All right, so I'm Ryan Maves, I'm a professor at Wake Forest, I'm an ID doctor and an intensivist and I chair the FCCS Crisis Management Committee within SCCM here. These are my disclaimers. I get some research support from Sound Pharmaceuticals, IACURUS, and Pfizer, none of this has anything to do with what we'll be talking about today. So I'm going to be talking about mistakes, about things we maybe didn't get right. And I think before you do that, I think it's honest to be apprise one's own mistakes. So three years ago at Congress, I gave a talk on COVID. And actually I had to change my slides that morning because COVID was named on the morning I did my talk. It was still 2019 NCOV that morning. And so I had this slide up like, how do we treat this emerging disease? There had been like 15 cases in the US at this point. And so I thought I'd look at that slide again and see how I did. So just kind of reviewing my own errors here. Glucocorticoids not recommended, okay, that was very, very wrong. Bacterial co-infection in 10% of admitted patients, no, that's off. Remdesivir works pretty good. Kaletra, which is an HIV protease inhibitor, does not. Standardized data collection, that was a pretty good idea. So not terrible, but room for improvement. So let's talk a little bit about what crisis standards of care is. And I think we all have an intuitive understanding of it, having lived through it for the last going on three years. So the Institute of Medicine defines crisis standards of care as a substantial change in usual healthcare operations and the level of care it is possible to deliver, which is made necessary by a pervasive or catastrophic disaster. They go on to emphasize that this decision, the transition to crisis standards of care, it is not optional. It is a choice forced upon us by circumstances. And the failure to do so, the failure to modify our regular standards, both in how we treat patients and what the standards of care are, is likely to lead to greater morbidity and mortality. So the goal of crisis standards of care then, during a public health emergency, is do the best you got with what you got, while trying as hard as you can to meet some reasonable standards of care. The goal of crisis standards of care is not to provide less care, it is not to decide who is delivered to die, and it does not serve to reduce our ethical burden. So when we think about crisis and organizing systems of care during a public health emergency, there's a sort of traditional rubric of stuff, staff, and space. And then people often put in a fourth S, which can be systems, which can be structure, it can be whatever overlying mechanism you have to control. But stuff is obviously vents, dialysis machines, medicines, PPE, staff, staff is us. Staff is us, and all of the other people who work with us, the janitors, the people who are providing food for our patients, the administrators, the ward clerk. And then space, physical ICU beds, physical ward beds, and then alternate sites of care, which can be within or without the hospital. And of course, during COVID, the availability of negative pressure became a major concern, although I think we did a pretty good job of expanding that. So within that rubric, we then have sort of areas or say, sort of zones of emergency. Conventional, during a crisis where we can maintain some manner of conventional care, we use our regular stuff. We have enough, we don't have to say no to anything. During contingency, where we're starting to burst at the seams a bit, but we're still holding it together, that's when we look to, for example, to repurpose, to modify our regular practices. And then in crisis, when we're essentially tapped out, that's when we get into issues of shortage, of triage, of finding methods for scarce resource allocation. And the problem we have faced after the pandemic, the problem, right? One of many problems we have faced after the pandemic is that we've been in a chronic contingency state, that a lot of our plans assumed a short period of emergency, right? Months, not years. So this is from a 2014 Chess paper that looked at, kind of described how this works. We're not going to read this graph, but just to show that it explains that there are certain levels of care that you try to provide and have an idea of, depending on what your demands are, whether you can consider yourself to be in conventional contingency or crisis standing. This is from a paper we did at the beginning of the pandemic. And you can see that you can move in and out of different levels here, that you can be, you know, kind of things are normal around here, and then you start to surge, but you're still doing okay. You can expand ICU capacity on average about 20% using existing resources without a lot of difficulty. But then once you start exceeding that, you start drifting into areas where your ability to provide the best possible care is compromised. And you can move in and out of that over time, depending on surges. And of course, we've all lived this now, right? I would say several years ago, this was sort of a secret wisdom that a handful of people knew. I think we've all, we all have a lot of lived experience in this. Our challenge here is let's see how disaster plans were originally designed for hospitals. When we think about conventional disasters, like an explosion or, you know, God help us, a mass shooting, casualties may be as high as the hundreds. They are largely trauma patients, but event recovery is pretty fast, right? Things go pretty fast. We, you know, it's an acute event and it's over. Now, there may be secondary issues. For example, with an earthquake, you may have infrastructure damage to the hospital that would further affect it, or you have infrastructure damage affecting your staff's homes. That can affect the timeline of recovery. But in a pandemic type disaster, in COVID, in Ebola, thankfully largely avoided in H1N1, these are in the thousands or greater. They are mostly medical patients. They don't come to one hospital. They go everywhere. And that prevents us from doing a lot of the load leveling and transferring of patients to overloaded systems when there's nowhere else to send anybody. Critical care is obviously in greatest demand. We are the rate limiting step of the hospital, and recovery requires months or longer. And we are still, quite frankly, in that recovery. So that brings us back to those three pillars, then, stuff, staff, and space. How has that been affected by having had a three-year-long emergency? The shortages we deal with, they have become chronic in many cases. Some of them, like PPE, aren't as bad as they were, and that's largely addressed. But drug shortages, like IV contrast, who ever thought we'd run out of IV contrast? It can be difficult to reallocate equipment when every area is impacted. And on the flip side, we have seen some shifting from single use to reusable equipment. For example, reusable PPEs, Elastomeric N95s, and things like that. In terms of staff, obviously, the great resignation, I think we're all fairly familiar with. Caregiver fatigue, burnout among the most heavily impacted staff, particularly, but far from exclusively nurses. The personal risks, I think we've all kind of gotten used to that, but very early on, that was a very acute concern. Now, on the flip side, is we're a lot more experienced at taking care of this now, and I think that's had a good impact on outcomes. We know how to take care of these diseases. And a generation of residents, you know, has ARDS down pretty cold. Space, we've gotten pretty good at doing space conversions, but we've also found that the limiting factor is staff, not space for us, for the majority of the time. So how have we responded to surges then? So the point of surge responses in a public health emergency is to increase health system capacity in emergency and avoid the need for crisis standards of care, to avoid the need to do those sorts of triage systems. In these contingency settings, normal standards of care should generally apply. There's different ways to do this. We can expand the workforce. For example, this is from a paper we published last year in CHEST with kind of surge response strategies. You can expand the acute care workforce by incorporating non-ICU trained staff who still have acute care experience, hospitalists, med-surg nurses, and the like. You can set up dedicated procedure teams. If your ORs are closed, you have a lot of anesthesiologists who can help you with tubes and lines. Same with surgeons, same with cardiology. Telemedicine as a force extender, as we put it in the military, a way to reach hospitals and entities where that kind of care is not otherwise available. Same thing applies for pediatrics. Early this year, and it seems to be on the wane now, thankfully, but there was talks about the triple-demic of RSV, influenza, and COVID-19 hitting pediatric populations in particular. There are about 85,000, 90,000 adult ICU beds in the United States if you count up medical, surgical trauma, cardiac, et cetera. There are 5,000 PICU beds. It doesn't take much to overwhelm pediatricians. So this is a strategy that we've described for adult ICUs to assume some of the care of older children, finding ways to spread it around to let that load be balanced. And then space, how do we repurpose space? We can use PACUs as overflow space, for example. Spaces that are configured already as critical care places, and with support staff and patient care staff who have some degree of familiarity with caring for the acutely ill patient. PACU nurses can handle an ICU patient quite comfortably for a good period of time, for example. Stuff, moving on into other big categories. So drug shortages are common normally, and they're exacerbated during times of crisis. There is a thing called the FDA essential medicines list that are recommended to be available for public health emergencies. The list is a little weird if you've ever looked at it. There are three fluoroquinolones on it. There's four neuromuscular blockade agents on it. Prednisone isn't listed for some reason. Inhaled zanamivir is considered an essential agent. It's a strange list, but it's a useful concept. What are the core drugs you need to maintain certain forms of patient care capabilities? This graph is from a paper by Lisa Burry in Ontario, who really does a beautiful job of laying out why do we continue to deal with these sorts of recurrent drug shortages. Things like stockpiling, things like just-in-time purchasing. We buy things at the last possible second, and then we're surprised when they run out. Those concepts apply to not just drugs, obviously. I think we all remember the great toilet paper shortage of 2020, for example. Other things with stuff, ventilator shortages. So we worried a great deal about running out of ventilators early in the pandemic. And then the Defense Production Act was invoked to make more ventilators. The problem, of course, was an excessive focus on things rather than capabilities. Over 20,000 ventilators were produced, most of them by companies who had never made a vent before. GM made vents. Ford made vents. And those worked about as well as a car made by Puritan Bennett Works, which is to say, badly. Only about 60,000 of those 200,000 ventilators could successfully support a patient with ARDS, which left us with this. Plenty of ventilators, too few of us. Not enough of us. Not enough people who knew how to run the vent. Not enough people who knew how to take care of a critically ill patient. Now, SCCM, the Society, has done a ton of fantastic work to extend that capability. Things like FCCS, ways to educate non-critical care trained staff to be able to support and augment our abilities. But it only goes so far, and it only lasts so long. So what happens to you? Why does this matter? Why does it matter if we have a massive surge of patients that exceeds our normal capacity? What's the big deal, right? Now, usually I have 15 patients on service. Now, I have 30. I'm pretty smart. I work pretty hard. I can probably do that. So it turns out I can't, or we can't, or we can't do it as well as we think. So there's a paper in the Annals of Internal Medicine by Samir Khadri about a year and a half ago. And what Dr. Khadri and colleagues did is look. This is a NIH-sponsored study looking at a very large database. And what they found is using this thing called a surge index, which is basically a way to standardize what is your normal hospital capacity, what do you normally do, and then how many patients do you have now, in terms of how many patients on the vent do you have, what's your ICU space, and so forth. What it turns out is the greater the extent of the surge, mortality goes up. And the odds ratio of death during the most extreme COVID surges could as much as triple compared to when we are operating at more normal capacity. And this is with the same staff. This is with the same people, equally experienced, if not more experienced, caring for it. But surge kills. That's the thing, surge kills. This is from CDC did a similar analysis. And they found a similar outcome, that with adjustment for illness severity, the higher your ICU occupancy goes, and certainly if it exceeds 100% occupancy, mortality tends to go up during COVID. So how do we cope with this? One of the strategies to deal with surge response is something we call load balancing or load leveling. The idea of having a mechanism to transfer patients smoothly from a more burdened to a less burdened institution. They're both still busy, but you're balancing out some of that load. And with that, you ideally have a mechanism for reducing mortality. So the decision to initiate these things, the decision to initiate crisis standards of care, this is a regional decision. This is a government decision. It is not a hospital decision. I read the other day about a, this is several months ago now, about a hospital in Alaska declaring crisis standards of care. A hospital can't declare crisis standards of care. That is a legal decision. It is a government decision. So what we need to do is figure out a mechanism for that sort of load balancing to reduce mortality. We did, I'm going to call our success rate spotty. So there were jurisdictions that did a fantastic job. And I would say the state of Washington, through something called the Northwest Health Care Response Network, and largely under the leadership of Dr. Steve Mitchell at Harborview, has done a fantastic job of finding ways to maintain this load balancing as best you can. If you're not familiar with the geography in Washington, there's a, you know, there's a large mountain range that runs right down the middle of the state. And particularly during the winter, that can make it difficult to transport patients from one side to the other. Actually, my uncle had to be medevaced from one side of the state to the other during peak Omicron. And that was, took some fairly heroic work by the people at Harborview to make that happen. This can also be a mechanism for sharing gear. If the problem is you have enough space, but you don't, say, have, you're short on dialysis machines, this sort of mechanism for borrowing, for sharing durable equipment can be effective. Also, loans and purchases of consumables, right? If one system had stockpiled all the rock uranium in the world and another hospital needed some, you would have a mechanism for loans, transfers, and such. So Minnesota Department of Health is another good example of a system. I'd say San Diego County, where I practiced the majority of my career, did a pretty good job internally as well. Other jurisdictions did not. Particularly, and it pains me to say it, New York City, where there continues to be an inability to communicate between health systems in the largest metropolitan part of our country. So this is just an example of, I'll just refer you to the paper. As I look at this slide, it occurs to me, there's absolutely no way anyone can see what this says. But it looks great on my laptop. So what this is, is a design of how you integrate that. What you need to make that kind of load balancing work is some central clearinghouse. Ideally, with some degree of government authority, county, state, whatever, to serve as an integrated operations center to share information. So every hospital should have an incident command system. All of those systems should then work together as part of what we call in San Diego an emergency operations center. These centers serve as a way for hospitals to communicate and health systems to communicate to move patients from one area to another, to find the best possible place to do the best possible good. Critical care has to be a key part of that. If it doesn't have... If there's not an ICU doctor and an ICU nurse on that, then it is unlikely to be successful. OK, so that gets us to triage. What happens when we overwhelm the capacity of the system? So this is a little infographic made by our dear friend Mike Christian, who couldn't be here today. So looking at what is the concept behind triage. It's very simply when demand exceeds supply. How do we address it? Well, we can address it by public health control measures to reduce demand. Masking, social distancing, things like that. We can try to increase supply by augmenting staff and stuff in space like we've talked about. Ultimately, if we're tapped out, then we need to have some manner of triage, meaning prioritizing who gets what. And that's a terribly difficult decision. So the whole point of those aforementioned surge responses is to find a way to avoid the need for triage to the greatest extent possible. Now, most of the talks about triage previously had to do with ventilators. They really focused on ventilator allocation. And it turned out we didn't run out of vents. We ran out of us before we ran out of ventilators. But any scarce resource can be limiting. Hemodialysis machines, particularly CRT circuits, were and remain a major limiting feature. Remdesivir, when it first came out and every state got a little allocation from the federal government, we triaged Remdesivir within San Diego County. And we set up a committee that I was a part of where we determined who got Remdesivir and how to prioritize that. Making that decision is best possible in the hands of the bedside clinicians. So the whole point of triage systems are to seek to identify patients most likely to benefit from critical care resources. Now, how you decide most is variable, right? Do you save the most lives? Do you save the most life years? Meaning focusing on younger people above older people. That's a challenge. And that is a little bit controversial. These systems have to apply when implemented to all patients potentially being admitted to the unit, not just patients with the disease under question, not just patients with COVID or Ebola or H1N1. The obligation remains to care for all patients, though. Even if a person is not allocated under a system like this to, say, receive a ventilator, they can still get high flow. They could still get non-invasive. They would still hopefully be offered palliative care. And that's an important thing to remember. Crisis standards of care are still standards. It is not a free-for-all. It is still a standard of care. So there are a number of ways that these things have been structured. People have used score, you know, attempting to quantify this, right, to give it a number. So long-term survival, things like your ASA score, frailty indices, comorbidity indices, ways to try to measure or predict that individual's long-term survival. These will often get incorporated with short-term survival scores, most often SOFA scores, occasionally Apache. But SOFA is, of course, a lot easier to calculate. And then there is a ethical decision to avoid what we would call exclusion criteria, although certain groups would have a lower priority for critical care resources. For example, persistent coma following return of spontaneous circulation would be not an exclusion criteria, but certainly reason to deprioritize an individual. Catastrophic intracranial hemorrhage would be another example. Then there's a question of tiebreakers, right? You know, if you have two people who are otherwise the same and there's one vent, what do you do? Pregnant women, I think, are broadly accepted as a group who would be a tiebreaker. Healthcare workers and first responders, not because there's anything special about us, but hopefully if we survive, we can then get back into the fight. Children, how do you prioritize children? That's not a thing I've ever had to think about, and I would hope to keep it that way. So what are the ethical principles that underlie this, right? Because this is fundamentally an ethical exercise. So in these triage systems, in this triage system of development, one can be utilitarian, simply say the most lives saved. If a system produces more lives saved than not, that's a good system. Egalitarian, recognizing that there are deep inequities in our societies and we should not use triage to amplify those inequities. Libertarian, communitarian, meaning that who does a particular group value? If an individual community, for example, particularly values its older members to a greater degree than another group, that group may prioritize the elderly in a way that another system may deprioritize the elderly. And then there's this whole notion of life cycle or fair innings, like saying that people have a right to grow up and graduate from school and meet a partner, and if they want to get married, get married, and if they want to have kids, have kids, and then retire, and then become a burden to their families later on, right? People have a right to go through those stages. And maybe that is sort of your cut points rather than hard age. And then again, we talked about considerations for key populations. These are somewhat mutually exclusive. So there are these triage decision algorithms you can look at. So again, this was something we published in 2020 where, you know, step one, does this person meet criteria for critical care services? Do they need a vent? Do they need vasopressors? If no, then no. If they do, do they agree to receive intensive care? You know, actually, I told the story about my uncle being medevaced across the state of Washington. My uncle had the most hardcore polst ever written. I mean, it was like, don't touch me. And it was changed because of an acute issue. But, you know, one would be perfectly justified in saying, we're not taking this patient, right? Difficult thing to say about one's own family, but it's true. And then going forward, assessing whether or not they're likely to benefit, right? Though I think we all appreciate, certainly from the pandemic, that an 80-year-old with advanced dementia is less likely to benefit from intubation and mechanical ventilation than a 25-year-old with no significant comorbidities. Overseeing this is this idea that triage needs to be separated from the bedside clinician, that the physicians, the nurses involved in triaging patients should not be the patients at the bedside. This is for two reasons. One, to reduce the moral burden on the bedside team. And two, to introduce objectivity, to try to reduce conscious and unconscious bias as a way to minimize the impact of that. And this is an example from a paper in 2008 explaining how these triage teams would be structured, how you would have outcomes and feedback, and they'd be communicating with the emergency operations center. Hospitals would be feeding them data. There's a committee. They would review the numbers. They would prioritize or reallocate events based on this. Now, how many of you at your hospitals for the last three years are aware of some plan like this? Probably most of us are aware of some plan like this. Who actually implemented something like this, where a triage team was checking in with you to let you know who you could intubate? Even when you are at your absolute peak of the pandemic, the worst surge ever, you're 10 deep intubated in the ER. You've got patients overflowing onto the wards. When did this ever happen? I'm going to make a bold statement here and say never. I don't believe this ever happened, at least not openly. So different states and different jurisdictions have different methods for their ventilator allocation. So all the ones with color incorporate something like SOFA score or an MSOFA score into it, saying that as part of our decision on how we're going to decide if someone gets a vent or not, we're going to use their illness severity. And very high SOFA scores, we are not going to allocate a ventilator to them because we consider them to be unlikely to survive their critical illness. Their ICU mortality is very high. And most jurisdictions who have plans use some element of the SOFA score. This is what we use in North Carolina, where I practice. People show up, and you get a number for your SOFA score up top. And then you get another score depending on how many comorbidities you have. And that produces a number. And if it's red, you're the highest priority. And you would be likely to be allocated a vent. And if it's yellow, you're the lowest priority. And if you're orange, you're in the middle. And that is how we would decide what to do if we run out of ventilators in North Carolina. This is not an unreasonable plan on the surface. If we look at SOFA scores, we know that based on pre-pandemic data, very high SOFA scores are pretty predictive of ICU mortality. That if you have a SOFA of 17, 18, your ICU mortality is about 80%. That's a real number. And in a time of shortage, we may be compelled to make the difficult decision to not intubate that patient. That's logical. It's painful. But it's logical. Unfortunately, SOFA at presentation is not strongly predictive of COVID-19 mortality. It's not. The ROC is barely more than 0.5. It is slightly better than a coin flip. Your SOFA score, when you hit the door in the emergency department, it doesn't really tell me if you're going to live and die. In fact, just using your age to calculate mortality is more predictive than your SOFA score. That's not helpful. What about other scores? NEWS and MUSE, these early warning systems. So these are a couple of papers that we did. This one was done using the data from the adaptive COVID treatment trials. And you can see the negative predict value is pretty good. The positive predict value isn't great. The receiver operator curve is at best around 0.7. Very low news, that patient's likely to do well. But we already knew that. I think I can usually tell you that the person who's on room air and has a normal blood pressure is probably going to be OK. Very high news score. It's about maybe 50% predictive of your risk of intubation. Similar data set, same people. This is from the military's EPIC cohort, which is a large multi-center COVID prospective cohort that we've been running since 2020. And we found about the same thing. What about the ISARIC score? So the ISARIC score, this is actually one of the better options. It seems to have a fairly good correlation between mortality and death. This is based on about 80,000 patients in 308 hospitals in the UK. And this seems to correlate relatively well with COVID-19 outcomes. But this is the challenge, right? Our ability to prognosticate. If we're going to develop some sort of resource allocation thing, we have to have a way to be able to predict who's going to do well and who's going to do badly. And we don't really have very much. The scores we have, and there are a million other COVID-19 scores, the ones we have are either not that great or don't tell you anything you already knew. I figured out that the guy on three vasopressors who's on an FIO2 of 100% in 12 people, prone, I figured out that he's probably pretty sick. I know that. So why don't these models work well? If we are going to find a way to allocate resources, we need to have a good way to predict. And we don't. So I make a few hypotheses here. One is that acute bacterial sepsis and influenza are not the same as COVID-19. The diseases are different. And on some level, we would require disease-specific scores. And that's very difficult. How does a triage team who is separated from bedside care know which calculator to take out, right? There's a relative rarity of human dynamic instability in acute COVID-19 at the time of presentation. A lot of these scores use a basically binary mode of scoring for hypoxemia. It's like yes, no. And obviously, the prolonged nature of COVID-19 compared to other comparable causes of critical illness, with patients on the ventilator for weeks to months. That's different. What else? I would argue that we are trying to use these predictive models for the wrong reasons. Predictive models like this are useful for standardization of clinical acuity in clinical trials. It's a reasonable use of these scores, right? If I want to know how sick a population is in a septic shock trial, looking at their SILFA scores, looking at their Apache scores as a way to normalize it, that's useful. Using it as a screening test, sensitive but inspecific, to identify patients at risk for deterioration, that's a reasonable risk. Using these sorts of scores for crisis triage, I think, is premature, at best. And when I think about it, if anyone's ever read the Foundation books by Isaac Asimov, HBO made a miniseries about it that it was apparently made by people who never read the book, but it's pretty good. Lee Pace as the emperor is a good bit of acting. So there's this concept in this book called psychohistory. And psychohistory is this thing in the book Foundation where a science that combines history and sociology and mathematical statistics to make general predictions about the future behavior of very large groups of people. And the idea is it's using math to predict the future. But they make a point in the books of saying this only works if you're looking at thousands or millions of people. It doesn't work on the level of the individual. There's no mathematical way to predict what one person is going to do reliably. I think these scores are the same. I think that using a SOFA score, which is a population-based tool to give me a general risk of how a certain population group is going to survive or not, doesn't work that well on the level of the individual patient. And also, psychohistory is fictional. How about frailty? Frailty is a pretty good prognostic tool. There's standardized scores, clinical frailty scores, and the like. And when we look at this, we see that your likelihood of ICU survival in multiple critical care settings, surgical ICUs, medical ICUs, COVID, and the like, that high frailty scores correlate with your risk of mortality. However, even the frailest of patients, half of them do survive. Half of them do survive. So is that good enough to make decisions about risk-risk resource allocation? I don't think it is. This is another study showing the same thing. But basically, the lower dashed lines here are the elderly frail, meaning greater than 75, with a very high frailty score. And this is COVID ICU survival. It's about 50% for the most elderly and the most frail. 50% survival is not that bad. It's obviously 50% worse than we'd like it to be. But it's hard to make a decision based on that. It's hard for me to not care for that person in the best way I have. Now, certainly, there's room for goals of care discussions. And those should be mandatory. How about if we just eyeball people? What if we just look at them and decide how sick is this person and are they likely to benefit from critical care? It turns out that we are about as good as a scoring system, which makes sense. I hope it makes sense. There's a number of papers about this. This is one I particularly like. This was in 2018 in the Annals of Intensive Care, looking at intensivists' assessment of whether or not given patients are likely to benefit from critical care, and then comparing it to internists, comparing it to hospitalists. And both of them, it's not perfect. AROC is about 0.63 for intensivists. Internists, interestingly, are better at it than we are, it turns out, at least in Switzerland, that the hospitalists did a better job of figuring out who was likely to benefit from the unit. I suspect maybe some of our innate pessimism got in the way. But that's sort of interesting. Now, there are all of these ethical challenges in triage. Life stages versus years of life. What do you do with pediatric patients and pediatric hospitals in triage? How do you make sure you're treating people with disabilities fairly? A number of these studies, I'm sorry, a number of these SOFA scores and other scoring systems that incorporate things like the Charleston Comorbidity Index, if you're living with a disability, you're going to rank lower. Is that fair? Is that right? What about avoiding disproportionate impact on already disadvantaged communities, particularly communities of color, where if you come in and your SOFA score is higher than mine, but you had no insurance and you couldn't go to a doctor for the last 30 years to get your hypertension taken care of, and now you have CKD, are we just amplifying inequities by using things like SOFA scores and comorbidity indices in the allocation of resources? Dr. Doug White at Pitt has developed a tool called the Area Deprivation Index. We're basically looking at where you live and using that as a way to mitigate some of those impacts. Again, it's another bit of math into a triage system, but it's an interesting concept of a way to get around this. And what it comes down to is, can we even perform prognosis-based triage versus first come, first serve or versus just random allocation, just a lottery? There are three of you. There's two vents. We draw straws. Whatever we pick, it has to have some sort of a defined process. It has to have a way to address these issues of liability and ways to provide the best care with what you have. There has to be documentation. There has to be a way to appeal. There has to be some mechanism for post-hoc review. There has to be an indication for admitting triage. And honestly, what you need is liability protection. You need liability protection from a government entity to protect us, the people that we're And you can't institute a system like this without that level of protection. But I just said we've never done this. We've never actually successfully implemented it. And I am starting to wonder if we can't implement it. I'm starting to wonder if it's not possible, at least not in the way that we have written it right now. And so Dr. Truog recently put a thing out in the Hastings Journal. Dr. Truog recently put a thing out in the Hastings Center Report late last year. And he made this observation. He said that we had all these plans, none of them happened, that effectively, it was first come, first serve. Because how did our patients present? They didn't come in groups of 20 all at once. They came one at a time. And they were intubated one at a time. And if there was event, and they presented, and they needed to be intubated, they got intubated. There was no opportunity for a triage team to sit down and review you, then you, then you, then you. So how are we going to try to reconcile this? Because we are still going to be faced with situations of shortage. So I think some of these prioritization protocols that we developed may still have a use, but for something that's a little less hyperacute, like initiating dialysis. We have a little bit of time to think about that. That's not as much of a spur-of-the-moment decision. The use of scarce medications. I said we tried this with remdesivir before, and I think that was a reasonably successful strategy. But for something so fast as intubation, I think it's going to be very difficult to implement. One potential option, and this is something Dr. Truog described, is the idea of a time-limited trial, in lieu of formal triage at the time of intubation. Rather, time-limited trial is a mechanical ventilation for peoples with people who have potentially worse prognoses, reassessing frequently. Now, that would involve potentially the withdrawal of invasive life support, and that can be more challenging. But that may be the only workable solution. Then another thing is we've been talking about triage teams. Should these decisions be made by separate teams? I've always thought so. I've always thought that that role needed to be separated from us at the bedside. But I've started to wonder. There was an article that came out in Critical Care Medicine relatively recently, and it was a simulation exercise of ICU triage in a German university hospital. And the decisions were made by the attendings. They had multiple teams, and then they would go and round together. And the attendings involved in the care of those patients were the ones who made the triage decisions at the bedside. And I actually had the opportunity to communicate with the authors about this, and I said, that's different. That's not what a lot of organizations and jurisdictions in North America do. Why did you choose that? They said, well, partly it was just practical. They said, and partly it's because we didn't feel it was right to deflect our moral burden onto someone else. Because it is interesting. There are some small studies looking at the moral burden on triage teams, and it may not be any better than the moral burden that we would experience for managing these people at the bedside. Are we just shifting that moral injury from one group to another? They also make an interesting point that, as we talk about how limited these scoring systems are, they make an interesting point that there are subtle prognostic features that you see at the bedside that cannot be rendered into a number. And that's an interesting thought. This does, again, raise concerns of issues of equity. I would hypothesize that the specifics of health inequity in Germany and in the United States are not the same. But nonetheless, it's a provocative thought. And I think something, as we start to plan for future emergencies, something we should at least consider. Lastly, health care worker support. I am certainly not the first and probably about the 900th person at Congress this year to talk about moral distress and injury among health care workers among us. I think we all know the symptoms. We're all familiar with the strain at this point. Major issues have always been insufficient training, high patient burdens, later in the pandemic, difficulties with families, and shortages of PPE. What are things that we do know help? Debriefs help. Having input into working conditions help. Open communications helps. This is a manuscript that we've submitted and are awaiting acceptance for. I know this isn't legible, but just showing broad categories of things that one can help with staff retention. This is a result of a modified Delphi process we did over the course of several months. Workload, mental health support, environment, institution-specific gaps, communication, and finances. It is interesting that we, I'll skip ahead a little bit. It is interesting that in the recent New York nursing strike that money was not the issue, that the hospitals offered the nurses more money, said that's not what we want. We want better working conditions. We want better staffing ratios, that we want to be able to provide safe care. I've often sort of joked at my own place of business that when one hears that hiring more staff is challenging and harder, that often paying people more money is a good way to get people to do things. But that's not the whole story. So what makes a response successful? Communications, right? Not in the abstract sense, but in a very tangible sense. Structured, formal mechanisms for communication within health systems to make sure staff have the maximum amount of information possible in a manner that is actionable. And then communication between health systems at the state, the region, at the national level. Again, there are some success stories out there, Washington, Minnesota, San Diego County. And there are some, we'll call them anti-success stories as well. Coordination, fostering this sort of teamwork in advance so that you're not building the emergency response as the emergency is unfolding. We have now a generation of critical care people who have lived through the greatest public health emergency of our lifetimes, right? We cannot forget. We cannot forget. We have to figure out a way to be better. We need to have an accurate inventory of our resources. We need to know how many beds there are. We need to know how many events there are. We need to know if those events work. We need to know how many staff we have. We need to know who we can call on if we get overwhelmed. Unfortunately, this is my current home state right now. There was a great deal of support provided to health systems from the government, from us, from taxpayers. And I will say that not every institution has used that money wisely or for the intended purpose. That we still have staffing shortages, but we are building a beautiful new tower, right, with no people to work in it. And if anyone from my employers is here, I'm sorry, but not sorry. One of the systems in our state was referred to in the news as a hedge fund with a hospital attached. I alluded to the New York nursing strike earlier, but just to say, it's not all about just paying people more. It's paying people better. So what is the bottom line then? Things are not critical. Capabilities are critical. It doesn't matter if you run out of sex signal choline. It matters if you run out of the ability to provide neuromuscular blockade. It doesn't matter if you run out of fentanyl. It matters if you run out of the ability to provide analgesia. And it doesn't matter if you have a vent if the vent can't take care of the right patient, or if you don't have an expert who knows how to run that vent. Capabilities matter. I think we need an honest reappraisal of crisis standards of care. I think we need to look hard at triage plans that we have learned defy practical implementation. I think we need to acknowledge limits in our ability to prognosticate. I think we need to take a look, and these are happening. Acknowledge the impact of delayed elective care. Some care is truly elective and can be delayed. Some care is not. Cancer care, cardiac care, and the like. And whatever we decide, we have to make sure that whatever we do does not exacerbate, and ideally reduces, existing inequities, right? We can't use an emergency excuse just to make things worse. And lastly, we cannot be caught with our pants down again. I've kind of joked on occasion when we talk about biodefense and bioterror, like the first case of plague that shows up in your hospital unless you practice in the Four Corners near New Mexico. The first case of plague, you'll always get caught with your pants down, right? But by the second or the third case, I hope you've caught on by then. This is the same story. We have to support our people. We have to rebuild the clinical workforce. We have to establish systems for cooperation and low-living between competing health systems and emergencies. And if this requires a force of government intervention, then so be it. And we have to reduce our dependence on just-in-time single-use goods, not just for environmental reasons, not just for climate reasons, but just because I don't wanna have to have six rotating paper bags with N95s in my office anymore. So, just to wrap up. A lot of this was new to us, but nothing is really new, right? This is outside of the old Walter Reed Army Medical Center in Washington, D.C., during the Spanish Influenza pandemic. We see masks, we see ventilation, we see patients separated in space. This was a protest during the Spanish Influenza. This was actually a counter-protest. These are people wearing a sign that says, wear a mask or go to jail. This could have been taken, you know, take away the hats, and this could have been taken two years ago, right? I mean, this lady's nose is even sticking out. She's even doing it wrong in the same way. And of course, there are tales of heroism, too. Javits Center in New York at the beginning of the pandemic. The volunteerism, people from all walks of life, all parts of the country, all parts of the world, in some cases, coming together to care for people. Taking care of people wherever they were, building new places to care, tents, healthcare workers, right at the beginning of the pandemic, so everyone wasn't wearing masks yet, flying to New York as volunteers. I see a few friends in the audience who did those very same things themselves. This is my friend, Sean McKay. Sean is a Navy pulmonologist. Sean was in command of what's called a rural response team, sent out to a badly stricken hospital in the Rio Grande Valley of Texas, and was there providing care and leading a team as a military officer and as the only intensivist there, caring for all of these patients. And I had to send this to him because Sean is not a surgeon, and he was very offended, and that's not also how you spell his name, but still, we appreciate the news doing their best. And this continues today. Rahm Stevens, member of a program committee of SCCM, one of the planners of this meeting, anesthesiologist and intensivist, continuing to volunteer in Ukraine now to teach critical care to Ukrainian physicians and nurses to provide for care for the victims in that war there. This continues. It never stops. So it's easy to think about the million people we've lost. It's good to remember that there are millions of people still alive because of you, because of us. That ain't bad. Thank you very much. We got some time for questions. Thank you.

COVID-19 pandemic

challenges

crisis standards of care

communication

healthcare worker support

inequalities

healthcare workers

load leveling