I want to thank everyone who joined us today for the webcast. You will receive a follow-up email that will include an evaluation. Please take five minutes to complete the evaluation as your feedback is greatly appreciated. On a final note, please join us for our next journal club, Critical Care Medicine, on Thursday, November 17th. And that concludes today's presentation.

out-of-hospital cardiac arrest

shockable rhythm

short no-flow time

ICU admissions

CPR

hospital survival rate

patient selection

early defibrillation

functional outcomes