All right, folks, good morning, and welcome to this exciting session on resuscitation and sepsis, the holy grail. We have a wonderful speaker lineup for you here today, and one of the things I would – by the way, I'm Rahul Manchal, I'm the moderator for this session. I am a practicing intensivist at the Medical College of Wisconsin. And if you could please hold your questions till everyone is done with their talks, then we will have a panel discussion at the end where you guys can get up and ask questions of all the experts. All right, so without ado, we are going – without much delay, we are going to get started. Our first speaker is Tiffany Osborne. Tiffany is professor of emergency medicine and surgery at Washington University in St. Louis. She has been with the Surviving Sepsis Campaign for over 15 years. She is – was the clinical trialist on the PROMIS trial, and she is going to be talking to us about lactate and lactate reduction. Welcome, Tiffany. All right, good morning. Gosh, y'all are here, and it's so fantastic in Orlando right now. I mean, what is it, 70 degrees outside, and very different from St. Louis right now. I can tell you that for sure. So my discussion today is going to be on lactate. And as we know, there's – are you guys hearing an echo, or is it just up here with me? Are you hearing an echo? Yes, sir. You are hearing an echo? No. Okay. All right, good. So there's a lot of confusion around lactate, right? Lactate, yes. Lactate, no. What are we really supposed to do? How do we put this at the bedside? And I'll be the first to say that lactate is imperfect. And my son, who is 14, also tells me that I fit into that category quite frequently, being imperfect. But lactate is imperfect. One of the issues that I hear frequently is, look, it doesn't tell us that someone is infected. It can be elevated in a non-infectious state. So is it still helpful? Well, let's think about some other areas where it's helpful. So lactate is used ubiquitously in trauma. But it doesn't identify trauma. Because someone has an elevated lactate, it doesn't mean that they have incurred trauma. Additionally, in trauma, lactate doesn't tell you the problem. It tells you there's a problem. We don't know if the reason for hypoperfusion is because of perforated bowel, or if it's because of bleeding, or if it's because of solid organ injury. We just know that there's a problem that we need to identify. So it doesn't give you the answer. It gives you a clue that you need to look for one. Think about troponin. I mean, troponin is elevated in numerous things that have nothing to do with myocardial infarction. Does that mean that if you thought someone might have a myocardial infarction, you're not going to use it? Absolutely not. The bottom line for that is that sepsis biomarkers, and to be quite frank, no biomarkers, all biomarkers, they augment clinical acumen. They don't replace it. What you do as a clinician is important, and you can't abdicate that to a biomarker. The biomarkers, they augment our clinical acumen. They don't replace it. So again, still, lactate, you know, all right, so it may be helpful, but look, does it really matter when it comes down to what we do at the bedside? Should we use it? And I'm going to say that the question isn't, should we use it? The question is, how and when do we use it? So there are pathological states of hyperlactemia associated with hypoperfusion, and then there's non-pathological states associated with hyperlactemia. It's partly up to us to get that sorted out. So is the non-pathological problem, is it associated with decreased metabolism because of shock liver? Is it associated with decreased removal from acute kidney injury? Is it the drugs that you have the patient on? There are plenty of non-pathological reasons for hyperlactemia. When you look at how lactate is reduced over time, during that initial resuscitation, those first few hours where you are resuscitating that acutely presenting patient, you see a sharp decline in lactate. And then after that 6, 8, 10-hour mark, you start to see a continual decrease, but it's a much slower decrease over time. So that initial decrease is when the patient is acutely presenting. That may be hyperlactemia associated with pathology, oxygen delivery not meeting demand, decreased oxygen uptake, things that you can address and modify by your treatment. Whereas over time, once all of your boxes are checked and you're seeing that the lactate is decreasing but not continuing to go down, all of your resuscitation boxes are checked, as our colleague is going to tell us about in a few minutes with ultrasound, then maybe it's not something you need to continue to resuscitate for. Maybe we need to look at other reasons for the fact that it hasn't normalized. Trying to try to use techniques to reduce lactate when the etiology is one thing, when the etiology becomes a second thing, is not helpful. That's where we get into trouble with over-resuscitation. It's not the fact that lactate is not helpful, it's the fact that we continue to use it for an etiology that we may have already addressed. So is lactate associated with mortality? I mean the answer there is absolutely yes. There is no question that there's more studies about lactate associated with mortality than I have slides I wanted to present. But lactate is not just associated with mortality and hypotension. It's also associated with mortality in people with abnormal blood pressure. So if we look at the sepsis 3 shock discussion, that shock paper, we look at this table. If we break that down into normal tension with an elevated lactate or hypotension with a normal lactate, what's the difference? Well, if your blood pressure is normal and your lactate is above 4, your mortality was around 30%. If you are hypotensive and your lactate is normal, your mortality was around 30%. This was again demonstrated in other areas too. You can see when we were getting the background for PROMIS, we looked at this and it was a similar mortality at 26% normal tension and elevated lactate. This isn't new. Dr. Rivers, esteemed panel member, when he came out with his 2001 study, this is one of the things he was talking about, is that shock doesn't just start when you see it. We define shock as when the blood pressure decreases because that's when it's easy for us to find. What he was demonstrating is that there's a shock, a physiologic shock that's building and if you can arrest it, then you may be able to make a change. Lactate is also helpful with early recognition. How does that help us with shock? How does that help us with septic shock? Because the best septic shock outcome is not getting it. The best septic shock outcome is not getting it. If we can identify it early, if we can treat it early, maybe we can avert the problem. Let's look at some other studies. When you look at PROCESS, PROMIS, and RISE, over 90% of these patients, once we thought they were infected, the majority of them got point-of-care lactate. That lactate was the first indicator that the patient was at risk and qualified for enrollment into the studies. 30-45% of patients, lactate was the first indicator that the patient was in trouble. How does that translate into actually impacting mortality? We can look at the New York State data of 40,000 patients and we saw that when the three-hour bundle was completed, for every hour delay that the three-hour bundle was not completed, there was an increase in mortality. What was that? That bundle was lactate measurement, blood cultures, antibiotics, within six hours of initial presentation, being initiated, and completed within 12. You may think, measuring lactate, come on. How does measuring something really impact mortality? When we look at 50,000 patients that had this elevated lactate within that data set, what we saw here is that there was an increase in mortality for every hour delay of checking lactate. Now, look at this. It was most significant in the normotensive patients. Two-thirds of these patients were normotensive, and time to lactate measurement had a significant impact on survival in an hourly fashion in these normotensive patients. That gives us a different thought when we're thinking about CUSOFA, where we were talking about borderline hypotension, altered mental status, respiratory rate, because we're already close to being hypotensive. Maybe in that situation, lactate wasn't adding anything to those patients, right? And that's what they saw. Maybe the reason why there was an increased mortality with a delay in lactate measurement was because when you see it, you do something about it. So Andromeda, this also makes us think the same thing. These are not inconsistent. These are all consistent data. So pre-ICU, the patient comes in, they're infected, they develop shock. Then they get admitted to the ICU. Upon ICU admission, that's when they're randomized. So if you were in shock, you were on vasopressors, and you had a low capillary refill, and then you as the clinician check that capillary refill and use that to determine your resuscitation every 30 minutes. That had the same outcomes as if you were measuring lactate every two hours. So there was no difference in 28-day mortality, no difference in 90-day mortality, no difference in hospital mortality. And SOFA score was a little higher with lactate, but it wasn't associated with increased mechanical ventilation days. It wasn't increased renal replacement days. There were no increased vasopressor days. So the idea here is this is an excellent study to say, if you already have a patient who's hypotensive, on vasopressors, and already has decreased capillary refill, you can use capillary refill as an indicator. It doesn't say you have to use it as your only indicator if you're in that immediate phase of resuscitation. So what about when you're dealing with this pathologic hyperlactemia? Now does serum reduction of lactate, is that actually associated with survival, or is it just seeing that it's there and treating? I mean, is there actual reduction? Does that change anything? Yes, in context. Remember again, as long as we're in that acute resuscitative phase where there is a pathologic system ongoing. So when we look here, you can see these were all early resuscitation patients. This was done in the emergency department. The odds of survival were five times higher once lactate was normalized, four times higher if it was reduced by 50%, and still of benefit when it was reduced by 10%. Again, here you can see there is a survival benefit to lactate clearance versus non-clearance, whatever that etiology is. And Jean-Louis Vincent does his systematic review, and he says at the end that there is an improved prognosis with decreasing serum lactate. On top of that, it does not matter if it's a septic patient or a general ICU patient, any critically ill patient in extremis, there was an improvement with reducing serum lactate. His recommendation was to check lactate every one to two hours, so checking it early and checking it often. That was his recommendation in his paper. So in summary, biomarkers augment your clinical acumen. They don't replace it. Lactate, not all lactate is treated the same. You have to think through where you are in the resuscitation phase. The best shock, the best treatment for shock is to never getting it. So if we can identify them early and try to arrest the process, our patients are better off. Point of care lactate can help with identification, and early serum lactate reduction improves survival. So thank you very much for your time, and I hope you're enjoying Orlando. Thank you for the great talk, Tiffany. And just to remind you folks that we will take questions at the end of everyone's talks as a panel. So I am going to be giving the next talk about physical examination as an endpoint of resuscitation. And again, I'm Rahul Manchal. I am a practicing intensivist at the Medical College of Wisconsin. All right, so I am going to, in the next few minutes, try and convince you that physical examination, at the very least, should be used as an integral part of our decision making in determining what the endpoints of resuscitation are. I have no relevant disclosures for this talk except these biases. I am primarily a clinician, and I have to confess that some of my ongoing studies are looking at the missed physical exam opportunities and their relationship to diagnostic error and outcomes in critically ill patients. So I am highly likely to be affected by all of these biases. All right, so the objectives of my talk are to talk about some of the history behind the use of clinical exam and critical illness, to discuss how we can use it to guide resuscitation and to debate whether it is better than other endpoints or not. So the use of physical exam and critical illness has mostly been used for prognostication. And the parameters mostly used have been vital signs, motling, and capillary refill time. Now, there are a variety of other markers that you guys intuitively use every day and incorporated into your clinical decision making, such as patterns of breathing, some components of the neurological exam, feeling the pulse volume, palpating the pulse volume, and so on and so forth. Now, these haven't been very well studied, but I'm sure all of you, every day when you are rounding in your ICUs or seeing patients, incorporate some of these into clinical decision making and using them as endpoints for something or the other. So let's talk about prognostication first. So this was a wonderful study now published in Nature in 2018, where Sweeney and his colleagues demonstrated that doing fancy gene expression analysis and adding it to severity of illness scores markedly improved the discriminatory ability of prognostication. But the real question is, can we use just a physical exam to reliably prognosticate or guide? I call this the Sherlock Holmes approach to septic shock. This is a quote from The Hound of Baskervilles. The world is full of obvious things, which no one by chance ever observes. And the history of resuscitation, or the history of using physical exam as an endpoint of resuscitation, is as old as the history of resuscitation itself. So this is the first ever published account of resuscitation during the epidemic of cholera, letter written to the Lancet by Thomas Latta and sent into the Lancet by his colleague Roger Lewins. And I want to read you an excerpt from the letter. Please, this is in 1832, so don't expect p-values. Don't think this is not a randomized control trial. And it's a little wordy, so please bear with me. But it's illustrative. Having inserted a tube into the basilic vein, cautiously, anxiously, I watched the effects. Ounce after ounce was injected, but no visible change was produced. Still persevering, I thought she began to breathe less laboriously. Soon the sharpened features and sunken eyes and fallen jaw, pale and cold, bearing the manifest impressive death signal, began to glow with returning animation. The pulse, which had long ceased, returned to the wrist, at first small and quick. By degrees it became more and more distinct, fuller, slower, and firmer, and so on and so forth. So this is using targeted serial physical exams as a resuscitation endpoint in the paper that was published in 1832. All right. So as opposed to gene expression, this is one of my favorite studies about patients admitted with severe sepsis to a medical ward in Uganda. And the frequency of monitoring vital signs, on average, was once a day for blood pressure, once a day for pulse, once a day for temperature, and once in two days for respiratory rate. Now, we need to prognosticate and have endpoints of resuscitation in resource-poor environments as well. So I have a sneaking suspicion that I will not be able to sell gene profile expression or serial labs in Uganda. But perhaps we could spend some more time looking at the pattern of breathing and looking at the respiratory rate. In the same study, a respiratory rate over 20, a temperature less than 36, and a CRB score, which is just confusion respiratory rate, age over 65, and blood pressure were predictive of mortality. Now, I don't know what happens in your institutions. But at mine, if I were to be blinded to the charted respiratory rate and someone asked me what is the respiratory rate of the patient, and if I say it is charted at 20, I will achieve statistical significance. OK. So some more on temperature. This was in 1969, a paper by Max Harriveen, the same person who gave us the definitions of shock, who clearly demonstrated that measuring the great toe temperature was highly correlated with cardiac index. A temperature less than 27 degrees centigrade would reliably predict a cardiac index of less than 2. And not only that, if he measured the temperature difference between the toe and ambient temperature, it would determine outcome reliably, 67% of the time, it would correctly determine outcome when it was measured within three hours of admission. If they measured it serially, which is like doing serial physical exams, he could reliably predict who died and who did not. In survivors, the temperature difference increased. And in non-survivors, the temperature between the great toe and the ambient temperature decreased. Now, several studies have now been published looking at skin temperature gradients as prognosis in critical illness. And all of them basically have shown the same thing. This is my favorite quote from a scandal in Bohemia. You see, but you do not observe. The distinction is clear. So this group of investigators decided to see if the motling score predicted survival in septic shock. All of you have seen motling, and everyone sees motling in a variety of shock states. And their method of determining motling was really simple. So zero for no motling, you've got a score of one if there was just a small area of motling in the center of the knee, an area of two if the motling area extended but did not cross the superior aspect of the knee, a motling score of three if it extended into the thigh but not up to the mid-thigh, four if it extended to the mid-thigh but not to the fold of the groin, and five if it extended to the fold of the groin and above. And this simple bedside measure had wonderful inter-rater reliability. And look at this. It actually predicted mortality very well. Look at the odds ratios of someone who had a motling score of four and five, 74. I have never seen such odds ratios. I mean, in most of the studies I've seen, I've never seen these odd ratios. And if you did this serially, okay, and by serially I just mean two times, at hour zero and hour six, they were able to predict who would die and who would survive at day 14. The people who survived had improvement of motling scores, and the people who died did not have improvement of motling scores. This is motling in general critical illness. Now, these investigators in their ICUs, their nurses are trained to look for motling. Now, they don't use a motling score, either motling is present or absent, and their nurses chart it every four hours. And even one episode of motling was a predictor of mortality. If there was long-lasting motling, over six hours, it was another predictor of mortality, and this is amongst the people who had motling. This is from the Bascom Valley Mystery. You know my methods. It's based on the observation of trifles. This is the capillary refill time that you will hear in a variety of lectures throughout Congress and even in this panel. You've already heard about the capillary refill time from Stephanie, but these are the same group of investigators who were looking at motling and decided to use capillary refill time as an exploration during septic shock. They measured capillary refill time at two areas, the index finger and the knee area. Again, wonderful inter-rater reliability. And at both the knee area and the index finger, capillary refill time was highly predictive of 14-day mortality. The area under the curve was 84% for the index finger and 90% for the knee capillary refill time. Now, I'm sure I have not seen better area under the curves for either apache or so for whatever things you may, or whatever severity of illness scores are available. All right, so now back to the study from Uganda. This was the most interesting part of the study. So whoever was taking care of the patients actually really knew intuitively who was sick and who was not. Remember, these were all patients with severe sepsis. The people who were sick had more frequency of vital sign monitoring. And this frequency of vital sign monitoring actually turned out that it predicted mortality. Not only did they monitor the sick people more frequently, but also they were able to detect respiratory rates accurately. Because each unit increase in respiratory rate was actually associated with the frequency of vital sign monitoring. People were breathing fast. The nurses or whoever were there were going and taking more vital signs. All right, but that is Uganda. What about the rest of the world? And this is another wonderful investigation, the simple studies in intensive care. And I have summarized the significant associations with mortality. And this was all about doing simple clinical exams in the study. And the respiratory rate, systolic blood pressure, consciousness, central temperature, urine output, again, all things that are part of our physical exams you can easily acquire at the bedside, all significantly associated with mortality. So what are the endpoints of resuscitation in septic shock? To name a few, we already heard about lactate and lactic clearance. There's base deficit. There's venous oxygenation. There is veno-arterial carbon dioxide difference. And I find, so I'm a chart stalker, and I like to sort of go through charts and read notes and things of that nature. I find that, and even during sign out and hand off, residents are telling each other, you know, please monitor lactate X times a day or every four hours. And in the chart, there are serial lactates and venous blood gases and base deficits. I have rarely, if ever, found someone handing off, please go and examine your patient every four hours, or we need to do serial physical exams on our patients. All right, so let's talk about lactate and septic shock. This was data now published by Luciano Gattonini in the Blue Journal. So wonderfully done study, a reanalysis, sort of a re-look at the data from the ALBIOS trial. And what they were trying to determine was, what is the origin of lactate in sepsis? And how does lactate and central venous oxygen compare? What is the mechanism of acidemia from lactate? And why do some people with lactate have acidosis, and why do some others don't have acidosis? We are not going to talk about the acidosis part. But if you looked at lactate by sex styles, yes, lactate was associated with mortality as lactate increased. Lactate was associated with increasing sofascores, you know, across increasing lactate sex styles. But if you look at SVO2, or central venous oxygen, the central venous oxygen across lactate sex styles was absolutely constant, except the sixth sex style, where it fell a little bit, significantly, but just a little bit. The average SCVO2, or central venous oxygen, in the sixth sex style was 67%. Similarly, the arterial pH was constant throughout lactate sex styles, except in the sixth sex style. However, if you looked at variables by central venous oxygen sex styles, most of the variables actually followed a U-shaped distribution, the lactate, the sofascores, and the mortality. So in terms of lactate, there was high lactate at really low SCVO2s, and there was high lactate at the really high SCVO2s. Now, the percentage of people who had low SCVO2s, less than 65%, was only one-fourth of the people in the study. So the conclusion of the study was that it is not oxygen transport, but actually oxygen utilization that is frequently affected in sepsis. And it called into account the one-size-treats-all strategy of giving X amount of fluid in septic shock. All right, so is the physical exam better? Now, I'm just going to go over this trial really quickly. You already heard about the Andromeda shock trial. And people sort of randomized to a lactate strategy versus a capillary refill time strategy. And they had parameters of, OK, step one was whether there was fluid responsiveness. Step two was in chronically hypertensive patients. They raised the mean arterial pressure. And step three was an inodilator test with dubutamine or milrinone if the endpoints had not been met. Capillary refill time was done every 30 minutes. Lactate was checked every two hours. Endpoints were a capillary refill time less than three seconds, or a lactate clearance of more than 20% on normalization of lactate. And yes, the mortality at 28 days was not very different. However, there was a strong signal towards the peripheral perfusion or the capillary refill time-guided strategy. The hazard ratio barely crossed one. Not only that, the SOFA was different at 72 hours. And the resuscitation fluids that were given within eight hours were way less, about 500 CCs, less in the peripheral perfusion-guided group. And in pre-specified subgroup analysis, the people who had a SOFA score of less than 10 actually demonstrated benefit. Okay, so this is a one-slide primer on Bayesian analysis. So show of hands, so you're waiting on a subway train or you're waiting on a subway platform for a train that is known to run on a regular schedule, only you don't know how much time is scheduled to pass between the train arrival or nor how long it has been since the last train arrived. As more time passes, would you choose A or B? A is you grow more confident that the train will arrive soon because its eventual arrival can only be getting closer, or B, will you grow less confident because you feel like that either the scheduled waiting times are far apart or you just happen to arrive at the last train. How many picked A? Okay, all right, lots of you. So if you picked A, you think like a frequentist. If you picked B, you think like a Bayesian thinker. All right, so this is, and that's what Bayesian analysis is all about. You are taking into account sort of prior beliefs and then determining your, the likelihood of something happening based on your prior beliefs. And so in a Bayesian reanalysis of the Andromeda trial, of the Andromeda shock trial, where the likelihood of something working was either taken as optimistic, neutral, pessimistic or null, optimistic being in favor of the peripheral perfusion guided strategy, neutral being both strategies are equal, pessimistic being the lactate-guided strategy is better, better and null being anything can happen, you know, whether it is better equal or whether one is better than the other or they're equal is exactly the same. The probability of getting a odds ratio of less than one, the probability of getting an odds ratio of less than one, which means in favor of the perfusion guided strategy was over 90% for all prior probabilities. Okay, so the physical, so we probably intuitively use physical exam markers to drive decision making all the time. You know, for example, I perform serial exams if I find abdominal distension, increased work of breathing, forced expiration, cool toes and so on and so forth. And we now have evidence to say that at least it should be incorporated as an integral part of determining the endpoints of resuscitation. And I end by asking you whether believing is seeing or seeing is believing. There is nothing more obvious than a deceptive, there is nothing more deceptive than an obvious fact and please open your eyes to the multitude of physical exam findings that can be followed serially, some of which are really apparent early during critical illness. Thank you. All right. All right. So and to our next speaker is Emmanuel Rivers. So Dr. Rivers is professor of emergency medicine at Henry Ford and, you know, obviously he is the person who, why I think we have this session today, is the person who changed the way that we resuscitate people for severe sepsis. I think everyone in the room knows him, Dr. Rivers, please, and he's going to be talking to you about central venous oxygen saturations. Good. Thank you so much. Thank you for that warm introduction. Now if you gave me five more minutes, I would appreciate it. But what I'm going to do today is talk about the holy grail of resuscitation and to convince you that measurement of oxygen utilization variables are still important. I'll start off with describing the concept of SpO2 as an endpoint and then talk about the physiology, the history, and the outcome implications as well as the follow-up limitations. So why is measuring oxygenation so important? Well, if you look at who my staff was when I was a fellow at University of Pittsburgh, it was Dr. Jim Snyder. And his specialty was oxygen transport and utilization. So if you didn't end your morning presentation without some mention of these variables, you were not a good clinician. And to be more forceful is that some of my admirers also received a Nobel Prize for the same. So if you look at last year, Nobel Prize winners, it was on how cells use oxygen. So a little about the physiology is that this gentleman did something centuries ago that we rely upon on a daily basis to diagnose shock, and that's blood pressure. He basically took horses and put a glass pipe in their necks and measured their blood pressure up and down, up and down until they died, and that's the concept of blood pressure. And if you look at today and you look at how we measure or differentiate the differences in shock, we basically take a patient who's severe sepsis, use blood pressure as a means of detecting whether or not they're in shock, and that mortality goes up 46%. But we all know the limitations of blood pressure, because if you decrease your cardiac output, increase your vascular resistance, your blood pressure actually can remain stable until you get really sick. And so that underlines the equation of why blood pressure perhaps is not the end all. And if you look at the whole idea of oxygen transport, which you'll be an expert in 30 seconds if you're not, you take oxygen, put it on hemoglobin, multiply it times cardiac output, and that gives you a tissue oxygen delivery of about a liter per minute. We consume about 25% of that, so the net oxygen returning to the heart is the venous oxygen saturation. And so when you look at the simplistic concept, it all makes sense that if you're measuring oxygen saturation, you're measuring the most important part of resuscitation. The pulmonary artery is a normal source of your venous oxygen saturation, but Dr. Conrad Reinhart years ago did studies comparing the two sites, and what he showed are a value of approximately .9, which indicates the correlation is pretty good. If you look at the whole idea of what we're trying to accomplish, when we get a number called a venous saturation, immediate things like improved delivery, oxygenation, hemoglobin, increasing that cardiac output, or decreasing those consumptive demands by treating stress, pain, hyperthermia, shivering, or taking away that work of breathing. Now if the number is good, then that means everything may work out okay, which most of the time it does. But if it's not, if it's high, and you have evidence of increased lactate, that can indicate what we call cytopathic tissue hypoxia, and that's perhaps not a good sign. So when you look at a resuscitation endpoint, it must be important to look at it in context of other variables, and so gives that you must do that with oxygen transport variables. But the whole idea is what are we trying to accomplish? Dr. Max Harry Wild almost 50 years ago told us that lactate was a good predictor or prognosticator of mortality, and he related the fact that tissue hypoxia, which is what we're trying to eradicate, if made severe, will cause a lactate of four, or more, as a prediction of outcome. So the whole acronym we use is a lactate of four should never go to the floor, and he showed that in 1954. I thought you'd get a kick out of that one. So what are we trying to say? Well, if you look at oxygen transport variables, they tell you just another dimension of what we're trying to accomplish, and the earlier you can basically correct it, the better you correct the patient's underlying physiology. So we define shock not just as a blood pressure phenomenon, but as evidence of tissue hypoxia that is really the definition of shock. Many people don't really know that prior to the co-directed study in 2001, we did almost 10 years of research in various shock states, looking at this phenomenon of what we call cryptic shock, global tissue hypoxia, which leads to oxygen debt. We took patients at triage with abnormal vital signs, and we resuscitated those patients. We actually used oxygen transport variables like SCV02 and lactate, and what we found is that 31 out of 36 patients who were resuscitated to normal vital signs continued to have evidence of tissue hypoxia. So this gives rise to that early phase of correcting a more perhaps physiological cause of your shock state. So what about the resuscitation itself? Where did that come from? Well, Dr. Bill Shoemaker, a former president of the Society of Critical Care Medicine, had an idea. He said that while you're generating lactate and you're fermenting wherever you are, you perhaps develop oxygen debt. So if you're not meeting your consumptive demands, you're accumulating debt, and the greater that debt's repaid for recovery is possible, delayed debt repayment means you have renal failure, perhaps liver failure, and if that deficit is not repaid, there's irreversible organ dysfunction, which really makes sense. And this slide is probably over 60 years old. And so when you look at the whole idea of what we're trying to accomplish is to rid that debt as early as possible in the shock phase. So Bill Shoemaker said that one of your expertises should be resuscitation. So why don't every person who sees emergency patients or perhaps critically ill patients have an algorithm to treat shock states? And this algorithm was based on some of the premises of tissue oxygenation. Dr. Max Harry Weil, who said the VIP approach to shock, if you can note the date, 1969. And what he had is a mnemonic called VIP, adequate ventilation infused and restored a pump or cardiac function. In addition, he described pharmacologic and surgical specific indications if your patient needed source control or perhaps eradication. So he had a concept of using CVP, using mean arterial pressure, cardiac index, and also looking at reversal of endocyanin green. And lo and behold, he looked at hematocrit and red cell blood mass. And what he described is a case where he took an elderly lady and he gave her over four to five liters of fluid. He saw her CVP go down, her hemoglobin remained stable, her lactate clear, and lo and behold, a concept of early resuscitation came about. And then Dr. Wilson, who said it's rational to measure your venous oxygen saturation and this is in 1965. And he had the ideas that you could do this both intermittent and continuous. And what he said is the final statement is that let's hope this time is not too far off. And the theory and practice meet halfway for the sake of our poor patients in clinical shock. This was in 1965. So here we are moved forward after we examine our patients and they actually have been in the ER or perhaps somewhere else in shock for hours and we say well maybe we ought to get them to the unit. Well this concept of Dr. Frank who said patients should perhaps be treated just like our animals who are in shock for two hours and perhaps we should reverse that as short as possible with the concept of a shock team. So this gave rise to an idea because as we go home and we see our patients are in our EDs and perhaps other areas of the hospital where ICU beds are shortage. This was told to us almost 30 years ago and here even in 2018 this continues to be an issue. So when you look back at the 90s and you say well what was the landscape of the treatment of early sepsis? This was the landscape. It was a nice beautiful evening with nothing. And then if we look at how we treated our septic patients, this is how we got notified back in 1997. Probably many of you in this room don't know what that is. And then it progressed to that which I had one of those as well. And now we have more modern means. Remember this is within the last two decades. So here gives rise to why we needed a perhaps resuscitation protocol is because we had no means to describe sepsis, no SERS, no risk stratification. Lactate was not used uniformly in my hospital until the year 2000. In addition, antibiotic recommendations, resuscitation standards, no fluid recommendations, no mean pressure, no myocardial dysfunction assessment and not even SCV02. So when you look at the landscape, we now have a lot, two decades of research but we didn't have anything. So what do we do? Well move forward and say that this is where we want all of our patients to be in this beautiful ICU. However that wasn't the situation two decades ago so we took what the American College of Critical Care Medicine said is give your patients fluids, vasopressors, give inotropes, perhaps check that lactate, get that imidacrit to 30, not 7 or 8, 30 and check that SCV02. Remember this was for ICU patients. So we decided to look at it urgently in our modest ER setting, provide the same concepts and that led to a protocol that I just showed you that was described by Dr. Shoemaker, Dr. Max Harry Weil, etc. and that was a concept of early goal-directed therapy. So not rocket science but listening to your peers, elders and learning the basis of what we call tissue oxygenation. And so now we're here two decades later and where is the evidence? Well this is a study from Dr. Pope, 619 patients she examined the predictability as well as the prognostication of SCV02. And in patients who are normooxic they had a mortality, again SCV02 is in the 70 range. The patients who were hyperoxic elevated SCV02s over 85% and then a hypoxic group, those patients who had low SCV02s averaged 55% and what she showed was a correlation with mortality. So those patients who SCV02s were low, mortality over 40% but it's important that we understand that hyperoxic group, that dysoxic group that we explained, combination of SCV02s and increased lactates high mortality and then a normoxic group, a mortality of 21%. And if you move forward and say well what did she say back then? Well the findings even in spite of patients who are hyperoxic is that SCV02 highlighted the potential and significance of impaired tissue oxygenation but remained a good measuring tool for those patients who are resuscitated. And if you move forward and say what about the ICU phase? This was a multi-center study out of France that measured SCV02 upon ICU arrival not ED arrival and what they showed is that the incidence was about 25% and that those patients who were less than 70 had a 10% higher mortality. And so when you look at the result of the scientific community to SCV02 or to go direct to therapy, one of the issues was is SCV02 helpful? But when you look at these clinical trials and you look at how are they conducted compared to the original study, they were conducted in an expeditious fashion but in a different setting. These patients could be taken to an ICU and were taken to an ICU on an average of two and a half to three hours. That is not a replication of Henry Ford, perhaps not a lot of hospitals. If you take animals and you perforate their cecum and you measure their penis oxygen saturation and hemodynamics, what you'll see early is a drop in blood pressure, a drop in central venous pressure. You'll see a drop in cardiac index and also SCV02 is in the 40% range. However, after fluid administration, you may see a reversal and perhaps an improvement and this is exactly what we saw in the process of rising PROMIS trials. And so when you go back and look at what we call the kind of patients or the stage of resuscitation, the patients had higher lactates than the go-directed, higher, lower SCV02s and as you can see, in a different phase of resuscitation. There was basically something to be said that resuscitation works, it's just a phase that you're in in terms of where you are in your whole continuum of care. And this is a recent study published by the ALVEAS trial and this is over 1,700 patients. And what they did is they went back and they looked at venous oxygen saturation and its impact on mortality. And out of the 1,800 patients, 300 were excluded, 1,400 included and when you looked at SCV02 on enrollment, they basically found that an SCV02 of less than 70% carried a mortality of over 50%. And more importantly, if you didn't correct it, you actually had a mortality that was decreased to 37%. Now let's check those patients who at arrival had an SCV02 of 70% or greater and at six hours it was less than 70%, that mortality is 48% and if you look at SCV02 corrected at six hours, mortality 37%. So we're talking about almost a 14% reduction. And if you look at the original go-directed trial, the mortality was 46.7% which ironically is almost the same as this trial as well. And so when you look at the overall commentary, this is coming from Gadenone, not me. Process of rise and promise enrolled many subjects greater than 70%. And in addition, the results were positive when you had an SCV02 of less than 70%. And that idea of cryptic shock is that it was present in about a third of patients who were resuscitated with normal vital signs and that persistence of a low SCV02 indicated mild cardio dysfunction as known by looking at your troponins and your BMP products and that patients may benefit from continued resuscitation even after normalization of vital signs. And the same trial was actually published in this article in the Blue Journal just recently and what they showed is that there was an inflection between SCV02 and lactate together. So if a patient who had low SCV02, lactates are high, improvement in hemodynamics reversed the lactate but again as it increased with an SCV02 that was normal, patients actually did worse which indicates cytopathic tissue hypoxia and so goes SOFA score as well as mortality. So again, getting back to the idea that it's the combination that's key of interpretation. And so this slide basically is a slide that's over 40 years old but basically tells us that you can be at different phases of your resuscitation. If you're resuscitated, lactate's normal, SCV02's normal, mortality of about 15%. And if you continue to be unresuscitated, mortality of about 27%. And Dr. Shoemaker identified a patient who were what we call supply dependent at higher deliveries, 51% mortality in those patients that we described as cytopathic tissue hypoxia, high lactates, high SCV02's and that mortality is about 40%. So of course the limitations is that SCV02 does not tell you which organs are hypoxic. It is a general number, that's a compilation of the hepatic veins, mesenteric veins, renal veins, etc. including the myocardium. And so when you look at the idea of one organ dysfunction based on SCV02, you're not going to see it. And in addition, the idea of venous hyperoxia, again based on the fact that supernormal venous saturations actually are helpful. It indicates that perhaps as oxygen goes to cells, normally it should be extracted. But when you develop microcirculatory defects, whether they're from cellular or anatomical issues, your SCV02's may actually be increased, giving rise to the fact that cytopathic tissue hypoxia may be an indicator of microcirculatory dysfunction. And so if you look at Pope and go back and look at that graph and say, well, yes, low SCV02's, high mortality, you also understand why you see hyperoxia associated also with increased mortality. So in conclusion, I don't know if you've ever seen this movie, but the Holy Grail is really an elusive idea in that even with all the knowledge that we have in research about SCV02, we know that it's more sensitive to vital signs. It's a key piece of information in high-risk patients, not for every patient, but high-risk patients. And that low levels are reversible. That's the most important thing is to understand it's reversible. And if detected, improves outcome. And if elevated with high lactates, we should think about other aspects such as cytopathic tissue hypoxia. So it gives rise to the bundle, which we all have in summary talked about in the first three talks we've made, is that these were included in a bundle for reasons. And if you look at what we call the resuscitation piece of the bundle, it was expanded to include multiple concepts. But the idea is to restore tissue perfusion, whether you do it from invasive, non-invasive, or perhaps to other products such as pulse pressure, stroke prior variation, or even from physical exam, the basis of using the concepts of your resuscitation piece to improve perfusion, as well as remeasuring that lactate, is kind of the whole idea of this session. So with that, I want to thank you. I hope we made it on time. You haven't put a hook on me, so I'm all right. So thank you so much. Thanks, Dr. Rivers. Wonderful talk. Our next speaker is Sean Townsend, who is going to be talking to us about SAP. So Sean is at Sutter Health and is a nationally known expert in patient safety and quality. He has been on the Surviving Sepsis Campaign for, I think, 10 years now, and he is also the steward of SAP. So welcome, Sean. Looking forward to your talk. Thank you, Rahul. So for those of you that are paying attention to your program, the first thing you should probably notice is I'm not Laura. Laura Evans was supposed to give this talk, and I was fortunate enough to be able to fill in for her today. I don't know about the rest of you, but at this point, I don't know if I should examine my patient, if I should check an SCV02, or if I should check a lactate, or if I should do all of those things in some combination or another. But the good news is the federal government has made this easy for you. There's a sepsis quality measure, and it's called SEP1, and I'm going to talk to you about it today and what it might be able to do for your patients who have sepsis and septic shock. I have to first make a disclosure, since I didn't fill out a form prior to this meeting since I'm subbing in for Laura. I have no industry relationships or funding to disclose. I am the CMS measure steward along with Dr. Rivers for SEP1, and I'm a member of the Surviving Sepsis Campaign Steering Committee, and today is my last day in that role. So what elements of SEP1, the sepsis measure, apply to septic shock? We have three elements, and these are listed here. These include checking an initial lactate, administering a broad spectrum antibiotic, and obtaining blood cultures before giving antibiotics, and all those things are done within three hours. But those all have to happen to patients that have shock as well. The measure requires this of all patients with severe sepsis or septic shock. There are requirements within six hours in the measure, and it says that we should repeat a lactate if the initial lactate was greater than two. If the patient is in septic shock because they have hypotension, initial hypotension that is, or a lactate greater than four, the patient should receive 30 mLs per kilogram of crystalloid fluids. And if that hypotension persists after the fluid is administered, patients get vasopressors applied. So clearly all of those things were applied to our patients who have septic shock. And then lastly, the measure calls for patients who remain hypotensive after they get that fluid bolus or their initial lactate had been greater than four, that they get some kind of a repeat assessment of their volume status and tissue perfusion. So looking at those things overall, I guess I'd have to say all of them apply to shock. And the measure itself is designed to take care of patients with both severe sepsis and septic shock. Here's an interesting question. I'm going to show you data from the first quarter, or rather the fourth quarter of 2015. This is the first quarter in which the measure was actually effective in the United States of America. And so I'm going to represent to you that about 100,000 patients were collected in that quarter. And that's true. On average we gather data on about 100,000 patients in the United States who have severe sepsis and septic shock. About half of those patients end up being excluded for a number of reasons. The largest reason for exclusion is they didn't meet definition or criteria for severe sepsis or septic shock. And the second largest exclusion is that they were transferred from one facility to another. So you end up with entering about 50,000 patients into the measure. So as regards this three-hour bundle, the first pieces of care, 100% of those 50,000 patients receive the measure. And so that gives you a sense that all of these patients are going to get treated with those three elements. But if we look at the patients who end up getting the fluid bolus, these patients have hypotension, initial hypotension, or lactate greater than four. It turns out only 25% of the patients qualify for that within the SEP1 measure. So 75% of the patients then don't meet criteria to receive a fluid bolus under SEP1. And then if we look at patients who remain hypotensive after they get that fluid bolus or they had an initial lactate greater than four that qualify for the repeat physical exam, that's 8% of the total that came in in the beginning. So SEP1, as you're starting to be able to tell, is largely a measure that treats patients with severe sepsis. But clearly, there's a good fraction, a quarter of them, who present with hypotension and end up being treated for shock ultimately. And that fraction that ends up on vasopressors at the end of all of this is 5% of that initial 50,000 patients in a quarter that came in. So I find that fascinating, actually. I love those numbers when I look at them. I think it's really interesting to see how this plays out. And I didn't expect this when we first built the measure, that this is exactly how we would see it. But one of the lessons you can take from this is that SEP1 is pretty strict about defining shock. In order to qualify for shock, you have to have initial hypotension measured by two or more blood pressure elements in the chart. So it's not just one that triggers that. And then you have to get a fluid bolus, and that fluid bolus has to be completed. And then you have to be persistently hypotensive after that with a check of two or more blood pressures to prove that you're persistently hypotensive. So the measure is very rigorous about defining who actually ends up being in shock, and a lactate greater than four is the exception to that. That also is included, as opposed to patients who are relegated to the category of severe sepsis under the measure. But the point I wanted to make principally with this slide is, let's look at the 100%. How can we prevent shock? How can we prevent the progression from severe sepsis to septic shock? Well, what does everybody need in this measure? That is, they need the first part, the three-hour bundle. The 100% of people who receive that, the progression to the other elements down below. And sepsis is a progression, right? It's the most biologically plausible explanation of what sepsis is. This doesn't compete with other understandings that we have about sepsis. The new definition under sepsis-3 calls sepsis a dysregulated host response to infection. So right away, it begins with the premise that sepsis is a progression. It's progressing from infection to sepsis. And so that's the starting point. And then we know, and we're starting to understand, more about the protean manifestations of sepsis and how people react and respond, what that dysregulated response looks like. In this paper by Seymour et al., with these pretty graphs to the right, these are different phenotypes of sepsis based upon their expression of different patient characteristics that are starting to be identified. So you progress from infection to these various phenotypes. There's some evidence in the literature for progression, and this is sort of an important point. I'm trying to advocate here today that the way you prevent getting into shock in the first instance is to not allow yourself to become more and more ill. But the question on the table is, is it truly a progression? Surprisingly, there's very little data that looks at this in the literature as regards the nature of sepsis as a progressive disease. We do have these three studies. The first one says that 20% of ICU patients with infection in SIRS criteria progressed to severe sepsis or septic shock within 10 days of ICU admission. The second study concluded that among ED patients, patients with sepsis without organ dysfunction, progressed to severe sepsis, so with organ dysfunction or shock, in 18.6% of patients. And then the last one found these characteristics that were associated among ED patients with progression to septic shock from severe sepsis. That was age, female sex, comorbid conditions, and a vascular access site infection. But this study found that antimicrobial administration at 24 hours among these patients was similar among the groups. But that was at the 24-hour mark. So the real question is, something like SEP1, which treats patients within the first six hours of care, what happens when you apply antibiotics early to patients? We have some evidence that's emerged in the literature in the last couple of years. This is a paper done at the University of Kansas by Wiles and others. And this study looked at the progression from severe sepsis to septic shock. So they enrolled about 4,000 patients in the trial. Ultimately, about 1,000 of them progressed to septic shock. And this progression, as it turns out, if you're giving antimicrobials, you can prevent this progression. And through multivariate analysis, they demonstrated that that's about an 8% progression per hour of patients with severe sepsis if they don't receive proper antimicrobials before they become in septic shock. The study actually has more meaning than that. It's not just that there's an 8% progression per hour. But in fact, there were some very other interesting findings that are kind of buried beneath that you have to look at. The first one is that the median time to antimicrobials in these patients was 3.77 hours in those who progressed versus 2.76 hours in those who did not. So SEP1 has a three-hour antibiotic requirement, right? We're telling people, get broad-spectrum antibiotics into the patient when they've got confirmed severe sepsis within three hours. And so that meets this threshold, that 2.76 would be within that range. Antimicrobial progression showed that the time to first antibiotic as well was associated with progression, as well as the time to broad-spectrum antibiotics. So both of these had positive odds ratios. But not only were these associated with progression, the time to initial antimicrobial in this study was also associated with in-hospital mortality, so leading to death. So I think you therefore have the progression we're looking for. The progression is infection, sepsis, progression to shock, mortality. And we're seeing this now pretty clearly in this paper by Wiles, and there are many other publications which demonstrate that time to antibiotics make a difference. But what's happening right now is kind of interesting. Starting within the last two to three years, what we've seen in the literature are a variety of editorials, opinion pieces, appearing in major journals. Here in JAMA, the Analysis of Thoracic Surgery Society, as well as in the American Journal of Respiratory and Critical Care. And by esteemed authors who we all know and respect. And an argument is being advanced that essentially early antibiotics and severe sepsis is not necessary, and that one hour doesn't matter in terms of delay. These papers, all three of them, and there are others, but they tend to cite each other, all call for delay in administration for cases of diagnostic uncertainty. And they cite a fear of retrospective measurement, such as something like SEP1, right? SEP1 which looks and says, of your sepsis patients, how many got antibiotics? And then you get a label that you failed to achieve early antibiotics at a certain point of time. And so they cite this fear of being measured. But they all have one other assumption, which is that they presume that physicians will precipitously act on the fact that this data is out there, that every hour of delay makes a difference in terms of delivering antibiotics. That somehow or another, because data has now been published that says each hour of delay makes a difference, therefore physicians will not think, they will immediately leap to antibiotics, and they will give the antibiotics to that patient simply to meet a quality measure. This has started to reach the doors of CMS, and presently CMS is being lobbied to eliminate the antibiotic requirement for patients with severe sepsis. I told you that 75% of those patients in the measure presently meet the criteria to get in the measure because they have severe sepsis. But this concept that doctors won't think and that doctors will just precipitously give antibiotics, I think is not really borne out by our experience with SEP1. This is a figure of the compliance rate with the National Quality Measure, SEP1, published in 2018, done by Ian Barbash and other colleagues. What you can see, SEP1 compliance here is on the y-axis. In the meeting compliance, every hospital in the country is listed on the x-axis. So all of these bars represent a hospital. There's about 4,000 of them. It's clear that SEP1 compliance, the mean, is about 50% across the country. Hospitals and doctors have no trouble at all deviating from SEP1. If they should want to not go ahead with some of the therapy that we provide, they don't provide it. And I think this really challenges the notion that a sepsis measure is promoting the administration of antibiotics in the country. If you don't think a patient has sepsis and they don't need to be treated with antibiotics, because of SEP1, it's highly unlikely, based on data like these, that you're going to go ahead and just give them them because you want to meet a quality measure. We do know, though, that even though people deviate from SEP1, SEP1 has shown good construct validity. That measure has actually, compared to other measures of quality, which are well established in hospitals across the country, it correlates well with them. So even if you score 59% on your SEP1 compliance score, that same score correlates with how you do in other quality measures. This study, again by Barbash, suggested that measures that are established, such as time to head CT for suspected stroke patients, aspirate administration to chest pain or patients that have AMI, or time to EKG, all these quality measures, SEP1 performs similarly. Hospitals that do well on those do well on SEP1. And so construct validity is an important point to understand. It means that the measure itself is a marker of quality. People who follow SEP1 at a hospital's best compliance rate indicates how well they take care of sepsis patients. Let's talk about this movement to delay antibiotics and severe sepsis and what their arguments are. The first argument that these individuals and these papers and these editorials, I should say, advance is that the association between time to antibiotics and mortality is much stronger for septic shock than it is for severe sepsis or sepsis without shock. They point to the paper in the New York State Publications, and it was in the New England Journal of Medicine, as well as an analysis in Northern California, which showed that there was a strong association between hourly delays and septic shock, but not as strong of an association in cases of severe sepsis. And then a couple of these editorials cite a randomized control trial of pre-hospital care where antibiotics were given in the ambulance to patients that had sepsis, but most of them did not have septic shock, and it showed no difference in mortality despite a 90-minute lead time in delivering antibiotics. And these papers argue the following points. Many conditions can mimic sepsis, so giving antibiotics is inappropriate and causes excessive and unnecessary usage. The paper cites a single paper to say that 40% of patients treated with possible sepsis actually have non-infectious conditions. They point to the experience of the pneumonia metric introduced by CMS many years ago that led to actual increases in prescribing because doctors did want to meet that pneumonia metric. And then finally, just that antibiotic exposure is itself harmful, leading to increases in C. difficile infections, other adverse events like organ failure, theoretical injuries such as mitochondrial injury, the emergence of resistant fungi, and C. diff. So those scary things make a case that perhaps we should not be giving antibiotics to patients that have severe sepsis, but that's the limit of the evidence. Notice that you're just questioning the strength of an association. They have no actual trial that says the association doesn't exist. They question the adjustments that these observational trials made to say, well, there could be bias, but there's no demonstration of actual bias. The trials don't show that there is no association between survival in early antibiotics and severe sepsis. We don't counter the evidence that's available in the literature. It's just a suggestion. There's no quantification of how inappropriate these antibiotics are, how often it is that CEP1 or other measures are causing the indiscriminate administration of antibiotics. We can't say that it's increased from 2015 by 10%. There's no number attached to it. And we don't know how much harm, how much C. diff, how much emergent resistant organisms. None of this is quantified. It's all theoretical. And so what we're left with is a bunch of editorials, perspectives, opinions, viewpoints that cite each other. They literally point to each other in the literature as evidence for themselves, but there's simply no total evidence overall that there's safety in delay. There's no study done that says it's okay to wait three or four hours. Moreover, there's no real evidence that CEP1 has increased antibiotic usage. The last analysis that was done was done by the CDC, published in 2016. And it showed in the status quo before CEP1 that overall rates of antibiotic usage did not change from 2006 to 2012, but there were changes in prescribing patterns that has been happening. You would think that in order to make an argument that we shouldn't be covering severe sepsis patients with antibiotics, you would have to show somehow that there was excessive and unwarranted usage with the measure. But there's been no showing of that. In fact, we have no evidence that antibiotic usage has increased at all since 2006. And this shouldn't surprise any of you in the room. Misuse of antibiotics is pervasive and it precedes the measure CEP1's development completely. This is a paper from 2003 stating that one-third of antibiotic prescriptions in hospitals involve potential prescribing problems such as giving antibiotics without proper testing or evaluation, prescribing an antibiotic when it's not needed, or giving an antibiotic for too long. It seems like I've heard this since I was in medical school. It's the same story over and over again. And so it's not clear again that there's been any unique association to CEP1 increasing antibiotic prescriptions. This is the paper that they cite to say that 40% of patients who are treated with antibiotics in hospitals actually do not have sepsis. But it's a post hoc analysis. In fact, it's done in two ICUs in the Netherlands. And it occurred before CEP1 was published. So again, it describes a status quo before this measure existed. That ICU physicians will treat patients where they suspect with sepsis and organ dysfunction with antibiotics. It doesn't describe after CEP1 what the effect is. They said 40% of patients didn't have sepsis, but that's incorrect. If you look at the paper in the graph to the far left on the overall, it's only 13% of patients that did not have sepsis. The other 30% had possible sepsis. And you tell me, if I'm a critical care physician and someone shows up in my ICU with an infection, and they're hypotensive, or they've got an organ dysfunction, and I think they might need an antibiotic, should I withhold it? Sure enough, in a post hoc analysis like this, you've got the advantage of a study team that looks at culture data and all of that type of stuff after the fact. And they can decide what you didn't need to give it. But in the moment, when I'm there treating these patients, I'm stuck making a decision. And so because of the high and unacceptable mortality risk with sepsis, of course I go ahead and give the medication. In summary, with regard to mortality risk, the Infectious Disease Society of America said in 2012, because patients with severe sepsis or septic shock have little margin for error in the choice of therapy, the initial selection of antimicrobial therapy should be broad enough to cover all likely pathogens. That society has changed their position today. They do not believe this any longer, and they suggest that there should be further delay before you make these decisions. I would point out that IDSA used to agree with us in the Thriving Sepsis Campaign when they endorsed the guidelines in 2012, that once a positive pathogen has been identified, you should apply antimicrobial stewardship within 48 hours. And that's probably the right answer here. Don't continue the antibiotics, take them off, but treat those people who present to you who are sick. I'm very quickly going to show you the amount of evidence in favor of early antibiotics and severe sepsis, because I think the point was well made by the WILES trial. But it's out there. This is a meta-analysis that was done by authors from the National Institutes of Health. In the green panel, you see eight studies, all of which studied early antibiotics in patients with severe sepsis and septic shock. And you can see in this meta-analysis that the odds ratio was positive, 1.34, with a significant confidence interval for delivering early antibiotics to these patients. This is a paper done in Northern California of severe sepsis patients, 35,000 patients. And I would just draw your attention towards the bottom, where it says severe sepsis only in the fully adjusted model. In those patients, the odds ratio was 1.07, with a significant confidence interval suggesting these patients should have early antibiotics. That odds ratio reflects each elapsed hour of delay. So for each hour, there's an increased risk with failure to administer antibiotics in these patients. In New York State, 50,000 patients were analyzed in the New England Journal of Medicine. And there we see the association between the in-hospital mortality and each hour of delay. People who have advocated opposite to this opinion have said that because in the fraction of patients that didn't have vasopressors, the odds ratio becomes statistically non-significant, that these patients maybe shouldn't receive antibiotics. But there are questions about this. Just not receiving vasopressors doesn't mean you're in shock. Are these patients fully volume resuscitated, or were they just initially applied a vasopressor? There's no indication in this paper that these patients were fully volume resuscitated before receiving the vasopressor. How about, did these patients qualify for shock under sepsis 3? Did they both have a vasopressor requirement, were volume resuscitated, and had a lactate greater than 2? We don't know. And we finally don't know how this paper analyzed patients who had cryptic shock. Lactate greater than 4, you've heard about some of that earlier today. So again, drawing a conclusion that these patients all were in shock, I think is premature. I'll point out that the New York State analysis had double the patients analyzed in a separate journal, the American Journal of Respiratory and Critical Care. And on the bottom, you see that the odds ratio associated with antibiotics in 3 hours was 0.78. So these patients benefited from insevere sepsis and septic shock from the early administration of antibiotics in a larger sample than in the paper published in the New England Journal of Medicine. Another paper doing an elegant analysis using a different technique. It was a retrospective cohort study using a propensity analysis to demonstrate the effect of antibiotics in time. Here we see again at about 125 minutes into delayed therapy, there's a breakdown and a separation between those who live and those who survive in septic shock, or rather in severe sepsis with the administration of antibiotics. So, getting close to the end. The question now, back to delay versus progression. All these trials show in summary that a delay in severe sepsis leads to more than just a progression to shock. They show that there's an association with death. We contrast this with the position that's emerging in the literature that some delay is okay, based upon editorials and no trials. In the end, there's no evidentiary basis to remove severe sepsis from SEP1. And so, my conclusion then is, what's the best approach to treating shock? Preventing progression in the first instance. Screening in the three-hour bundle. Thank you. Thanks, Sean. Our next talk is on ultrasound, and I have the pleasure to invite Dr. Enyo Ablodepe. Hopefully, I'm pronouncing that right. To give the talk, she is an assistant professor of anesthesiology and emergency medicine at Washington University in St. Louis. All right. Good afternoon. So, I've learned a little bit about how to continue to take care of septic patients in this holy grail effort that we take. We're going to be checking lactates. We need to do our physical exam. We need to check SCV02s. We need to make sure we do our bundles and reassess. And you heard within there, things about volume, in terms of giving a bolus and then resuscitating with questions about fluid responsiveness. That's a big question that's been in the literature for many years, and people are still not sure exactly how to discuss it. I wanted to give an opportunity to talk about how you can use ultrasound and kind of a personalized approach with each patient in terms of fluid resuscitation and assessing for when to become fluid tolerant. I have no disclosures. So, we'll start with a case, let's say a 63-year-old female, spontaneously breathing lupus in your ICU, has urosepsis, hydroureteran sepsis, hypotensive, did the bundle, did the physical exam, SCV02s have been checked, lactate is high, has received five laser fluids, is still hypotensive while on vasopressors. Part of their physical exam is listed there. Some of their labs are consistent with somebody who might be this sick. And the question is, well, what do I do now? Do I continue the fluid resuscitation? Do I continue vasopressors? What else could I consider doing? Obviously, when we're giving fluids, our basic question is the same. If I give fluid, what will happen to the patient's hemodynamics. Is it that my cardiac output will increase, which is what we want, that is the definition of fluid responsiveness, or is it that there will be no change? And it's hard for us to assess that with some of our current markers, such as our physical exams sometimes, and sometimes our vital signs. Perhaps there's sort of graphic ways to look to see if there are improvements in our cardiac output after fluid. But it depends on what your heart kind of looks like. So baseline echo can be helpful in this holy grail resuscitation with fluid. So let's say scenario one, your heart looks like this, it's a classic parasternal long axis view, and you can see here that it looks pretty typical of what you might expect for somebody who has septic shock. It's a heart that's beating hard and fast with chambers that look to be pretty euphemic, especially in that case where they got five liters. But if the heart sort of looked like this, where clearly the systolic function appears to be decreased, perhaps after your fifth liter, you'd probably start to maybe slow down or rethink if I need to give them more, or if this was in the beginning of resuscitation, perhaps you'd be cautious about that 30 per kick bolus. And more importantly, in cases like this of obstructive shock, where the RV is very large and the LV is decompressed, again, how you would manage fluid resuscitation is going to change depending on the case. So the old way that we do it, or the way that people are most historically familiar with is IVC. And I'm not here to talk bad about IVC. It's one way that some people use it to assess volume status. There are some caveats that we should always be aware of. So the punchline is that in extremes, IVC has pretty good correlations with pressure systems or pressure assessments in the body. It doesn't answer the question of fluid responsiveness always, but then in other extremes or other scenarios you may see in your critically ill patient, IVC may not be helpful. The picture in the center is just demonstrating that you as the person taking the image can influence the size of that IVC, whether it looks large or looks small. So in the video that's going right now, you can see an IVC that looks very small. And we would assume that this person's venous pressure is low, and so giving 30 per kick bolus may not be harmful here. Wait for the other clip. I may have to move this because it's not playing. I'll go here. So in this one, for example, here you can see a large plethoric IVC, and here again after 5 liter bolus, perhaps giving additional may not be useful. There are scenarios, though, where IVC may look small, but actually central venous pressures are actually high. This was a case with somebody who had compartment syndrome, and you can see the IVC looks very small, but their CVP was very elevated. And in ventilated patients, the correlation is challenging as well in terms of IVC versus central venous pressure. So these are ways that people have historically looked at volume status assessments, but may not be the most accurate. So back to IVC again and looking at what the literature supports, and again I'll just say thank you to a colleague of mine, Jan Casal, for some of these slides here. So in spontaneously breathing patients, you can see that IVC size was very useful in assessment of CVP, and again in extremes is where this was very useful. A small IVC suggested a lower central venous pressure, dilated IVC suggested a higher one. But the key here is that whether it's large or small did not necessarily predict that the person would be preload responsive, meaning that their cardiac output would be augmented after the fluid bolus. And in mechanically ventilated patients, the data is maybe not as clear. What we do see in the literature is this concept of IVC dispensability being sort of a better marker and there's some numbers that support that. Those numbers for mechanically ventilated patients are somewhere along 12 to 18% if there's a change in your IVC diameter. And these are again patients that are mechanically ventilated with a stable tidal volume. In fact, those tidal volumes needed to be somewhere around 8 to 10, which is different from how we are typically taking care of, let's say, ARDS patients and using lower tidal volumes. So just be aware of that when trying to apply this number. Those patients also all needed to have no arrhythmias and have no dyssynchrony, so have kind of stable passive breathing. These are the calculations that were used in those studies. So punchline here is that IVC dispensability can be useful. Those 12 to 18% can suggest that the person is fluid responsive. They needed to be in mechanically ventilated patients and there are some restrictions related. Maybe there's a new or better way to assess volume responsiveness in patients besides using IVC or perhaps you can use them concurrently. And the question here is, can we individually tell a person's Stirling curve? That's the word we all talk about when we talk about volume status. And so sonographically there are ways that we can potentially determine patient A from patient B because they don't come with a label that tells us what Stirling curve potentially that they have. And the point here is that when I give a certain amount of volume to that person, that I get some augmentation in stroke volume and it depends on the person's Stirling curve. And so sonographically, again, there are ways for us to determine that. One of the ways sonographically we can do that is by estimations of stroke volume. And it seems like it may be challenging, but there are some simplified ways that we can do that. That's been reported in the literature. A simplified way to assess stroke volume sonographically is by measuring an area times a velocity, essentially. You do that in a parastronal long axis view. There's a calculation that's used by measuring the left ventricle outflow diameter, and then you multiply that by a velocity through the aortic valve, and essentially that gives you the stroke volume. But that's just a number, the stroke volume, which we're really interested in is after that 30 per kg bolus that we give the patient, did the cardiac output then increase? So if it was 45 before and their lactate was elevated and they had mottled skin or whatever else, and then we gave them that 30 per kg bolus, did their stroke volume go up, suggesting that they were responsive to that volume that we gave them? In mechanical ventilated patients, this has been studied. So I gave you that stroke volume calculation, and one of those was the LVOT velocity or what we call the VTI. In patients who are mechanically ventilated, that has been looked at so that if you don't want to take the extra steps to measure the entire stroke volume, which includes the area, perhaps you could just look at the variability of that velocity time integral, similar to a pulse pressure variation, where if there are changes in the peak of that velocity time integral, that would correlate with somebody being likely preloaded responsive based on some of this literature. That number, again, tends to be about the same as the numbers I gave you in IVC, about 12 or 18 percent change in the peak of that LVOT velocity. Again, some limitations here, so patients who are spontaneously breathing, it was challenging to do that on them as they can adjust some of those values and if they had an arrhythmia and if you had lung compliance. So passive leg raise, we've heard a lot about that in the last couple of years, using that to help assess somebody's fluid responsiveness. The point of the passive leg raise, as you recall, is to simulate a bolus for our patients. So instead of giving them that fluid bolus, you would simulate a fluid bolus of about 200, somewhere to 250 cc's by raising their legs up, making sure you're holding it about at least two minutes. And you still need some way to measure a stroke volume, though. So whether you do it sonographically or some of these other non-invasive modalities that are out here, you would do a baseline, then you would simulate the bolus by doing the passive leg raise, and then measure that stroke volume pre and post. And again, an increase of about 12.5 percent suggests in that patient that they would be fluid responsive. Again, limitations always exist. So if somebody has an arrhythmia, again, if they're spontaneously breathing, some of the literature was not as clear cut. So this is an example of that velocity-time integral that would be measured pre, then you would do the passive leg raise, re-measure it, simulating a bolus of about 300, 250 cc's, and if that velocity-time integral increased, you would then presume that the person was preload responsive. So let's go back to the case. I made them younger, but they were actually an older case. This could be any age, really. Same case, they have urosepsis, they're hypotensive, they've gotten a lot of fluid resuscitation, all their markers suggest that they have sepsis, and in this holy grail search, we want to know should we give more fluid, what should we do about the vasopressors, et cetera. And what you can see here, again, is that you would check a baseline velocity-time integral at that apical four-chamber view, trace it, it'll give you a value, and then you would give the person that fluid bolus, or passive leg raise, or albumin, whatever your preference is, or bundle, and then you would recheck it and see if it has augmented. If it has, again, it would suggest that the person is fluid responsive. Changes in this particular case, the example here I'm showing you, that person went up by 28%, suggesting that even after five liters, that person right at that time that this assessment was done would still be fluid responsive. And so this may be a way to individually, again, guide and determine where the Stirling curve of that person is. So this is a schematic of the same visual that I just gave you here, that you would check velocity-time integrals before, give a bolus, and you would expect to see an increase in that velocity, thus increasing the stroke volume, which is the goal of giving fluid. And at some point, as you continue to give fluid, or if you're simulating it by passive leg raise, at some point you're going to become fluid tolerant, which means that your velocity-time integral, or your stroke volume, no longer really increases. And at that point, for that patient, that means that the resuscitation in terms of volume is no longer indicated. That doesn't mean that for the rest of the shift, or the rest of the day, that's the case, because we think that people's Stirling curve sort of changes throughout their course, so this will have to be reassessed. So there's a serial reassessment nature of this. Okay, so what about fluid tolerance? When I get there, to that point where the stroke volume now no longer changes with fluid resuscitation, do I keep giving fluid, or perhaps I should think about diuresis? We run into this case a lot, where at some point we overshoot, and now we need to diurese the patient, or perhaps they've gotten better with their illness because we're on the correct antibiotics, and we've done source control, et cetera. So here I'll talk about a concept called left ventricular filling pressures, and this is well-known in the literature, but hasn't quite permeated, I think, the critical care environment, whether in the ICU or the ER, and I think that that is this, about fluid being harmful is a concept, again, that we're starting to discuss. Fluid obviously very useful in the initial resuscitation, but at some point it becomes harmful when the person only becomes tolerant as opposed to responsive. So I think in this case here, with this heart beating like this, I think, again, if that case, that patient on their fifth liter, we probably would start to be cautious about giving more volume here. The case that we're probably missing where we overshoot and are then having to spend a lot of time diuresing might be this case here, where the person's systolic function is fine, but they have impaired filling that's hard for us to see because visually we're used to only measuring systolic volume, and so there are ways, sort of graphically, that we, with point-of-care ultrasound, can assess their filling capabilities as well. So fluid status, again, the Holy Grail of assessments, we're previously doing things like IVC here all the way to the left-hand side, and really the information we want is over here at the left ventricle and diastolic volume, and there are circuits that we have right now, even the wedge pressure is getting close, but the closer we can get to this information, the more accurate we believe it will be regarding each patient and their resuscitation during their septic episode. So if you've heard of diastolic dysfunction, there's a lot of variables for how to do it, and it's very complicated, so we appreciate those who are very facile with it. There are several components. Typically listed are five validated ways to assess diastolic volume, and that's challenging for us to do. We've limited it to probably only two of those five, and we think that that, which is reported in the literature I'll share with you here, is useful for, again, determining each person's sterling curve and determining when they become fluid-tolerant, when we should perhaps either stop our volume resuscitation or starting to consider diuresis. We're going to do that with what's called an E-to-E prime. You do that by first getting an apical four-chamber view. It needs to be a pretty good apical four-chamber view, so if your image quality is poor, it's probably best to not make some of these assessments. And after you get your apical four-chamber view, you're going to make some Doppler measurements. The first one will be pulse-wave Doppler at the level of the mitral valve, where it's opening and closing here. You're going to get a tracing that looks like this, and it just looks at the velocity of the blood going from the atria into the ventricle, and we're going to be measuring the peak of this, called the E, or the early diastolic filling. Then we're going to measure the tissue velocity instead of just the blood velocity. We'll do that by putting, again, Doppler, in this case tissue Doppler, at the septal and the lateral annulus of the mitral valve, so simulated here, and we're going to get a tracing, and we're going to look beneath the baseline now at what's called the E-prime. It's still the first early diastolic filling, how the tissue moves when blood first fills in that ventricle. The ratio between the E and the E-prime is reported in the literature to help us understand what that left atrial pressure is, which is really what we all want to know, again, what that left ventricular end diastolic volume is. In a real case, it might look like this, so this is mitral inflow. This is that first E-wave. You would measure the peak of that. Again, you would do the tissue Doppler, and this is the E-prime. You would measure that, and the ratio of those two will give us a sense of where the filling of that patient might be. Here's a schematic of how that might look, and if you've ordered an echo or looked at your own echoes you've ordered before, you might see when you look at the diastolic area, they say things like normal, grade one, pseudonormal, or restrictive. Again, we've decreased it to just two assessments, inflow velocity, and then the ratio of the two instead of all the assessments that the sonographers are typically doing, but it's still very useful for us at the bedside to assess our patient's filling. So, normal would be an inflow velocity that sort of looks like so, and you're measuring the E to the E-prime, and normal here would be less than eight. Impaired relaxation, you notice that the E is smaller than the A with a smaller tissue velocity at the bottom. The scary thing here for us is that the pseudonormal or the grade two really looks like the normal, and the only distinction between those two here is our tissue velocity where this is high and this is low, so the ratio between the E to E-prime here would be elevated, suggesting again that somebody has high filling pressures in their ventricle, additional volume for them will not be helpful. And then restrictive patterns are clear cases of people who have amyloid or other infiltrative diseases, a high E-velocity and a very low E-prime, suggesting a large number. It sort of looks like this. Take away from it these numbers, so an E to E-prime that's greater than 15 means that that patient who has urosepsis and is on their fifth liter, giving them additional volume will not help them, and they're unlikely to benefit from that volume. Their cardiac output will not increase, and you have an increased risk of actually putting them into pulmonary edema by saturating the interstitium. People with an E to E-prime are nine to 14 with caution, so I say yellow light here, and then E to E-prime of less than eight, it doesn't mean that they will be fluid responsive, it just means that their wedge pressure is low, and so fluid here is not harmful. You would do the same assessments you did when you gave them 30 per kick to see if it got better by checking their stroke volume. So then you resuscitate the patient, then their x-ray sort of looks like this, or perhaps their x-ray looks like this when you first get them because their substance is from pneumonia, and of course I'm showing you a bad x-ray with a bad CT scan, and the question is should we give them diuresis? This is often a clinical question. This is an ultrasound here of lung looking at what's called B-lines, suggesting that there's lung water, but we know that there's cardiogenic reasons for B-lines and non-cardiogenic reasons for B-lines, and the assessment that I just went over, E to E-prime, can help us decide for a patient with that x-ray or that CT scan, would diuresis likely help that patient? Everybody that has that x-ray does not necessarily need diuresis, or if I'm resuscitating them with fluids and I finally get to this place, I know I need to stop. So looking at this schematic here, it's hard to see, but this is again that same lung ultrasound. There are one or two B-lines, that's normal, so you might look at that as you're starting your fluid resuscitation for your patient, suggesting that they have low filling pressures, you would give them your fluids or do your passive leg raise, and when you start to saturate here and have a lot of B-lines, it's time to stop, or you can do the assessment of that E to E-prime. There are limitations, again, with B, with just looking at B-lines, because you don't know if it is from hydrostatic edema or capillary leak, again, that's why the E to E-prime can be helpful. A high E to E-prime suggests to you that this is probably related to a cardiogenic edema, diuresis would be helpful. A low E to E-prime, even if they're B-lines, means that it's not from a cardiogenic reason, giving diuresis is not going to be helpful. That E to E-prime in the literature, just a comment, also is an estimate of your left atrial pressure when you add four to it, and in the cardiology literature, it's reported to have good association with your wedge pressure, so that's what that E to E-prime can be useful for. In the literature, these references here, E to E-prime has been used to estimate that left atrial pressure or that wedge pressure, again, helping you at the bedside decide the person's Starling curve on an individual basis. Half estimates suggest that an E to E-prime less than eight means that they have low wedge pressure. It doesn't mean they'll be fluid responsive. You'd still have to do your assessments of stroke volume, but it means that giving them fluid here would not be harmful. If it's greater than 15, it means that it's likely to be harmful, additional fluid resuscitation, and you would consider other things like increasing your vasopressors or better source control, et cetera. And then the gray zone, really, in the literature, between eight and 15. Here's literature looking at mechanical ventilative patients with that E to E-prime and sensitivity and specificities with that, with their wedge pressure in cath lab, and they're pretty close. So my summary statement here for you, you want to know if your patient is fluid responsive, they're on their fifth liter, they're still hypotensive, use ultrasound to help you estimate where their Starling curve is by measuring their velocity time integral or measuring their stroke volume. That's what that means. At some point, you're going to get fluid tolerant. At some point, that next bolus you want to give your patient will not help their cardiac output and may increase the risk potentially of pushing fluid into the interstitium and causing pulmonary edema. In that case, you want to use things like lung ultrasound as well as LV filling pressure assessments with E to E-prime to determine when that part of the curve would occur. And that's different for every patient, depending on their pre-test probability, what their previous heart would look like, and what their current conditions are. And with that, I'll end. Thank you very much. All right. Thanks, Anil. Time for questions. We are open for questions. from Indianapolis. Now that we have the lactic acid checked in the triage area in the ER, fluids, antibiotics given very quickly, do we have an estimate what percentage of these patients actually end up hemorrhagic shock or MI after we give them the antibiotics and the 30 mL per kilogram of fluids? Maybe what if Sean can answer the question? I think I understood your question to be how often is it that we find conditions that mimic sepsis after they've been initially treated for sepsis. It's actually a space where there's not a lot of study done. There's actually one paper in the United States that said in the U.S. Emergency Department that patients who were initially treated for sepsis, 18.6% of them, actually did not end up having a sepsis-like syndrome. So that plus the Clauenberg paper that I talked about, the 40% figure that is actually more like 13%, give you a sense of where it might be. So 13% to 18%, something like that. And so that's the direct answer to your question, what's the number? And then I think to add on to your question, it becomes a question of what is the mortality risk for the patient if you don't treat them? What's the cost-benefit ratio to deciding whether to give that patient antibiotics or not? And I think people do their best to be careful, but I think the high mortality rate in severe sepsis and septic shock itself, 20, 30, 40%, depending on what condition you're looking at, causes people to treat. Very good question. The other issue is you treat other diseases in a similar fashion. If a patient comes in with a stroke, 10% of them who get TPA are stroke mimics. They're not strokes. So when you look across the board in terms of how we practice other diseases, it's the same issue. The problem is is that stroke treatment is based on morbidity, not even mortality. But we'll stop, do a 30-minute stroke scale, we'll have a CT scan and a neurology consult, and 90% of the patients never get treated with TPA. So when you look in terms of how we manage other patients, it's a similar fashion. I look at it as a combination of both what's my clinical suspicion and how sick is the patient. So if I have a low clinical suspicion but the patient is hypotensive, I'm going to treat them. If I have a low clinical suspicion of infection and the patient is not sick, then I feel like that gives me more time to think a little bit more fully, do my diagnostics, do my x-rays, do my whatever. And I'm not saying prolonged delay, I'm saying the difference between what's an emergent administration of antibiotics, patient's hypotensive, or you've got a high index of suspicion and the patient's moderately sick, I call that emergent. And then you have urgent antibiotic administration, and that is maybe I have a lower index of suspicion if I haven't sorted it out in an hour, two hours, or up to my three-hour mark. Once you're getting close to the three hours, then I would consider whether I needed to pull the trigger or not. But if they're sick, or I have a high index of suspicion and they're moderately sick, then they get treated like I try to get that within one hour if possible. Thank you for a great presentation. My question is, we know time to antibiotics improves outcome. How about time to starting pressers? We know that we have to give 30 mL, but is earlier pressers, especially patients who are not preload responsive? I think the best answer I can give to you is that's probably what happens in practice in any event. The bundle that I showed you, the sepsis CMS measure, shows it happening sequentially. It pretends that patients who remain hypotensive after full volume resuscitation then get vasopressors added. But in reality, I think in my ICU and probably yours, someone who comes in 60 over 40, while they're getting their first liter of fluid is probably also getting their norepinephrine infusion started if I have central access. So as a matter of what happens practically, I think we live that. And then what the science says about that is a whole other question. There is a trial ongoing now, the Clover's trial, which is designed to look at restricted fluid administration and the addition of vasopressors earlier. Dr. Rivers is an investigator on that trial and I'm not sure where it stands at this point in time, but I think that answer is coming. There are a few retrospective studies that may suggest that earlier use of vasopressors in cases of severe hypotension might be beneficial. But again, that's with the caveat that they're all observational retrospective studies. Thank you all for the lectures. Just I want to ask Dr. Mani Rivers about early goal-directed therapy. I know that the other challenging three trials, all before randomization, they all give fluids. So I think the fluid issue is not the question. But what about the blood and the dupitamine? I mean, do you still practice giving blood to a target of 10 and also dupitamine before the NOEP or do you change your practice? Thank you. Very, very good question. Otherwise, remember people treat goal-directed therapy, again as a noun. It's not a noun, it's a verb. And it is comprised of all of the prior work that's been done to basically treat a septic patient. Remember, this was in 1997 when the protocol was written. So you have to understand that it reflects the standard of care expert recommendations from the American College of Critical Care Medicine. So I didn't make it up. It was something that was a component. Now if you look at the result in literature, there's an article published in New England that looked at hemoglobin strategies for septic shock. And the question whether or not seven or eight versus nine versus 10. If you look at that paper, it said there is no mortality benefit from giving blood for the way that trial was conducted. But if you go into the appendix and you look at what the baseline lactate and SVO2 was, they were normal. So you're giving blood to people who don't need it. So I don't give blood to people who don't need it. If you've got tissue hypoxia and you've got evidence of myocardial dysfunction, I'll give you blood if that's a pathway. But remember, only 15% of people get it. So when you look at the isolation of blood in and of itself, it's very controversial, yes. But you're talking about a patient with a mortality of 50%. And if your lactate is high or SVO2 is, what I see a lot of times is 40 or 50, your CVP is high, then you need blood. You need oxygen carrying capacity. Now, if we say that we're not going to do that, we'll conduct a trial and put yourself in that clinical situation and say, I won't give blood, but that's never been done. All right. Go ahead. Thank you. I'm a huge fan of ultrasound and I love the fact that diastology is creeping into our ICU. Regarding the E to E prime, should we be cautious that it still is a static index? I mean, people have now shown that dynamic indices are better predictors and even if it does represent wedge, we know that wedge is really a poor predictor of fluid responsiveness. So how can you, can you think of ways that we can use E to E prime as a dynamic variable? Yeah, that's a great question. So first caveat, and I didn't get into it, I only had 20 minutes. I just gave you an estimate about E to E prime. There are lots of times you should not use that. So if somebody has any mitral valve pathology, that is not an accurate reflection of their filling pressure. So if they have macular or annular calcification, they've had a replacement of that valve. Sometimes if they are very tachycardic, heart rate greater than 100, this may not necessarily be very reflective and absolutely it's a static marker. So I tend to use E to E prime not to tell me when to give fluids because you're absolutely right. Wedge pressure does not tell us, even when low, that we will be fluid responsive because all of us in here should have a low wedge pressure and giving us a 30 per kick bullish right now is not likely to augment our stroke volume. I tend to use it for who I should not give or be cautious to give fluid to because their filling pressure is already high, telling me that likely giving additional volume will not improve cardiac output. And I tend to use it for who to diurese or not diurese based on the radiograph. So I don't tend to use it for the first part of the Starling curve. I tend to use it more for the second part of the Starling curve. Does that answer that? It's likely used as a stopping point and not for assessing fluid responsiveness. So the extravasation of water depends on what the filling pressure is. And so it's probably a stopping rule and not for fluid responsiveness. Go ahead please. Thank you for the lectures. My question might not be clinical but most likely related to the comment that was made earlier about the SEP1 status and the perceived pressure in terms of the antibiotic administration, which we all know that the antibiotic administration probably have the highest evidence in terms of the whole bundle. What is the status on the other parts of the bundle? Are we receiving the same pressure and hypothetically if we drop the antibiotic administration, what's going to happen to the whole bundle issue? Well, you'll end up with new CMS measure stewards probably if we drop the antibiotic issue. I'm kidding. Yes, the whole experience of being the measure stewards for SEP1, I think many would agree, has been an interesting exercise in advocating a national standard for sepsis care and having virtually every professional society of America to have something to say about how and what's wrong with the measure. There's lots of pressure on different components. Primarily, I guess, the next area that would get the most attention is the fluid bolus, the requirement for a strict 30 mls per kilogram for patients with severe sepsis and septic shock, or just in shock rather. And so those two areas I think are the most sensitive right now. But that question has kind of not been as the societies that would be concerned about that have not pressed that point so much at the level of CMS. There's just academic debate on that front primarily. I think the slide I showed that people comply with the measure variably, about 50%, and that the top performers of the measure, actually about 70%, suggest to us it's a relative comparison of what people are able to do at their own institutions. There's no 100% compliance with SEP1. There never will be. And I think a lot of the debate that's driven about these particular elements is that people think that it's a mandate and you have to do it all and you have to achieve 100% compliance. But that's not what we've ever said. We're just asking you show us how often you can do this. Just a quick comment. If I can make an advertisement. This issue of critical care medicine has three articles on an examination of the Medicare sepsis population. The mortality is astronomical. So if you're a Medicare beneficiary or recipient and you go in with severe sepsis, you're still more likely to die at a 30% range. If you have septic shock, you're 40, 50% mortality. So what drives us as stewards, in spite of all of the arrows that shot at us, is the mortality is still very high. And I think it's very striking that this month issue has three articles. Tim Buckman, I think, is the primary author, and there are many others. That's a very astounding. And they put an economic as well as a mortality cost on sepsis and it's still the worst disease as treated. This is more asking for when is the data coming with regards to compliance and mortality reduction because we're seeing it on other things like the ICU liberation research data showing more compliance with various bundle elements, better outcome measures, and I think that would shut a lot of people. It would quiet the critics significantly if we had that mass data being able to show. And I know you previewed a little of it the last time I saw you present, and so I'm just curious. It's always difficult to be in a session and to want to be able to tell you things that I can't say. But I can say this much. The answer to your question is this year, that will be out. We're holding you to it. Oh, this year, absolutely. There's no doubt that's coming this year. The manuscript's written, and both Manny and I have approved it, and it's just a matter of getting it finalized through submission. And I can't tell you another word about it. I have one question for maybe Enyo and Dr. Rivers. This concept of fluid responsiveness and unresponsiveness, and whatever study you look at, for example, the Andromeda shock trial, the concept was to give, in appropriate patients, of course, to give fluid still fluid unresponsiveness. Once someone is fluid unresponsive, they're on the flat portion of their cardiac output curves or their Starling curves. By definition, none of us are on flat portions of our Starling curves. And by definition, if you are there, you have probably gone too far with your fluid resuscitation. So should we be actually targeting fluid unresponsiveness, or should we be doing something else? Or should we be doing something else? Should we actually be targeting fluid unresponsiveness? Is that the right thing to do, or do we do something else? Okay, well, I'm not an expert in that answer. I just speak strongly about the ultrasound. I have a feeling that the Starling curve, there's more to it. I think it probably has a modulating effect that we never talk about. Like the other part of the Starling curve, we never really talk about. So we don't know what happens. Does it go like this, meaning like going down, so you're unresponsive, and now you're, I don't know what's worse than unresponsive. Or is it that at some point after several hours, you become responsive again? Or is it that you shift back and forth on the same curve? So I don't know the answer to that. I know in my own clinical practice and what some of the data suggests, people have only looked at the extremes of that Starling curve, never really answering the question the way that you have asked. Yes, sir. The other thing is, I want people to understand, as you practice, when you go back and you take care of the next septic patient, and we all do, I never take one variable. I mean, this panel talked about, ECHO talked about SpO2. I don't practice that way. I use everything. So remember, the cardiovascular system is closed. There's preload, there's afterload, there's contractility rate, coronary perfusion pressure. So we don't just treat one number. You have to treat it as a unit, a closed vessel. And when you look at fluid administration, perfusion, you look at lactate, venous saturation, use base deficit, urine output, you put them all together. So I think when you go into an ICU and you look at all the different monitors and what have you, and you just focus on one, it may come across that way as we speak at these meetings and what have you. But if you gave me ultrasound, you gave me stroke volume variation, you gave me lactate, you gave me SpO2, you gave me blood pressure, I'll use them all. Because these people are sick. They deserve every possible monitoring that you can have, and you know the limitations, you know the good things and the bad things about them, and you just live with that. And that's kind of how I practice. We have time for probably two more questions, and then we have to conclude. Can the panel comment on the surviving sepsis campaign one-hour bundle, and if it's likely to be adopted by CRS? Okay. So the one-hour bundle, currently as it stands, I think in septic shock patients, you know, initiating the bundle as soon as possible is of benefit. I think that that is a marked improvement to where it started, where it was to initiate all elements of the one-hour bundle within, I mean, initiate all elements of the three-hour bundle within one hour of presentation to the emergency department was more challenging. So I think that where we are right now is probably a better place to be. And sick patients, they deserve your attention. So you can talk about CMS. CMS won't adopt the one-hour bundle. That won't happen. The field's not there yet. It may be aspirational on the part of the sepsis campaign that we could recommend that strategy and try to raise people to that level of performance. But hospitals themselves probably don't have the capacity to function just as organizations with that level of performance, and CMS is not likely to push hospitals to that extreme when we can't master SEP-1 yet. Last question. Good question, though. Hi. Jeanette Berg, Kootenai Health in Idaho. I'd like your thoughts on the high SVO2 patient who you think you have adequately resuscitated with fluids and given antibiotics. How do you approach their supportive care? Again, you look at all your different variables. Now, if you've got a high SVO2, your lactate's normal, you're urinating, your organ dysfunction is gone, that's a good sign. But what it reflects is actually a low oxygen consumption. We don't use PA catheters anymore. If you have a metabolic heart that can measure VO2 directly, that's a good monitoring device. But actually, it's your VO2 that's your basis for life because that's oxygen consumption. So what you're doing is measuring a variable that perhaps reflects it but doesn't really tell you directly that that's really your oxygen consumption. And I think we all got to get back to the basics of knowing what are the basis of supply and demand. Well, your consumption, your ability to consume oxygen, cellular respiration is the basis for life. And if you don't have that adequately, you're going to die. So SVO2 perhaps can help you predict that or help you reflect that, but it is not a direct reflection of oxygen consumption. So just to give you my ideas, it's one thing, but that is a variable that has correlation and perhaps not the greatest. All right. Thank you. The session is now concluded.

sepsis management

shock management

oxygenation variables

timely administration of antibiotics

central venous oxygen saturation

lactate levels

resuscitation indicators

prognosis indicators

SEP1 sepsis quality measure

initial lactate levels

blood cultures

severe sepsis

antimicrobial treatment

outcomes improvement

ultrasound in fluid responsiveness

Sepsis

Lactate

Shock

Guidelines

Presentation

Select

Professional

Associate

2020

Foundational

Intermediate

Advanced

Nonmember

Sepsis Resources