Hello and welcome to today's Journal Club Critical Care Medicine webcast. This webcast hosted and supported by the Society of Critical Care Medicine is part of the Journal Club Critical Care Medicine series. In today's webcast we feature two articles from the July issue of Critical Care Medicine. This webcast will be available to registrants on demand within five business days. Please log in to mysccm.org and navigate to the My Learning tab. My name is Tamás Zagmany and I'm a Senior Lecturer in Intensive Care at Cardiff University in the United Kingdom. I will be moderating today's webcast. Thank you for joining us. Just a few housekeeping items before we get started. First, during the presentation you will have the opportunity to participate in several interactive polls. When you see a poll, simply click the bubble next to your choice. Second, there will be a Q&A session at the conclusion of the presentations. To submit questions throughout the presentation, type into the question box located on your control panel. Third, if you have a comment to share during the presentation, you 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 the five minutes to complete the evaluation. Your feedback is greatly appreciated and will help us to improve. Please note, this presentation is for educational purposes only. The material presented is intended to represent the approach, view, statement or opinion of the presenter which may be helpful to others. The views and opinions expressed herein are those of the presenters and do not necessarily reflect the opinions or views of SCCM. SCCM does not recommend or endorse any specific test, physician, product, procedure, opinion or other information that may be mentioned. And now, I would like to introduce today's two presenters. Dr. Kara Savro is an assistant professor in the Department of Community Health Sciences with crossed appointments in the Departments of Surgery and Oncology. She's a member of the O'Brien Institute for Public Health and the Arne Charbonnet Cancer Institute. Dr. Savro holds a PhD in Community Health Sciences from the University of Calgary and completed a postdoctoral fellowship in Implementation Science within the Department of Critical Care Medicine at the University of Calgary. As a Head and Neck Cancer Health Services Researcher with the Olson Research Initiative, Dr. Savro's research focuses on improving the quality and safety of care for patients with head and neck cancer, especially those who undergo major surgical treatment to optimise the health of these patients. Dr. Savro has a particular interest in leveraging routinely collected administrative data to evaluate the quality of healthcare and the application of knowledge translation and quality improvement methodologies for improving healthcare delivery. Our second presenter will be Dr. Max Martin, who is currently in his first year of Pulmonary and Critical Care Fellowship at the Mayo Clinic in Rochester. He's originally from Kansas and attended the University of Kansas for both his undergraduate studies and medical school. He then transitioned to Mayo, where he recently completed three years of internal medicine training. His research interests focus on acute exacerbations of interstitial lung disease, particularly how they are managed in the intensive care unit. Thank you, Kara and Max, for joining us today. Before we begin, could you each tell us if you have any disclosures to note? Kara? I don't have any disclosures to note. Max? I do not have any disclosures either. Okay, I will now turn things over to our first presenter, Kara. Thank you, Tomas, for that kind introduction. I'm excited to be here today to talk about a multicenter cohort study that I did as part of my postdoctoral fellowship, and it looks at adverse events after the transition from ICU to hospital ward, and specifically the seven days following ICU discharge. All right, to start things off, I'd like to get a survey of everyone understanding of the frequency of adverse events among ICU patients. So, how common are adverse events among ICU patients? The first option is less than 10%, the second option is 10 to 19%, and the third option is 20 to 50%, and the fourth is greater than 50%. Okay, I'm going to start with the first question. How common are adverse events among ICU patients? Okay, interesting. So, most people chose 20 to 50%, but a good proportion also chose greater than 50%. So, the correct answer is somewhere in the 20 to 50% range, depending on the ICU patient population and the methods that are used to evaluate adverse events. And to be clear, what I mean by an adverse event is an unintended negative consequence of healthcare, not the patient's disease process or progression. In a general hospital population, about 7.5% to 10% of admissions experience an adverse event. So, they are fairly common. And 90% of admissions are fairly common. And these adverse events result in about 98,000 preventable, and that's a key word there, preventable deaths per year. So, obviously, they compromise the health of patients and sometimes result in death, but they also cost the healthcare system considerably. So, some slightly older data now showed that patients who had an adverse event cost the healthcare system almost $7,500 more than patients who did not have an adverse event. And critically ill patients are particularly vulnerable. As we found out from the poll, between 20 and 50% is the estimated number of adverse events among critically ill patients compared to 7 to 10% in the general hospital population. And indeed, the rate of adverse events does vary within the hospital between clinical patient populations from 3.2% up to 20% between departments. A recent retrospective cohort study from our group found that 25% of critically ill patients experienced an adverse event during their hospital stay. 6% were readmitted to the ICU, and 7% died in hospital, further showing that critically ill patients are particularly vulnerable to safety events. So, we know critically ill patients are vulnerable to safety events, and we also know that transitions in care, such as the transition from ICU to ward, are very challenging periods in care that are vulnerable to errors and breakdown in the quality of care. So, transitions in care, which is any change of responsibility from one provider to the next or between environments, are associated with communication breakdown, which can lead to medical errors or adverse events, overuse, and inappropriate use of diagnostic testing. So, it's not entirely unsurprising that patients become quite dissatisfied with their care during these periods of care, and that they cost the healthcare system. So, taken together, the critically ill patients are a vulnerable population, and transitions in care are vulnerable time in care. So, we were interested to see how frequent adverse events were in critically ill patients as they transitioned from the ICU to a hospital ward, and factors and outcomes that are associated with these adverse events. So, we leveraged a larger prospective study from our group that looked at the transition from ICU to a ward with a very heavy focus on communication. So, we used the 451 patients who were included in that study and did a retrospective chart review to identify, within the seven days following ICU discharge, how many patients had adverse events. To do that, we used two independent blind physicians to review the charts, and a third physician was available to adjudicate any discrepancies between the data collected from the two independent reviewers. We used a standardized data abstraction form, and then we used the surveys that were collected as part of the prospective study to look at hospital characteristics, as well as physicians' perspectives on the quality of the transition, as well as risk of having an adverse event, ICU readmission, or hospital death. So, in the end, we included 10 medical surgical ICUs from every major centre within Canada. The data was collected between July 2014 and January 2016. We included adult patients, so over 18 years of age, that stayed at least 24 hours within the ICU, and we only included patients that were transferred within the same hospital. So, if a patient was transferred from the ICU to a different hospital, they were not included. The primary outcome was adverse event, and we had a very clear definition in line with the definition I gave before, with three key components, that the healthcare was the causation, not the patient's disease, that there was harm, and then also the type of adverse event was another primary outcome. We also looked at the preventability and severity of adverse events. We looked at ICU readmission, hospital mortality, as potential other variables related to safety incidents, and then, of course, we looked at patient and hospital characteristics and the physician's perspectives as per the surveys. So, what did we find? Well, we found that about 19% of patients experienced an adverse event within seven days of being discharged from the ICU. That was 19% of the 451 patients. The most common type of adverse event was supportive care failure, which was falls, fluid imbalance, pressure ulcers, that type of thing, and that was followed closely by drug-related adverse events. More than a third, so 36% of adverse events, were considered preventable by the data abstractors. Of the 19% of adverse events, 6% resulted in death, 17% resulted in disability, and 77% resulted in some symptoms that didn't lead to disability or death. Patients who experienced an adverse event were more likely to be readmitted to the ICU, more than five times more likely to be readmitted to the ICU. They stayed 16 days longer in hospital than patients who did not have an adverse event, and were almost five times more likely to die in hospital than patients who did not have an adverse event. So, we were interested in trying to understand when patients were at a greater risk of experiencing a safety event after they were discharged from the ICU. Was it close to the discharge date, or was it sometime after? So, you can see on the x-axis the number of days after ICU discharge, so ICU discharge date was zero, and the bars represent the proportion of patients who in green had an adverse event, in orange had an ICU readmission, and in yellow had hospital mortality. So, most of the adverse events occurred within the first three days after ICU discharge. All of the hospital mortality occurred within three days after ICU discharge, and a good proportion of the ICU readmissions also occurred within three days, so that seems to be a critical period after discharge. So, we identified that adverse events are fairly common, and these lead to negative outcomes and are potentially preventable, but is there anything we can look at that predicts who will have an adverse event? So, we looked at several variables and decided we did a univariate analysis, and based on our univariate analysis included these variables in a multivariable analysis, and we found that when we're just looking at any adverse event, there was no real association with sex, but females were more likely to have a preventable adverse event than males. Age was not associated with adverse events, which is a bit contrary to other studies that have found age is associated with adverse events. Having comorbidities was associated with having adverse events, and we found within our univariate analysis that surgical patients who had surgery before ICU, so surgical patients, were more likely, more than had adverse events, and indeed those who had emergency surgery before their ICU admission were more likely to have adverse events. Also, the APACHE-2 score was associated with the occurrence of adverse events, and particularly interesting is that we found when patients were discharged from an ICU where more or at least 80 percent of the beds were occupied, they were almost two times more likely to experience an adverse event. So, staying within the same theme of what might predict adverse events, we wanted to kind of look at the almost like the diagnostic utility of ward physicians' predictions of adverse events, ICU readmission, hospital death, ICU physicians' predictions, as well as those patient variables that we found to be associated with adverse events. And here you can see the ROC curves looking at how well patient variables and the two types of physicians were at predicting. You can see for adverse events that patient variables, ward physicians and ICU physicians, were equally poor at predicting whether a patient would have an adverse event or not. For ICU readmission, physicians, regardless whether it was ward or ICU, did a better job at predicting ICU readmission than patient variables. Conversely, for hospital death, patient variables were a slightly better predictor than the physicians. Unfortunately, none of, neither the patient variables, the ward physicians or ICU physicians did a particularly good job of predicting any of these variables at the time of transfer. So, in summary, adverse events are common in critically ill patients. One third of adverse events are preventable and one fourth result in disability or death. Adverse events increase ICU readmission, length of hospital stay and hospital death. Most adverse events, ICU readmissions and hospital deaths occur within three days of ICU discharge. APOCHE-2 comorbidities and ICU bed capacity were associated with adverse events, but adverse events, as well as ICU readmission and hospital death, were really hard to predict at the time of transfer. So, what does this all mean? So, adverse events clearly are occurring fairly frequently among critically ill patients who are transferred from ICU to ward, and a third of these can be prevented. So, focusing on those that are able to be prevented is a key, is a key feature for future safety initiatives to improve the safety of these patients. Since most adverse events occurred within three days of the ICU discharge, that presents a time period for, for maybe some additional support for critical patients as they transition out of the very resource and intensity unit to less resource intense units. And also when discharging patients during a time where bed, ICU bed capacity is high, patients who have a high risk of patients who have a high APOCHE-2 score of comorbidity, maybe we need to give pause and think about, is it appropriate to discharge these patients? And these are all areas for the focus of future initiatives to improve patient safety. And now I'll turn it over to Max. All right. Thank you, Cara, for that wonderful presentation. So, my talk today is going to focus on my research in exacerbations of interstitial lung disease. And the title of my recent publication was Mechanical Ventilation and Predictors of In-Hospital Mortality in Fibrotic Interstitial Lung Disease with Acute Respiratory Failure, a cohort analysis through the paradigm of acute respiratory distress syndrome. And before I begin, I just want to give a quick thank you to my mentor, Dr. Teng Moa, who provided me with an incredible amount of guidance and expertise on this project. So, we're going to start out with a case here. So, we have a 72-year-old gentleman. He has underlying idiopathic pulmonary fibrosis, and he's admitted to the ICU with acute respiratory worsening of unclear etiology. And because of his progressive respiratory distress, the decision is made to intubate and start mechanical ventilation. And so, I ask, in light of this patient's underlying lung disease, do you agree with the decision to intubate him? And our answer choices are yes, no, or you would like more information. Okay. So, everyone wants more information. Very good. And if we move on to one more polling question, and that's, so which factors do you feel are predictive of in-hospital mortality, particularly in this situation? And so, answer choices include initial FiO2, a P-to-F ratio within three hours of mechanical ventilation initiation, targeted plateau pressures less than 30, overall net volume status, or all of the above. Okay. Very good. So, and I'll hopefully give you some light on those answers as I go through this talk here, but so my objectives for today are that I want to discuss how mechanically ventilated patients with fibrotic ILD and acute respiratory failure were managed at our particular institution, and then identify which factors influenced in-hospital mortality. And that includes underlying demographic characteristics and specific ventilator strategies. And then I also want to review the evidence-based management strategies that go along with ARDS, and then determine if our patients, those with fibrotic ILD, were managed in a similar fashion. So, some of the background kind of behind our study was that we knew that patients with fibrotic ILD who undergo mechanical ventilation, their mortality is extremely high. The data on this that existed previously typically ranged anywhere from 85 to 100%. And, you know, somewhat because of this, there's been some thought in current practice that maybe it's best to just avoid or delay as long as possible the initiation of ventilation for these particular patients with the thought that perhaps it's just a futile initiative. And with that in mind, that kind of got us thinking about how we identified that there are several similarities between acute exacerbations of ILD and ARDS, and particularly with the definitions. And so, they have the similarities as of acuity, superimposed pulmonary infiltrates, lack of volume overload. The ILD part just lacks the severity grading based upon PDF ratios. But otherwise, their actual definitions are fairly similar. And they also have similar histopathologic findings of diffuse alveolar damage on biopsy. And so, that got us thinking about the management, and that there's lots of literature that exists on ARDS management, but there's not a lot out there about how these fibrotic ILD patients have been managed in the past and how they should be managed in the future. And so, what we did was we created a retrospective cohort review. And what we wanted to do was we wanted to assess the ventilator characteristics of all the patients that were, had fibrotic ILD and underwet mechanical ventilation, and were admitted particularly for acute respiratory failure. And the purpose of this was to determine whether or not there were any associations within hospital mortality. And our hypothesis was that if a ARDS type of model was pursued, perhaps those patients had better outcomes. And so, we included a total of 111 patients that accounted for 114 admissions that spanned over about an 18-year period, so from 2000 to 2018. And our inclusion criteria was that we included anyone with new or established diagnosis of fibrotic interstitial lung disease. They had to be admitted to the ICU for the particular reason of acute respiratory failure, but it could be of any etiology. And then they had to undergo non-elective intubation, so they had to be intubated for the particular reason of respiratory failure, and so not for example, like a procedural indication. And so, taking a look at some of the baseline demographics here, I know this is kind of a busy slide, but just to highlight some things. So, the average age of ILD diagnosis was around 63. We included more males than females. About 60% were smokers. 27 of them had biopsy-confirmed ILD, whereas the rest were clinically diagnosed. We took a look at underlying or previous pulmonary function testing, those that were most recent prior to their admission, and included that information of note. Their median FVC was around 55. And then we also did a breakdown of their particular ILD subtype. And so, about 35% of the patients had idiopathic pulmonary fibrosis while the rest had a non-IPF ILD. And then this table shows their ICU admission characteristics. So I think some of the highlights here include, so we looked at a Charleston comorbidity index score to try to get a sense of their underlying disease severity. And that median number was four. And then we also broke down their admission type. As I said, we included all types of admissions. And when you break that down, about 65% of them had an idiopathic acute exacerbation, followed closely by, or 21% was due to an infection. And then you can see the other reasons listed here. And then we also, just a few other things here. For example, bronchoscopy, 87 of the 114 admissions underwent bronchoscopy in the ICU, but only nine of those actually had a positive finding, whether that was signs of infection or of hemorrhage. And then we also included other data on use of nitric oxide, paralytics, diuretics, vasopressors, use of proning, high-dose steroids, just a few other things there. This next table here looks at the, so these are the mechanical ventilator characteristics. And so some of the ones to highlight include, so the median time from admission to intubation for these patients was 10 hours. And that was from when they hit either the ED or the ICU to when they were intubated. And that did include a pretty wide interquartile range of anywhere from two to 57 hours. Their median P to F ratio taken just prior to intubation was 122. A majority of these patients were managed on CMV or continuous mandatory ventilation mode in a volume control setting. And then here, this is important to highlight. So 85% of the patients met an initial low tidal volume strategy. So a six cc's per kg strategy. However, only 50% of them met a early targeted plateau pressure strategy, or defined as a plateau pressure less than 30 centimeters of water. And then we also included initial FIO2, PEEP, peak pressures, mean airway pressure, and plateau pressures. And this is furthermore a continuation of the same table, but we looked at whether or not their FIO2 and PEEP was able to be weaned. And a good percentage of the patients were able to wean their FIO2, but not necessarily the PEEP. We also looked at overall volume status at 2448 and end of mechanical ventilation time. And those were significantly positive at all stages. We looked at, so the total amount of time that patients were on the ventilator, the median was 4.8 days. And then this is kind of our mortality data here. So our overall inpatient mortality was 75%. And then there's a little bit of a breakdown as far as whether they were terminally extubated, extubated but died within 30 days, and then whether or not they underwent tracheostomy. But overall, so 29 of the 114 admissions were able to discharge from the hospital. So that's all our data and kind of moving forward with how we analyze this data. And so in order to determine predictors of mortality, we did both a univariable and multivariable logistic regression. And what we did was we wanted to create a theoretical causation model beforehand. And reason being is that we wanted to do direct and specific comparisons. And we wanted to avoid kind of data mining trying to find predictors that are not really there. And so we created this theoretical causation model that I'll show you on the next slide. But basically what using that, we identified four variables that we wanted to correct for. And those included age at admission, P to F ratio at presentation, RVSP and time to intubation. And this is the model that we came up with. And basically what it shows is that we have an exposure and we have an outcome. And so our exposure is mechanical ventilation and our outcome is death. And we took all the potential variables and kind of divided them into whether they're a confounder, a proxy confounder or a mediator. And a confounder would be ones that are associated with both the exposure and the outcome. A proxy confounder is kind of between the confounder and the outcome. And then mediators are those that are directly between. And so based on this, that's how we decided to go ahead and correct for age, P to F ratio, RVSP and time to intubation. So continuing with our results here. So these are our univariable or unadjusted predictors. And we actually found that there were several predictors for in-hospital mortality. And I went ahead and bolded those here on the right. So those included age at admission, RVSP, time to intubation, P to F ratios across the board and then the plateau pressure, early targeted plateau pressure strategy. I do think it's notable to note that, so underlying ILD subtype, whether it was IPF versus non-IPF was not predictive as well as the admission type, whether it was an acute, idiopathic acute exacerbation versus a secondary cause also was not predictive of mortality. And this is, sorry, just more of the univariable predictors. And then, so now we look at, so when we adjusted the variables for what I left list below the predictive variables when adjusted, there were also several, those included the early targeted plateau pressures, P to F ratio at three and 48 hours, initial FIO2, initial mean airway pressure, pressures, paralytics, and then a net overall volume status. So several predictors of mortality, even when we corrected for multiple things. So in the discussion or, so what does all this mean, or at least what did it all mean to us? And so, there were, as I mentioned, there were several factors that were predictive of mortality some of which were modifiable and some are of course not modifiable. So, some of the ventilator strategies we can modify, whereas for example, a patient's presenting P to F ratio, we can't modify. So there were kind of two groups there, but we did note that there were some that were similar to ARDS and that includes the targeted plateau pressures, the time to intubation and the overall volume status. However, when we look at our data, we really see that our patients had some elements of an ARDS strategy, but really it was not complete. And for example, so in regards to the early targeted plateau pressures, really only 50% of our patients were able to achieve that. And there was a significant discrepancy between, so the patients who received low tidal volume ventilation, that was much higher, that was 86%. So there's a discrepancy there. And when we thought about as to maybe why that exists, we came up with several theories, but one thought would be whether or not physicians are just kind of doing reflexive dialing in of the tidal volumes and then not coming back for subsequent adjustments to meet those pressure goals. Another thought is that, due to the patient's severity of their lung injury or in their underlying fibrosis, perhaps you just can't maintain these targets, whether that be secondary to hypercapnia or hemodynamic compromise, we just weren't able to get there. And then lastly would be just simply they weren't targeted at all. A couple other examples of where we didn't follow with the ARDS management plan was only 4% of the patients underwent prone positioning. And then the median time to intubation was 10 hours, which given the severity of their underlying hypoxemia at the PDF ratio of around 122, you could argue that that might be considered delayed in that situation. So it goes without saying that the mortality in our study was still very high at 75%. So yes, it's perhaps a little bit better than previous studies have shown, but still obviously very high. And so when you think about why that's the case, is it due to mechanical ventilation strategy alone or is it due to their underlying disease? It's a difficult question to answer. Certainly their underlying disease absolutely plays a role. And it's kind of, we don't know quite yet how much the ventilator strategy plays a role, but there's a chance that it's a combination of both. Our study certainly had several limitations. First and foremost, it is a retrospective study. So difficult to make a clear causation. We tried to identify and correct for different confounders, but of course we probably missed some. And then also fibrotic ILD is a very heterogeneous disease. The disease, there's multiple disease subtypes and including there's also multiple reasons to have acute respiratory worsening. And so, despite our analysis did suggest that this doesn't necessarily matter or that it wasn't predictive of mortality, but it likely still plays a role. And so looking towards the future, we identified several predictors of mortality. And we wonder that if we further targeted these variables in a way similar to ARDS, could we perhaps improve mortality? And I think that's a question that still is to be answered, but we shall see in the future. So I think that's all I have. Thank you. I will now turn it over to Tomas to monitor the Q&A session. Thank you very much for both of you. These were very interesting presentations. And there are already questions coming from the audience. So the first one is actually to you, Max. Is it worthwhile or would it be worthwhile looking into driving pressure rather than plateau pressure for mechanical ventilation in your patient population? Yeah, so I think the short answer is absolutely. I think that some of the feedback that we've gotten is that driving pressure would be an important variable to look into, and I completely agree. And I think if, you know, I'm hopeful that we still might use our data and expand on it. And if we do, we'll certainly look into driving pressure. And we've been talking amongst our department about looking into respiratory power and a few other variables. So yeah, I completely agree with you that I do think driving pressure is an important variable that I do wish we would have looked into, but we're certainly gonna consider it in the future. Thank you very much. Another question for Cara is about how could we or should we record these adverse events routinely in the hospital? And does it happen in Canada? That's a great question. Should we? I think the answer is probably yes. Could we, I think is maybe a different answer, maybe a more pragmatic answer is the way that we've collected data on adverse events for this study is very time-consuming and not really practical from a health system perspective. I think we could do it if we had a way of measuring these adverse events from data that's already collected and have a more streamlined approach to measurement. And so the retrospective study that I mentioned earlier on adverse events that we published some time ago did use some algorithms that are ICD-10 based because in Canada we're using ICD-10. So if we have mechanisms like that to measure, I think that it is feasible that we could measure them, but I think it's really also important to make sure that we do something with the data. So if we're going to be collecting data, it's important that we feed it into a system that can try and improve the quality and safety of care for ICU patients. Is it being done in Canada? Fairly inconsistently, I think. I think every kind of province and health system has a different approach. I think fairly universally across Canada is there are some reporting systems, but I think a lot of times the concern and question and disillusionment with those is that there's not a lot of feedback of the data to people who can actually make changes. Thank you very much. Another question for Max. On the slides, you showed that about 50%, 48% of the patients had high-dose steroids. Did you find any signal that it might be working towards benefit or indeed harm? Yeah, so just to review, yeah, we had, I think 42% of our patients did receive high-dose steroids when they were in the ICU. And that was not, when we did our analysis, it wasn't predictive of mortality. So I think as far as the benefit versus harm question, I think it's almost something that you have to look at on a case-to-case basis. I don't know if I can say broadly whether or not we saw harm or benefit. I can't say that for sure, but I think individually on a case-by-case basis, perhaps I would think that some patients showed some improvement, but at least within our study, the use of it was not predictive of mortality. Thank you. And another question is, it is a sort of broad question to you, Max, again. What is the incidence or prevalence of ILDs because you have presented 100-odd patients over an 18-year period in one health center? What would it mean for others? Yeah, so, yeah, it's worth pointing out that even to get, say, 111 patients, 114 admissions, we had to look at a pretty big period, so over an 18-year period. And so I think at least at our institution here at the Mayo Clinic in Rochester, I think you could say that perhaps on average we see about one acute exacerbation of ILD per month. Granted, I think perhaps some medical centers in larger cities would see more. And so I think that overall, we may see more and more of this in the future, though, particularly as we're able to hopefully improve our outpatient management of ILD and perhaps as we're able to extend patients' lives, and we're also getting better at diagnosing the disease, that I think we could see more of this. So I suspect that in the future, we're gonna see more patients in our ICUs with fibrotic ILD, and I think particularly more in the bigger cities. So I think the incidence and prevalence could increase in the future. Thank you. Back to Cara. Do you think that the in-hospital critical care services, which they've got different names, Outreach, MAT, other teams, et cetera, could they help to reduce the frequency of these adverse events that you have seen? Yeah, so I think in theory, yes. I think the evidence is slightly inconclusive, and I think it depends also on the type of outreach and how well it's used. So we did look at the use of MAT among the hospitals and patients who had adverse events and did not, and we didn't find a difference there. I know there is a systematic review from one of my co-authors, Dan Niven, looking at the benefits of some critical care support services that suggest that there are some benefits, but then there is also evidence to suggest the opposite as well. So I think there is a potential for them to improve the care. I think it's probably more about how they're implemented and how they're adopted within the setting that makes a bit more of a difference. Thank you. And yeah, the follow-on question would have been that how, by what mechanism they can work. But I agree that it's probably the implementation of these services which would do the difference rather than a specific mechanism. Yeah, I think how they're viewed and if they're actually used. So I think that was brought up in some of the literatures when you're looking at them. It depends on if there is ICU follow-up within the wards, if it's not viewed as useful, then it's not going to be used. And of course then it will not have any benefit. But I think there definitely is an argument from our study as well as some evidence from other studies saying that that first kind of 72 hours after ICU discharge is a very critical period in the care for patients who are critically ill. And I think there probably is a need for additional support. Another way to look at improving that transition from ICU to ward is maybe there needs to be a pause in discharging these patients. Maybe they need to be staying in the ICU longer. Now there isn't any evidence to support that, but that could be an area for future research. Thank you very much. I would say that there is some circumstantial evidence that it might help. In the UK, we've got a terrible problem with delayed discharges from the critical care units. And one of the recent ICNARC data analysis showed that actually those patients who were experiencing a delayed transfer of care from the ICU to the ward, they had better outcomes. So, you know, it might be that we just need to keep them a little bit longer. I don't know. It's really interesting, yeah. I'd like to put Max on the spot a little bit. And sorry about that, but this is a question that given the high number of these patients with a potentially end-stage flare-up and the high mortality, which you have seen, how much of this you could think is down to ventilation or, you know, what would be your effect size if you would do a trial? Yeah, so that's a great question. And I think that that, you know, really gets to the heart of the study here. And so, you know, I think, you know, what you're getting at is that there's no doubt that these patients at baseline are sick. And we saw that based upon their Charleston comorbidity scores, as well as, you know, their underlying PFTs. So we know they're not coming to the ICU with perfect lung function. And so, you know, it's difficult to say whether a better ventilator strategy could have prevented mortality. I think in many patients, the answer is definitely no. I think they're, unfortunately, their underlying disease is too much that no matter what we do, you know, we're not gonna be able to change that outcome. But I, you know, I do think that not all patients had, at least within our study, you know, not all patients had end-stage disease. You know, several were admitted for potentially reversible causes. And so I think you, of course, have to look at it on a case-to-case basis. And I think if it's something that's, you know, within the patient's wishes and, you know, I would argue that a trial of mechanical ventilation is reasonable. And I think if that's the case, then, you know, perhaps our paper can give a little bit of guidance to that or at least spark some thought about how these patients could or should be managed. And again, I think in some patients, the perfect ventilator strategy still wouldn't have prevented a poor outcome, but perhaps in others, if we could at least optimize their strategy, then we could give them the best chance. So if that answers your question. Yes, it does, it does. Thank you very much for both of you for the presentations and for this really good discussion at the end. This concludes our Q&A session. And I would like to thank our presenters and the audience for attending. Again, everyone who joined us for today's webcast, we receive a follow-up email that will include another evaluation. Please take the five minutes to complete the evaluation. Your feedback is greatly appreciated. And on a final note, please join us for our next journal club on August 27th for a discussion on articles about the impact of acute respiratory distress syndrome. And this concludes our presentation today.

adverse events

ICU discharge

supportive care failure

ICU readmission

hospital stays

mortality rates

comorbidities

mechanically ventilated patients

in-hospital mortality