OK, it's not working. There we go. I have no financial conflicts of interest to disclose. I am an author of one of the articles that's utilized in the guidelines. So I want to start off with a little bit of rationale for the guidelines. Emergency airway management is a commonly performed procedure in critically ill patients. And it is associated with a significant risk of life-threatening complications. It certainly involves very complex decision-making. And just in one of the most common approaches to emergency airway management is rapid sequence intubation. That is the administration of a sedative hypnotic together with a neuromuscular blocking agent in rapid succession to facilitate the endotracheal intubation. Now, RSI is designed to minimize the risk of some of these life-threatening complications, in particular aspiration and hypoxemia, but also needs to take into account some of the other complications related to severe hypotension and cardiovascular collapse. And despite this being a very common practice, there are a number of controversies and in-practice variations that exist related to RSI and hence the motivation for these guidelines. In terms of the methods that the guidelines used, the American College of Critical Care Medicine organized a multidisciplinary panel of 20 practitioners. These were physicians with backgrounds in emergency medicine, anesthesiology, critical care. Also include pharmacists, respiratory therapists, and there was one nurse practitioner. And these were providers with experience and expertise in airway management. The panel submitted 35 PICO questions for consideration to be considered in the guidelines. The panelists voted on those questions and ultimately 10 questions were selected for inclusion. The guidelines used the GRADE methodology and had access and highly utilized both a methodologist as well as a librarian in this process. Consisted of a literature search, evidence evaluation, consensus meetings, and then ultimately the panelists voted on these statements that were issued. And those statements were based on the quality of the evidence that was available, broken down into recommendations, suggestions, and then a couple of best practice statements. One question actually had insufficient evidence to make any statement at all. So I'm going to be focusing on the first five questions and four of those questions were related to the pre-induction period. The fifth one was peri-intubation. Question one was focused on the semi-fouler position versus other positioning. The second question was high flow or non-invasive positive pressure ventilation for pre-oxygenation. The third question was related to medication assisted pre-oxygenation. Question four was related to the issue of nasogastric tube decompression. And finally, question five focused on peri-intubation vasopressors versus fluids. Now I'm going to talk about questions one through three based on the evidence that was available to come up with a recommendation. In questions four and five, there were best practice statements that we won't be talking about today. So question one, again, this is focused on positioning. In critically ill patients undergoing RSI, is there a difference between the semi-fouler position, that is the head and trunk inclined during intubation versus the supine position with regard to the outcomes of first pass success, the incidence of oxygen desaturation, or pulmonary aspiration? Now the rationale that was utilized by the committee focused on a number of observational studies, a randomized controlled trial, as well as some other studies that were performed in the operating room or on cadavers, but not on critically ill patients. So an accumulation of that evidence, the evidence table you can see to the right here for those outcomes that I mentioned. And basically, to kind of summarize the findings, three of the observational studies found improvements in first pass intubation success with a semi-fouler positioning. The fourth observational study, there was no difference. Unfortunately, despite there being a randomized controlled trial, it was felt by the panel that it had some significant limitations. There was no standardization of the degree of positioning, intubator experience, or even the laryngoscopic technique. And as I mentioned before, there were a number of studies that were considered and utilized by the committee that came from the operating room that suggested benefits of the semi-fouler position in terms of speed of intubation, prolonged time until critical desaturation. In terms of evidence gaps, again, really focusing on the critical variables that were not standardized across studies. Again, the angle of positioning, intubator experience, and laryngoscopic technique. And certainly, the studies were limited in terms of some of those subgroups of patients where this positioning might be particularly beneficial, such as patients with morbid obesity, increased aspiration risk, pregnancy, and anticipated difficult intubation. Nonetheless, the ultimate statement from the guidelines was a suggestion for the use of the semi-fouler position during RSI. Question two, related to pre-oxygenation. In critically ill patients undergoing planned RSI, is there a difference in pre-oxygenating with high flow nasal cannula versus using face mask, bag mask, or non-invasive positive pressure ventilation with respect to the occurrence of desaturation, gastric insufflation, or pulmonary aspiration risk? Slightly larger number of studies compared to question one, including a number of randomized controlled trials. I think what we have to say about these is that they were variable in terms of definitions of severity of hypoxemia, the subgroups of critically ill patients that were examined, as well as measures of success. But what was concluded in general from the panel was that compared to face mask, high flow nasal cannula appeared to reduce desaturation rates, prolong the safe apnea times, and limit the degree of desaturation. Furthermore, and for this particular question, the panel came up with a second statement, and that was with regard to non-invasive positive pressure ventilation. And that was based on the evidence of a strong evidence base for the use of non-invasive positive pressure ventilation for preventing critical desaturation in the setting of severe hypoxemia. Again, as I mentioned before, there were some significant evidence gaps. In particular, the subgroup of patients where there was an anticipated difficult airway or prolonged laryngoscopy. Nonetheless, the suggestions from the committee were pre-oxygenation with high flow nasal cannula when laryngoscopy was expected to be challenging, and the suggestion for the use of non-invasive positive pressure ventilation in the setting of severe hypoxemia, in particular when P to F ratios were less than 150. The third question, medication-assisted pre-oxygenation. So in critically ill patients in whom RSI is planned, but the patient is agitated, delirious, or uncooperative, is there a difference between medication-assisted pre-oxygenation versus the usual care with face mask, assisted mask, non-invasive positive pressure ventilation, or high flow with regard to the outcomes of desaturation or hemodynamic instability? The evidence that was utilized by the committee was fairly limited. Two observational studies, one multi-center study from the emergency department, another one that was a pre-hospital study from the helicopter. Helicopter. Both of those studies used ketamine as the medication to provide sedation in providing the pre-oxygenation. But based on those two studies, there was at least a, there was either an improvement in oxygen saturation or not a decline with the utilization of ketamine. Again, very low certainty of evidence, and certainly there are significant evidence gaps with this statement. Things such as, what is the optimal medication for assisted pre-oxygenation? What are the safe doses? What are the hemodynamic consequences? And are there other risks, such as increasing aspiration and hemodynamic instability? But based on the evaluation by the panel, the suggestion was medication-assisted pre-oxygenation for patients undergoing RSI who are not able to tolerate pre-oxygenation by other means. Thank you for your time, and I'll turn it over to Dr. Patanwalla for the second five questions. Thank you. I'm Sid Patnawala, I am a pharmacist, and I'm based at the University of Sydney and the Royal Prince Alfred Hospital in Sydney, Australia. And I am going to present the second half of the guidelines that pertain to medications within the guidelines. So I have no conflicts of interest to disclose. I am an author for articles in PICO questions 7 and 10. So the questions that I'm going to talk about today, and hopefully I'm here to summarize and give you some insight into the thought processes behind the panel and how we came to some of these decisions. So there are three questions that I'm going to talk about that pertain to the induction agents. So there's induction versus no induction, etomidate versus other induction agents, etomidate with or without using a corticosteroid, and then the other two pertain to neuromuscular blockade. Now question 6 is a best practice statement, so I won't go into details about that. But let's just say that the background on this is that we considered, you know, would we not use an induction agent in some specific patients who were in a minimally conscious state and who are hemodynamically instable? The bottom line is that there was really no direct evidence to support that, but the panel still felt that the benefits outweighed the risks of, you know, giving, so we should give an induction agent. So that was the best practice statement. So I'll cover questions 7 to 10 in a little bit more detail. The outcomes listed on the right here, you know, if you look at them, it's intuitive in how they pertain to the question. So for example, with etomidate, we're concerned about adrenal suppression, so that's why, you know, the outcomes are hypotension and multiple organ dysfunction. And the questions related to neuromuscular blockers, it's more about, you know, first pass success and complications. So that's how the questions are set up. So question 7 is, in critically ill patients undergoing RSI, the difference between etomidate and other induction agents with respect to some of these outcomes. Now acknowledging that these different agents, although they're combined, they have different pharmacological properties, they were still combined in that question. But in the guidelines, we've sort of separated out while discussing them. So here are some of the reasons of why we came to our recommendation. On the right is the certainty assessment. It's part of the grade evidence decision framework, but it's not the whole thing, it's just one part of it, so I'll just draw your attention to that. But the rationale is, so there was no, you know, significant difference in mortality between etomidate and the other agents. And most of them had, you know, etomidate has a favorable profile, hemodynamic profile. When you look at the risk of bias assessment, they're all serious. So most of the studies had some flaws in them. And so that's why they, at the end of the day, you know, it was a moderate sort of a recommendation. There was some evidence gaps in certain subsets of patients. So if people had, you know, pre-existing HPA suppression, they were on corticosteroids to begin with, and there was one meta-analysis showing that perhaps in people with high severity of illness, they were at higher risk for poor outcomes with etomidate. But that being said, it's really hard to differentiate what are those cut points of where you make that decision. So at the end of the day, we came down to, we suggest that there is no difference between etomidate and the other induction agents, and it was a conditional recommendation with a moderate quality of evidence. So building on question eight, sort of builds on question seven, and this is, should you use a corticosteroid in addition to using etomidate? So use etomidate with or without a corticosteroid? And the outcomes being, you know, infection, mortality, base suppressor use, and these others listed. And most, so the studies, a lot of them showed overall, you know, there was no difference in mortality or multiple organ dysfunction syndrome, but there was a lot of heterogeneity in the corticosteroids used. So for example, one study would have this 40-plus hour infusion of hydrocortisone 200 milligrams over that time period, versus another one had 200 milligrams as an infusion per day. So they were very different doses. Also sometimes it was an infusion versus other times it would be a bolus dose. So there was some heterogeneity in the regimens used. So there was overall a low certainty of evidence, and the evidence gaps, you know, pertain to, there was some populations, for example, people with sepsis, with cirrhosis, there were some small trials showing that those patients could potentially benefit. But when you look at the trials, they were really small sample, and that's why when you look at this, there was some serious imprecision in a lot of this because of wide confidence intervals, and a lot of that is driven by small sample size. So we suggest against administering corticosteroids following RSI with etomidate. So remember that beyond the evidence, the framework for GRADE, also we look at other things, right? So the values, the adverse effects, other risks and benefits by the panel, and we came on, and while we were sort of debating this, we came on the position that we wouldn't recommend giving, you know, people corticosteroids because they also have their own adverse effects, and the evidence wasn't that great. Okay, moving on to question nine is, so now we're shifting gears to neuromuscular blockade, and this question is, in people, particularly adults undergoing intubation, is there a difference between using a neuromuscular blockade with a sedative or using the sedative by itself? So the background here is that, so we're considering the situation where you have the cannot intubate and cannot oxygenate scenario, and so are there situations maybe where you don't use neuromuscular blockade? But when you look at a lot of the studies, when they look at first-class success rates, they were higher with neuromuscular blockade. So I've listed the ranges there, so there were some wide ranges for some of these, and that's where you see some serious imprecision here, and also risk of bias. A lot of these were observational studies, but generally it was higher success rates with having a neuromuscular blockade. The other thing was with safety outcomes. So the safety outcomes, you know, they're not reported as well. Which are the safety outcomes that are different between the different studies? Sometimes they're grouped together, sometimes they're mentioned separately. So overall it was a low, a very low certainty of evidence. And we, you know, in terms of an evidence gap, we considered that, is it really feasible in a future study looking at safety outcomes? So that's something to, how would you actually do a study, and that's something that we thought about. So we recommend administering a neuromuscular blockade agent when an induction agent is used. And that was a strong recommendation overall because of the success rates, but the quality of evidence was low. And finally, the last question is that in adults undergoing RSI, is there a difference between rocuronium and succinylcholine? This is the perpetual debate of rock versus sucks, and so that was of value to have in the guidelines. So there was no significant difference in forest path success. It was driven by, so there was a lot of observational studies, and a lot of them were small samples. But you know, there was one large clinical trial you probably saw the French study, which was a pre-hospital study where it was inconclusive basically. But there was one small retrospective study which showed a higher mortality rate in those patients, subset of patients with high severity traumatic brain injury. But we considered low certainty because of the, you know, it's an observational study and small sample. So evidence gaps, risks of patients with TBI, perhaps another future trial should consider that. The optimal dose is, it depends, so the doses might vary. And then post-intubation outcomes, because when you use rocuronium, there's the perception that someone is sedated when they may not be sedated, they're just paralyzed. So some of the studies that have looked at post-intubation sedation use have shown that there have been delays in using that sedation. So maybe a future study looking at outcomes such as awareness might be useful. So we suggest administering either rocuronium or succinylcholine when there's no known contraindications to succinylcholine. And these are the final recommendations summarized. Thank you very much. Thank you. All right, I'm Kimia Honerman, and I'm an adult intensivist. I work at McKenzie Health, which is in Ontario, Canada. I'm also a grade guideline methodologist with the guide group at McMaster University, also in Canada. I'm probably too short to see the slides, but I'll do my best here. So I'll be presenting to you the methodology and a bit of background regarding the guideline on recognizing and responding to clinical deterioration outside the ICU. And then I'll pass it on to my co-presenters to present some of the additional recommendations. The presenters have no conflicts of interest to disclose. We'd like to acknowledge our wonderful group of panelists, which included patient and family advocates for their contributions to this guideline. And in addition, I should mention that I was serving as the methodologist for this guideline together with my colleague, Dr. Bram Roschburg at McMaster. So a little bit of background regarding this topic and why we chose to tackle this. When critical illness occurs in patients that are outside the ICU, it's important to promptly recognize and respond to these patients in order to improve their outcomes. Some centers have established rapid response systems, which may include early warning scores, may include rapid response teams. There's some variability in the availability of these rapid response systems and in their composition. So in this clinical practice guideline, we're providing evidence-based recommendations on the early recognition of critical illness in non-ICU patients, as well as approaches to rapidly respond to clinical deterioration. So in terms of methodology, I'll try and be very brief here. Our 25-member panel, which included patient family representatives, generated eight PICO questions pertaining to this topic. We performed a vast systematic review of the literature, which addressed each of the eight PICO questions. We then used well-established grade methodology to generate recommendations using the evidence to decision framework, which takes into account not just the evidence, certainly the evidence, but also takes into account values and preferences, resource requirements, equity, among other factors. In terms of our meta-analysis, where we were not able to meta-analyze data, we summarized the data narratively and we were able to use that narrative summary in generating a recommendation. In terms of our final recommendations, they can be categorized as per grade into for or against a particular intervention, and within each of those groups, separated into strong recommendation, which indicates that the benefits largely outweigh any harms or risks, or conditional recommendation, where the risk-benefit balance is a little bit less certain, but still indicates that clinicians and hospitals should consider applying that particular intervention. I'll review briefly two of our recommendations, and then I'll pass it on to my co-presenters. The first question, which was actually initially started out as two questions, was related to early warning scores and rapid response teams. I'll summarize the recommendation here. We recommend hospital-wide deployment of rapid response systems, as in rapid response teams or medical emergency teams, for non-ICU patients that includes explicit activation criteria for obtaining help from a designated response team. This was a strong recommendation based on moderate certainty evidence. The challenge that we ran into, so we did review six RCTs that were included for this question, as well as 112 before-after studies, so we really tried to get at this answer and tried to answer, if possible, the question of early warning score versus rapid response team and the relative contribution of each separately, but unfortunately, given the heterogeneity in the study designs and interventions, we were not able to tease apart the relative contributions of each of these components, so we generated a combined statement. In terms of the outcomes that led us to this recommendation in part, just briefly, we did find that mortality rate is likely one that favors activation criteria with a rapid response team, so there is probably a lower mortality rate when these interventions are implemented, and in addition to that, cardiac arrests outside ICU were also pointing towards lower frequency or lower incidence, rather, when it came to this intervention, and this is just based on the RCTs, the observational data was very similar to this, and the other question that we tackled was the role of patients, families, and care partners, informal care partners. We generated a good practice statement that reads as follows, patients, families, and care partners of hospitalized patients are able to recognize subtle differences in clinical status that may signify deterioration and should be empowered to alert appropriate personnel, including the rapid response system. We felt it was important to generate this statement because it was highly advocated by our patient and family partners, as well as our clinicians, and it's important to respect and value the contributions that patients, families, and care partners can provide and the insights that they have. We also had a related recommendation, which was a conditional recommendation. We suggest that patient, family, and care partner concerns be incorporated into hospital early warning systems, and this was based on low-certainty evidence, based on five studies with very different methodologies and variable findings, but we still felt, the panel still felt that it was important to generate this conditional recommendation. I'll now pass it on to my colleague to present the remainder of the recommendation. Good afternoon. Hi, I'm Dr. Frank Sabat. I've just retired after 44 years of clinical practice, but I'm still doing research and process improvement, originally from Northern California, Redding, California. And I'm going to, this afternoon, basically go over a couple of guidelines as it relates to earlier recognition of clinical deterioration outside the ICU. So, which button do I push? Okay. So, vital signs. This was the first topic that we decided to tackle, and you might wonder, why in the world would you tackle something about vital signs? And I'm going to jump to the bottom first and then go to the top, which is our recommendation was, is that ward staff should make a good faith attempt to accurately and timely obtain an accurate, and the operative word here, I should have underlined it, is an accurate vital sign assessment, when it's ordered, when there's additional cause of concern, and to report significant abnormalities to the appropriate clinician. Why was this even needed? I mean, the word vital is in the sign, so why do you have to have a recommendation about this? Well, the problem is, is that they're frequently inaccurate or incomplete, and the literature supports that, and as a result, leads to failure to recognize key signs of decline, and thus, leads to increased failure to rescue. And their importance is self-evident. So, the conventional vital signs above the line here, temperature, heart rate, blood pressure, respiratory rate, mental status, the problem with the first three is at least temperature and blood pressure is that they're late. So they're accurate because they're automated, and the automation is pretty good for these, but they're not abnormal until very late in the course of clinical decline, with the exception of heart rate, but it's so nonspecific, it's not very sensitive. On the other hand, the sentinel signs of decline, the very earliest that the patient starts to experience a deterioration in which their respiratory rate starts going up or down if they're on narcotics, or their mental status changes, they get anxiety, agitation, apathy. But these are missed by the instruments of GCS and AVPU in terms of the mental status, and in terms of respiratory rate, and I'm going to show you some data on the next slide, that basically the respiratory rate is not measured. It is estimated with a bias towards normality, therefore leading to late detection of elevated respiratory rates, only when the patient's sick enough that somebody goes in to actually count it. Now capillary refill, it's not a vital sign, but it's a much better estimation of perfusion than blood pressure, and I would advocate that it become a vital sign. So the problem with respiratory rate measurement. On this graph you see on the left, these are the manual respiratory rates determined, and they're all clustered, and by the way, this is almost 37,000 patients that these vital signs were done on, and they're clustered on 18, 20, 22, 24. If you superimpose in the gold bars, the graph on the right, the automated respiratory rates that were accurate done by RespiroSense, a particular instrument that does this, you can see that the rates at 18 and 20 were grossly overestimated, and there weren't anywhere near the number of patients that were really breathing those rates, but as you go further to the right on the graph, you can see that the respiratory rates at 22, 24, 26 are grossly underestimated by manual method as compared to the automated method. Now what's the problem with the mental status scales? Well, first of all, GCS was never intended to be used on the general wards. It's a coma scale, and we don't have very many patients in acute coma on the general wards. It was never validated, and it just somehow got there. AVPU, which is used by many hospitals, is a very blunt instrument. It only detects whether you're alert or you respond to verbal pain or you're unresponsive. A solution to this is some hospitals are modifying the RAS scale, which does pick up some of the earlier signs of mental status changes, or you could just add three As to the AVPU scale and add agitation, anxiety, and apathy, and you would pick up these subtle signs much earlier. Our recommendation regarding clinical education dovetails with the prior recommendation because we suggest that bedside clinicians, the nurses actually caring for the patients, should receive education, whether it be focused or not, on how to recognize earlier clinical deterioration. In clinical trials where they did this, there were improved outcomes, and there was one cluster randomized control trial in 20 before and after demonstrating benefit to patients. There was heterogeneity in the format and structure, and the education provided, nonetheless, there was benefit. Continuous vital sign monitoring, we make no recommendation regarding this particular intervention in unselected patients, and unselected is the operative word. Only in patients with continuous intravenous morphine, in tidal CO2, continuous monitoring has shown that you'll benefit, but in unselected patients, there is no benefit demonstrated. And you have to weigh the continuous monitoring against the downside of cost, sleep interruption, alarm fatigue on the part of the nurses. The fix? Well, the fix is back to the bedside, back to the basics, from my own personal bias. Automation is going to help, but I think better bedside clinical assessment and more time at the bedside is likely to go much further into improving this situation, and in addition, when you're doing those assessments of mental status, capillary refill, respiratory rate, you're at the bedside to glean more information. This often leads to more satisfaction of the physician, certainly more satisfaction from the patient, and significant clinical value. Thank you. Hello, everybody. My name's Randy Wax. I'm an intensivist at Lake Ridge Health in Durham region, just outside of Toronto, and an associate professor in the Department of Critical Care Medicine at Queen's University. I'm going to be speaking about a few of the additional recommendations out of the guidelines, particularly focusing on the characteristics of rapid response systems. No disclosures as well. So one of the perpetual questions that comes up when you're designing rapid response systems is who should actually be on that team that responds? And so we did ask a few questions about this. One of the questions we were asking, is there a difference between having a nurse respond versus what we called a prescribing clinician, and in the studies we saw, typically it was looking at a nurse practitioner. And we were able to make no recommendation, despite looking at only two observational studies, and they looked at hospital mortality, hospital length of stay, unplanned transfer to a higher level of care, and there was really no difference, but again, limited data. Another question was, should a physician be attending versus not having a physician? There were only three observational studies, and these really looked at a physician defined as a resident or an attending physician, depending on the study, and in all three studies, the non-physician they described as an alternative was actually a nurse practitioner. And again, looking at outcomes of hospital mortality, cardiac arrest, and hospital length of stay, could not really detect any significant differences there, and were really unable to make any recommendations. So unfortunately, based on the current literature, we can't answer this question, and it may be because there's no difference between these different providers, or maybe we just haven't studied it properly. But at any rate, this is a question that may be an ongoing question in the future. We also looked at the role of palliative care and how that fits into rapid response systems. So one of the questions we asked, if you have palliative care providers, trained providers as part of your rapid response team or medical emergency team, will that make a difference? Unfortunately, this is a great question, but no answers because there's no literature to tell us anything about that one, so we were unable to give really any recommendation on that. However, there was a question, as part of your rapid response system, should you provide education or guidance for the clinicians outside of the ICU on the wards about how to elicit patients' goals of care and making sure that they're able to establish treatment plans that will help guide their care if they're deteriorating, hopefully in advance of deterioration or even during that process while you're making the decision about a disposition for the patient. And there was one randomized control trial and two before-after studies, and we were able to make a suggestion that, in fact, a rapid response system should be including that as part of their system with conditional recommendation, with low certainty evidence. But overall, looking at these studies, it did favour that. And just to give you a bit of a flavour of what that meant, some of the outcomes we were looking for when this intervention was provided, it did look like there were potentially increased rates of reevaluation of the documentation of goals of care and increased rates of changing patient resuscitation status with this focused education for staff outside of the ICU. It did not increase the rate of involvement of a formal palliative care consultation or service and unclear effects on ICU length of stay. But overall, it was felt, you know, with the conditional recommendation, this is something that many rapid response systems should be looking at if they're not doing already. And then there's a number of different interventions suggested as part of these studies because they weren't just sort of a single intervention that was multifactorial, but things like trying to make sure you knew who the substitute decision-maker was, discussing what the plan of escalation would be if the patient got sick, various kinds of education initiatives or standardized documentation. So there's lots of different initiatives, but in general, the recommendation is still people should think about it and see if they can potentially influence what's happening outside of the ICU before they have to come to the ICU. So we're bringing the right patients that should come and providing the appropriate care to patients who shouldn't come to the ICU. There was also a good practice statement that came out of the question, is there evidence to support quality improvement strategies or initiatives as part of rapid response systems? Lots of studies, 23 observational studies, but very heterogeneous, kind of three big buckets of some of these quality improvement strategies or interventions, looking at organized plans for education and structured implementation of rapid response systems, having some sort of audit or feedback mechanism and providing feedback to clinicians, and also enhancing teamwork and communication strategies. And again, it was a bit all over the map, but there's a number of different metrics that many rapid response systems use, could look at the number of activations of their teams, look at delays in activation, look at cases of patients who end up in the ICU without prior activation, that was probably warranted, and also looking for patients who had a non-ICU cardiac arrest that we know from the literature in many cases, those patients had something that would have helped predict that and maybe an earlier intervention would have helped. But we also recognize as a panel that the resources in different hospitals, depending on how big the hospital is, what kind of system it is, whether you're in a developing country with more limited resources, it may differ in terms of what you can actually do as part of quality improvement interventions. So in places that don't have sophisticated databases or systems, you can still look at cases and do individual case audits and still learn a lot from individual cases that will still help tweak your rapid response system. But if you have the opportunity to develop a database, you can identify trends and look at numeric trends over time, particularly things that whenever you're looking for mortality or ICU admission, you're adjusting for patient volume over time, and these opportunities are there. And finally, again, we got great feedback from our patient and family representation in the guideline group. When you're trying to think about quality improvement, what outcomes should you monitor, what's really important for patients and families, what interventions are important, and in particular, if you do have patient and family participation in the activation of the rapid response teams or MET teams, and it was mentioned in one of the other recommendations that we think incorporating them as part of the early warning system was a recommendation, we should be getting family and patient input on that. And just to wrap things up, so really our guideline looked at sort of what we call the forearms of the rapid response system. There was the afferent and efferent arms, so how do you detect, how do you respond, the quality improvement piece and overall structure of the rapid response system. Lots of questions still to be answered, and so we look forward to feedback from people after reviewing the guidelines. In particular, a lot of interest in the detection part and early warning systems and lots of new data that's come out since we developed the guidelines, and so I suspect there's going to be a lot of work looking at that in the near future. And thanks to our panelists on the guideline panel, and we'll turn it over to Dean. Thank you. the podium and the glycemic control comes up with the first speaker. I just want to remind you that all these guidelines are available on SCCM.org. It was really fun the first time, so I asked to come back. Sorry. Oh, I see. Okay. Oh, perfect. All right. Thank you. And I'm still the same height as the last time, so. All right. I won't introduce myself again, but I do have the pleasure of discussing the methodology we used for the guidelines on glycemic control for critically ill children and adults. And thanks again to SCCM for this opportunity, as well as for supporting the guideline. The disclosures for myself and my co-presenters are listed here. And we'd like to thank our task force members, expert group of panelists who contributed to this guideline, as well as our patient and family partners. So I'll just be speaking with you briefly about methodology without trying to reiterate too much from what you've already heard. This guideline was an update on the 2012 SCCM guidelines related to the same topic. In this iteration, we had 22 panelists, which included two patient and family partners. We generated six PICO questions, each pertaining to pediatric population as well as adult populations. We conducted systematic reviews of the literature between the year 2000 and 2023, and then used grade methodology as standard grade methodology to generate the recommendations. In addition to looking at our main analysis, which was all critically ill patients, we were able to separately look at certain subgroups, including mixed ICU populations, patients post-cardiac surgery, and neuro ICU patients. And so we'll discuss those findings separately where it's relevant. The strength of the recommendation is as per grade approach. So if the desirable effects of an intervention clearly outweighed the undesirable effects, it was a strong recommendation. And if it was more equivocal, then it was a conditional recommendation. We did generate good practice statements, mostly when the net benefits were large and unequivocal, and the message was necessary and clear and actionable. So if it met the good practice statement criteria as listed here. There was one unique aspect to this guideline in terms of methodology that I just wanted to emphasize. When the evidence was insufficient to make a recommendation in either direction, we generated a no recommendation statement as a standard. However, in this guideline, we also generated in our practice statements. So these are summaries of what the panel experts tend to do in a situation where they're faced with that particular intervention. These are intended only to describe current practice patterns of our panel experts, and they are not graded statements and are not intended to be recommendations per se, but only to provide some additional guidance for practitioners. And here are the outcomes that we looked at, long list of outcomes that the panel rated in terms of their importance from a patient's perspective. And we looked at each of these outcomes as were available for each of the PICO questions. And I'll pass it on to my co-presenters. Thank you, and thanks to the journal production staff and our reviewers for helping us get this paper published on Friday as a head of print in critical care medicine, both electronically as the full document and as an executive summary. For our adult population, we had four statements that were graded recommendations and two good practice statements. We felt that it was important to identify a trigger for paying attention to hyperglycemia, but data were inadequate and not separate from the titration targets. And so, in terms of identifying a trigger value, we established a good practice statement that it is still important to pay attention to hyperglycemia and set that point at greater than 10 millimoles per liter or 180 milligrams per deciliter to do something. And it may or may not be an insulin infusion. There may be ways to reduce glucose intake, alter your nutritional support, or add insulin if it is persistent. But perhaps this is an area worthy of additional research. Our most important PICO was addressing the target range when you are on insulin therapy in terms of where you set your protocol to achieve. We set a good practice statement as a part of this document for the use of protocols with a low risk of hypoglycemia, implying that you ought to know your hypoglycemia rates, and with a process to treat hypoglycemia without delay. And this was more extensively discussed in the 2012 Insulin Utilization Guidelines, and I refer you to those. We had a lot of debate as to how we would frame this recommendation and chose to suggest against titrating to a lower or tight blood glucose target versus a higher target with an emphasis on minimizing hypoglycemia. And our target ranges that we were comparing for intensive was 4.4 to 7.7 or 80 to 139, so a little bit larger than the original tight glycemic control target of 80 to 110 versus conventional starting at 7.8 and going to 11.1. And you'll note there's a slight difference in our target. It's a bit higher than our trigger to be a little bit easier managing the patient at the bedside, and with a trigger being slightly lower, avoiding excursions of glucose well above 200. As you can see from our meta-analysis, looking at the numerous reports discussing severe hypoglycemia, the data are very clear-cut that using intensive insulin increases the risk of hypoglycemia by a relative risk of almost 3.8 times more often, and so therefore it favors conventional glucose control targets. Regarding hospital mortality as really one of our most important endpoints, the data were not clear-cut to differentiate a difference between intensive and conventional glucose control. And just a quick note, we did not include TGC-FAST in this meta-analysis. It had not yet been published. And thus, our takeaway is that conventional glucose control, which still addresses hypoglycemia but potentially reduces workload for our bedside caregivers and hopefully reduces hypoglycemia, is the most important PICO. Regarding monitoring frequency, we looked at less than one hour, and so it could be continuous or near-continuous monitoring compared with more than one hour intervals. In particular, when the patient is having glycemic instability, it doesn't mean that this will be your target throughout the entire course of insulin infusion therapy, and found no difference in hospital mortality, ICU mortality. There was, in a small number of studies, a difference in renal replacement therapy that favored frequent glucose monitoring. And regarding ICU length of stay, that favored less frequent glucose monitoring. But these were small studies with significant heterogeneity, leading us to a conditional recommendation or suggestion that you monitor hourly or less during periods of hyperglycemia or glycemic instability. Looking at administration method for your insulin therapy, comparing insulin continuous infusion versus intermittent dosing, we suggest continuous insulin infusion versus intermittent. Again, very low certainty of evidence and not well studied. I'll now turn this over to Ellie Hirshberg to discuss our pediatric statement. Thank you, Judy. I'm Ellie Hirshberg. I am an adult and pediatric intensivist at Intermountain Healthcare and a professor of medicine at the University of Utah. How do we go forward? There we go. Okay. I'm sorry. I messed that up completely. I'm not doing anything now. Oh, you're doing it for me. Great. All right. Just left. I apologize for that. So the pediatric statements followed the adult recommendations fairly, or the adult questions fairly significantly. We came up with two graded recommendations, two good practice statements, and two questions that resulted in no recommendations and in our practice statements. Our first was the glycemic target range with an insulin infusion. We came up with a good practice statement where essentially we recommend, or a good practice statement where we recommend against titrating to a lower blood glucose target versus a higher blood glucose target range. There are two good studies that really support that evidence, and our good practice statement aligns with the adult statement to use protocols with a low risk of hypoglycemia and to treat the hypoglycemia without delay. For the administration method, infusion versus intermittent, there's really no data in the pediatric literature that supports whether or not we use continuous IV insulin versus intermittent or subcutaneous. However, the panel consisted of experts who really only use continuous infusion. So in pediatric population, in our practice, we recommend the pediatric expert panels say that they use continuous infusion of insulin for the control of blood sugars. Monitoring frequency, again, there was no data in terms of what was superior, more frequently less than an hour or every two hours. However, our experts said that essentially they typically in their practice preferred to use continuous or near-continuous monitoring for glucose if available, and that that was helping prevent the incidence of hypoglycemia. The last question was the use of an explicit decision support tool, and we provided a recommendation statement where we suggest the use of an explicit decision support tool. That was a conditional recommendation with a very low certainty of evidence. However, we were very clear about defining what an explicit decision support tool was. It had the following features, explicit recommendations so people who were following it knew exactly what to do with reproducible actions, and that there were two or more patient-specific input variables so that the protocol could provide patient-specific care, two or more output variables, and an open-loop system in case there were needs to intervene. So this is the summary of the adult and children glucose guidelines, and you can see they're fairly similar. The biggest difference from the PED standpoint is no recommendation regarding the root of insulin and no recommendation regarding the frequency of BG monitoring. Thank you.

emergency airway management

rapid sequence intubation

glycemic control

critically ill patients

GRADE methodology

pre-oxygenation methods

continuous insulin infusion

hypoglycemia prevention

glucose monitoring