Good morning, and I'm going to talk about the use of minimally invasive or non-invasive cardiac output monitors to diagnose shock. My name is Siddharth Dugar. I am a staff at Cleveland Clinic 2. As you can see, I'm the director of point-of-care ultrasound, so they made sure I don't talk about focus. Otherwise, there will be a lot of bias and conflict of interest that will come into play. I have a grant from Chest Sonocyte Ultrasound to study COVID. When we are talking about, as Matt pointed out, what is the goal when we are thinking about putting a hemodynamic monitor? The first thing is we are looking at can we diagnose shock in the patient? What kind of shock do they have? If we are going to use fluids or vasopressors or diuretics, we want to see if we can choose intervention appropriately and assess the response. Those are the things we are asking from a monitoring device. This slide, everybody has gone through it multiple times, but the point being, if we know the feeling pressures and the cardiac output, we can differentiate obstructive hemorrhagic cardiogenic shock from distributive shock. Hence, most of the monitoring devices are targeting to find us what the cardiac output or cardiac index of the patient is. I'm going to spend some time here because this slide is going to be the basis on which we are going to assess all the different devices. This is basically a bullseye right here. There is something called accuracy and something called preciseness. Accuracy is a common term. How accurate is the value if you are using, let's say, for cardiac output monitor? If we are using a PA catheter, whatever device we are using and comparing it to PA catheter, how accurate is the cardiac output it's giving me as compared to a PA catheter? That's one question. The other is, if I keep repeating it again and again, will I get the same number again and again? That's called preciseness. This is the same thing that is represented on this Bland-Ottman curve. The blue line is what we are getting with the PA catheter or the gold standard. The red is what we are getting with the device we are testing. The closer they are to each other, the better the device. That is what we call bias. There will always be a bias, either a positive bias, which means the cardiac output I'm getting from the monitor is a little bit higher than what I will be getting with the PA catheter. If it is negative, then it is giving me a lower cardiac output than what I should be getting with the PA catheter. But it's a spread that is also very important because we want to be sure every time I repeat the measurement, I will get the same value or close to the truth. This is called limit of agreement. The closer they are, the narrower the limit of agreement, the better the device. The widespread they are, that's a device that you cannot trust because it can be all over the place. And there is another number that we will be talking a lot about is percent error. Simple term, a very numerical value. The closer it is to zero, the better the device. The closer it is to 100, the worse the device. And when we are looking at all the different devices, make sure you look at the limit of agreement and the percent error. So the first one is the finger cuff plethysmography. This is the most non-invasive method. It's the same what we have been doing in medical school since, like, you know, 1900s. You have a cuff that instead of you applying on the arm, you are putting on the finger. It uses what we call volume clamp method. So the cuff inflates to a point that the volume of blood in that finger compartment remains constant. Now the only thing that is changing is the pressure. It sends this to a microprocessor, which amplifies it. So even without any invasiveness, the finger cuff is able to give you an arterial pressure waveform. Pretty neat, very non-invasive, and, like, you know, can be applied in literally one second on the patient. So these are multiple studies that have been done over the years looking at the accuracy of its measurement in measuring cardiac output. So remember we talked about the limit of agreement. So it can be as wrong as minus 2.16 liters to as wrong as plus 1.76 liters. And all the studies have shown that you can be anywhere three liters below what is the truth to two liters above what is the truth. So you have a very wide range that you are working with. And the percentage error, the same thing. Closer to one is a better monitor. Closer to 100 is a worse monitor. And you can see it's close to 50%. So you have a lot of error that you should be okay with if you're using this measurement. It will give you a cardiac output. It can be wrong by plus minus three liters. So use it accordingly. And these are different patient population. Every time you look at it, it's the same. You know, the error is decent enough. The next is the bioreactants and bioimpedance. So this system was used a lot in early 2000. What it does is it has four electrodes that you place on the patient's chest. It runs a low amplitude, high frequency electrical charge through the thorax. And what it measures is the phase shift. The more the blood is flowing from the thorax, the higher the phase shift, the higher the cardiac output. So this is, again, the Blandt-Altman curve that we were talking about. So you can, in this study, they were comparing the NICOM, which is one of the monitors that uses the bioreactants, to thermodilution. So again, it can be as high as 4.1 liter above what we will be getting with thermodilution to minus 2.2. So you have a six liter range that you are working with. Again, if the patient has a cardiac output of six liter, it can give you zero or it can give you 10. So use it accordingly. The one that we are using a lot nowadays is a pulse contour analysis. So what it does is, you put an A-line, it gives you an arterial tracing, it goes into a black box. Literally, it's a black box. I don't think anybody has, anybody sitting here has the equation of how they're calculating the cardiac output. But they give you a cardiac output. So how do they do it? How do they do it? So it's very easy. Cardiac output is your heart rate into your stroke volume. But you have to change the pressure to volume. The simple equation we use is, if you know the compliance of the arterial system, the change in your pressure is the change in your volume. And that's how they convert your pressure to volume. The problem is, what is the compliance of the arterial system? How do you get it? So there are multiple ways to do it. The non-invasive is the one where they are like, we will do it in 10,000 patients, we'll collect their biometric data, and we'll approximate, based on your patient's biometric data, what the compliance of the arterial system is. Okay? Good enough? So we have the biometric data that we put in the patient height, weight, and everything. And it's like, okay, based on our data set, this should be the compliance of the arterial system. Doesn't matter how much norepinephrine they are on. Doesn't matter how much volume we have given. The other is just looking at the pressure tracing. They are like, we have an AI, very smart system. It will look at the waveform and will give you a very accurate cardiac output. Or other one is like, we will use thermodilution. This is the, again, I have no conflict of interest, but there is lithium thermodilution that uses the same thing. You inject lithium, it measures the lithium concentration. And based on that, it assesses the compliance of the arterial system and gives you a cardiac output. Little bit better. But, again, remember the percentage error? So the earlier study, the percentage error was 74%. Should not be using this device. As we are getting to 2013 and 20, the percentage error is getting better. But still, you have a very high bias in this group of patients. So your cardiac output is definitely wrong by 0.5 liters, but it can be wrong by 2 to 3 liters on both direction. So this is another systematic review. This is the patient population we deal with. Hemodynamically unstable patient. Different devices. The flow track is our pulse control analysis, 2 and 3. And the cardiac output, this is the mean cardiac output. And 2 liters is right here. Just make sure. I know the graph is big, so we might not think about it. But this is 2 liters, positive and negative. And you can see, it can be as wrong as 4 liters negative to 2.5 liters positive. And look at the variation that we are seeing here. So again, it's a good, like, you know, there is utility to it. But as Matt said, these are not going to give you the true cardiac output. There are some devices that may be better than others. If you go to lithium thermodilution, maybe there is more narrow, more precise values. But if your patient is realistic, I will go more invasive than all these devices. How often do you, if you are using a lithium thermodilution system, how often should you calibrate it? So the R square value is how accurate it is compared to the gold standard. You can see, like, within the first hour, this lithium thermodilution is doing pretty good. 0.79 is a pretty good value. But as each hour passes, the accuracy of that compared to the gold standard keeps declining. So you will have to recalibrate it every two to three hours to make sure your measurements are still accurate. So my opinion, I like this paper that was published in Current Opinion in Critical Care by Siegel Group, where they said that I don't think the non-invasive devices are there for you to tell what kind of shock the patient has. It can give you some idea, but the best value of these devices are when you are assessing fluid responsiveness. Use them, resuscitate the patient, if the patient is still not improving, maybe a more invasive device is needed. Thank you. Next up, Dr. Casey Cable will talk to us about point-of-care ultrasound approach to the diagnosis of shock. Thank you. So I apologize. I'm getting late. But you know what? It's Hawaiian time. So yeah. So let's talk about non-invasive methods to kind of phenotype shock. And what's the literature? Where are we? What are we going to do? And so I'm at VCU Health, English from Virginia. And I just got an FCCP. Sorry. I'm very excited. It's a big thing. It's a big resource. No financial disclosures yet. So I'm sure this has been shown. This is what we have memorized for medical students. I mean, this is what we know and love. But how do we implement this? How do we actually utilize this chart? Because we know this and we hate it. So ultrasound is great. Point-of-care ultrasound is amazing. So it's easily accessible, typically. It's at the bedside. It's non-invasive, relatively inexpensive. And so it's a great tool that we can utilize to differentiate shock, to figure out what's going on with patients. And then we can have an algorithm to figure out where and how to utilize that. Are there some downsides? Of course. There's always downsides. We have variability in expertise, of course. We've got trainees. We have people that are experts that are not. And of course, it works best in patients that are perfect. Do we get that? No. And so it works best in patients that are previously healthy and only have one type of shock. Do we get combo shock? Of course we do. But we have to realize that. But still, I still feel like the point-of-care ultrasound has great impact for realizing what type of shock a patient has. So what do we use? ACES was first. Who is familiar with that? Quiet audience. So abdominal cardiac evaluation of shock was first. And it looks at cardiac IVC, the aorta, right hipquadrant, left hipquadrant, and pelvis, which is phenomenal. And so that was kind of came about in the late 2000s in looking at how do we standardize looking at ultrasound, sonography in patients with undifferentiated shock. And I'm talking about I am not going to go into trauma, and I'm not going to go into cardiac arrest, because that is a completely different talk and completely different ballgame. So this is just our population, undifferentiated shock. Rush is fantastic. Who uses rush? We have some involvement. Yes. And so it's definitely, it's a sequenced analysis to look at how we evaluate patients with low blood pressure, with shock. POCUS has been used in undifferentiated shock, and it's been fascinating. So it's becoming standard of care. So previously, looking at the evidence, and this is a phenomenal study. I mean, just, I know, done in the last, you know, I don't know, I've got to say five years, but 2017, I know, right? More than that, looking at patients with where they looked at previously and then with ultrasound, and it really improved their diagnostic quality up to, you know, from 60% to 80%. And the effective management, which is very important. So ultrasound is becoming very important, an easy access tool for what we can do. This is, I love this. So shock is the sonography and hypertension cardiac arrest protocol. This is the hypertension protocol, where they have, what they look at is this core examination, what should be done when you have, initially, then after that, if you have time to do supplemental, as well as additional examination, which is phenomenal. And the trial they did, yes, and this is in the emergency department, and we can speak to, I would love to have a conversation with anyone out there speaking to the floor or in the medical ICU, because this is probably an underrepresented population. But it definitely improved the diagnostic recognition of shock. Ultrasound did, for sure. But then looking at the actual outcomes was, it was an fascinating study, looking to see did it actually improve outcomes in fluid resuscitation, inotope usage, admissions, actual length of stay, and it didn't, which is fascinating. Now, I do have to say, I do love this study, and I do a lot of research, and so, yes, I could go on forever about this. Among some of the limitations of the study, in terms of, in the middle of it, it kind of became more so utilization of ultrasound, and how did that impact it. But still, it questions us to say, okay, so we have identified, with ultrasound, for sure, which is a great utilization, we identified it early, then we impacted it early, we did things fast, but why did it not actually change outcomes? It's a fascinating question. For debate, for sure. So, once again, this looks at 30-day discharge survival, didn't do as much as we should have, but yeah. So, I'm just here, quick summary to say, I believe that POCUS, and I can go into full lectures on how to diagnose the various types of shock with ultrasound, I mean, that is, CHESS does an amazing job with that, but you can, all the different variabilities, and all the different, looking at where you can, I think we all can, but early recognition of shock is important. I think we all know that. What type of shock so we can treat that? Why does it not improve outcomes? I'm not sure. It's interesting. But I think going forward and adopting a standard of care with our colleagues out at other locations, so we can say, this is how we do this, so that they can also be abridged to that, I think is very important. Do you guys have thoughts? Yeah, I think one thing that's worth mentioning, so there's a lot of studies about diagnostic testing and how you use it to change your management, and I think the key is, we had a session yesterday where this came up, but if changing your diagnosis doesn't lead to you to change the treatment in a way that's going to be associated with improved mortality, then it won't be able to show that kind of benefit, so that's something to keep in mind. Absolutely agree, absolutely agree. I think waiting until you have frank manifestation of shock is a really difficult thing to salvage. I guess a more crass way of thinking about it is that if you look inside a sarcophagus and find a dead body, you can't be surprised. I think that once your patient has irreversible organ damage, or ultrasound, a PA catheter, or whatever diagnostic modality you choose, the clouds are not going to part and the sunshine is not going to come down. No, I completely agree. Yeah. I think the nice thing about ultrasound is, yes, while we have the luxury of that, it is becoming more and more available in other countries. I've got many colleagues that are using it in Africa and down in South America, which I think is also very key. I will just add, again, it's a diagnostic tool, we should not be comparing outcomes in a patient. In something we are using just for diagnosis, like NICOM, you found a cardiac output, but will it change outcome? If you do the right things, maybe, but again, the mortality is so high, and unless we say like there is one diagnostic or one therapeutic intervention that completely changes the outcome, we cannot say that the test itself changed the outcome. It's like if somebody studies using echo in patient with pericardial tamponade, I'm pretty sure it will change to outcome because there is a needle that you can put in and drain the fluid. But in a patient where majority are distributive shock with a mortality of 30 to 44%, finding that you had distributive shock, you know, it will give you more confirmation, but it should not change the, like, you know, overall outcome of the patient. You still have to do the right thing for the patient. Yeah. Okay. Great time. Thank you. Just about the invasive approach. All right. So for those that are Spaceballs fans, what's happening in the scene depicted here is that you have two people that are arguing is, you know, when is then going to be now? And the answer is soon. My name is Max, I'm from D.C., I'm a CVICU intensivist, and working in a CVICU, people come in on fire. We get calls from, they're on, you know, what's their NOREPI at, 999, and, you know, what's your carrier, VASO. And so what usually happens is that the ICU staff is like, let's go. You'll see a lot of our trainees, our APPs, the attendings, just kind of line up with like many things ready to put in invasive lines. And very commonly, this happens in rapid sequence. And one of the reasons that we put in these invasive lines is to get a little bit of control of hemodynamic mayhem. And so when you want to create more of a hemodynamic phenotype, a lot of these invasive lines can bestow a fair amount of information. There's a time when you need to transfuse blood very quickly, and eventually there's going to come a time where you need to give some potentially venocostic drugs, and that is often not possible with the access where they come from, from the floor or wherever they may happen to have been. The more practical reasons is because not everyone has great windows. I'm not, you know, the master sonographer. I'm not Sid. I am but a humble man with a phase array probe, and sometimes I can't see anything. And sometimes you also want to limit the patient's torture. The 22 in the fingernail that they come in with in the emergency department is probably not the right thing, and it's quite painful. And the clinical scenario is often dynamic in the way where hemodynamic trespass is all but imminent, and often vascular access sways the numerator of the amount of information that you can get in a very small amount of time. When we come up with reasons to put in access, if the reason you're putting in access starts with the word just, it's the wrong reason. So I want to put in the access just in case. Just in case what? Or just to have, or just want to see. That's not the right move. The way that I think about this is that whether it's something like a 20 gauge or as big as an ECMO cannula, you need to have a bidirectional relationship with the cannulas or whatever you have in the patient's body. That is to say, you have to be able to do something for the patient, and they have to be able to provide you with actionable information. Otherwise, that device inside the patient's body is probably there for the wrong reason. So we don't know a lot of things about when to get invasive in shock. There's no holy grail metric. We invoke cardiac output. We invoke cardiac index. Some people target mean arterial pressure, and there are many other fancy metrics that we come up with. But these are very poorly evidenced. And these are even worse evidenced in the intensive care unit. The most data that we have on focusing on cardiac output is actually from the perioperative data. Think way back to the Fedora trial, which looked at perioperative complications and length of stay in relatively low to moderate risk patients undergoing intermediate risk surgery. And what they found is it didn't save any lives, but there was a reduction in bad stuff that happened after the operation. So OPTIMIZE 1 and 2 and 2 is still in progress. They used cardiac output guided hemodynamic therapies, which followed a relatively rigorous algorithm. And they compared post-operative outcomes and 30-day mortalities. I am not a believer in the OPTIMIZE algorithm, believe it or not. FLOELLA is still in progress, and custom is underway in Germany. And all of these trials look at cardiac output and have been somewhat underwhelming in their performance. So when you think of the patient's presentation as an iceberg, there are a lot of different things that we can do to them that have been discussed at length today. We haven't really talked about the physical exam, because the year is 2023. Of course, we use ultrasound, we use arterial catheters, and we're going to talk about as we move down the iceberg. So this is, of course, an arterial line. And as far as I'm concerned, since I've rotated in the ICU as a medical student, this is virtually the price of admission. You walk in, the bouncer gives you an A-line, and the nurse hands you a Foley, asks you to sit there idly, don't bother anybody. And it's relatively common to get A-lines. And it'll give you a morphology that looks kind of like this. It goes up, and then it goes down, and then it goes up again, and then it goes down. So what information can we actually get from an A-line? You can actually get heart rate from an A-line. Well, why would you ever want to do that? Please tell me your patient is attached to the ECG monitor. Does anybody's patient's ECG, or the patient themselves, ever move? Anyone hear an alarm, for example, ventricular tachycardia, when the patient is in anything but ventricular tachycardia? Right. So one reason that I use heart rate on the A-line source is when I'm floating a transvenous pacer, and I want to know when I have mechanical capture, works quite nicely. Of course, the A-line will give you mean arterial pressure. And we have looked at mean arterial pressure relatively at length in ICU literature. And what we have found is that it's kind of a disappointing metric. I'm not really sure what map to target. We say 65, because it seems like polite ICU parlance, and everyone's like, map of 65? And it seems like a thing that we can all agree on that doesn't create consternation. One thing that I really like is DPDT. So DPDT is a change in pressure over a change in time. So if you look at the black curve here, you have a normal left ventricle pressure. And this is a scalar over time. And so the black curve shows you that the pressure goes up, and then it goes down. And with an increase in contractility, the change in pressure over change in time goes up. So it gets a little bit more steep if you look at the red line. And arterial DPDT actually performs very well when thinking about looking at patient contractility. It doesn't correlate very well with annoying stuff like stroke volume variation or SVR. It actually correlates quite nicely with a DPDT that you can get from an MR jet when you do a patient's echo. So what I really want to know is, is it really worth it to know a patient's DPDT continuously? I don't know the answer. I think the answer is no. I mean, I can't tell you that I will titrate a therapy from one minute to the next based on a patient's DPDT, or at least I don't think I should. But I think that if I start a patient on an inotrope, a very gross test would be if someone's DPDT starts to go up, I can gain some inference that the inotrope is doing what I want it to do. So of course, we also can get a little bit of insight of pulse pressure variation. This very much, terms and conditions apply. This has to apply to a very polite patient population. They have to be well-behaved. They may, on the ventilator, on a certain mode, and their heart rate has to be well-behaved. And this probably performs quite poorly in a lot of the patients that we are interested in who are, quite frankly, really sick. So the next step up is your central venous access. And when I say central venous access, this can apply to a large venous sheath introducer or a multi-lumen port, triple lumen or quad lumen, for example. Regardless of the cannula configuration, you can get two types of information. You can get either a, you can get a SVC gas or a SAT, or you can get a central venous pressure. So your SCV02 is SAT from the SVC, and it really has two chief determinants. And it represents a balance between the DO2, or the delivered amount of oxygen, and the amount of oxygen that a patient is consuming. So DO2 largely relies on cardiac output hemoglobin and SAO2. And for those following along at home, you'll ask about, what about the PAO2? So if you look at the coefficient of the PAO2 and the DO2 equation, it's like 0.003 or something. I don't know. Two zeroes. Two zeroes. Lots of zeroes, right? Yeah. OK. So is it worth mentioning? No. OK. Great. Moving on. So your DO2 is chiefly determined by all of these things. And we invoke lots of concerns about these things when we say we want to optimize DO2. That's why we give blood. That's why we start inotropes. And that's why we hook people up to absurd machines like ventilators. And then, of course, the DO2 represents the patient's metabolic demand. So for patients with septic shock or cardiogenic shock, they have really high metabolic demands. And the SAO2 measures quite nicely the balance between the two. So when you look at SAO2 versus cardiac output, you can see on the very, very steep part of the curve on the radiographic right side of the screen, very small errors in cardiac output measurements resulting in substantial errors in interpretation. So the TLDR version here is that the SAO2 is helpful when it's low. When it is high, notice how kind of lazy the slope is. So your cardiac output can be anywhere between, I don't know, 8 and 14 liters. And as far as I'm concerned, that's not really a helpful number by way of prediction. Additionally, if someone is critically ill, do you really care about discrimination between someone whose cardiac output is 10 and 14 liters? I don't. I'm happy when I have a cardiac index of 1.8. That's winning. So when you have both an arterial line and a central venous line, you can accomplish something called transpulmonary thermodilution, which is just a normal central line catheter in the IJ. You can also use the femoral and a femoral A-line. So transpulmonary thermodilution requires what are called thermistors. And a thermistor is this portmanteau of thermal and resistor. Basically these are two just fancy temperature sensors. And the goal of transpulmonary thermodilution is to manufacture some values on the screen that are supposed to give you some insight regarding cardiac output. So many, if not all, transpulmonary thermodilution use pulse contour analysis like Sid was talking about. And this gives you a real-time estimate of cardiac output. And so cardiac output remains valid in transpulmonary thermodilution in the scary cases that prohibit interpretation in other ways that you use cardiac output devices. So patients like atrial fibrillation, valvular disease, cardiomyopathies, you can rely on the cardiac output from transpulmonary thermodilution here. The problem with pulse contour analysis with transpulmonary thermodilution is that the normal ICU therapies that we do manipulate cardiac output. And so the longer that your patient is on this device without recalibration, the further away the cardiac output value will be from correct. So to calibrate these devices, you do what's called just a normal thermodilution. So this machine requires just a normal thermodilution that we have done since the 1970s, which is akin to the pulmonary artery catheter. And at each bolus injection, so you inject something into the venous line, which is usually cold. It may or may not have an indicator. What this does is reset the pulse contour analysis metric for what your cardiac output should be. So the classical thermodilution technique relies on the Stuart-Hamilton principle, which is really why calculus is integral. Yeah, that's fair. So the important part of the equation is on the bottom here, is namely that the area under the curve is inversely proportional to cardiac output. So when you see an area on the screen that's a very, very large curve, it takes a long time to come back to baseline, you can be sure that your cardiac output is probably really bad. So there are a couple of things on the transpulmonary thermodilution screen that I want to draw your attention to. I would say there are too many variables on this screen for me to pay attention to. I have the attention span of a gnat, and I see all of these numbers, and it's like watching a laser show. This is all very interesting. What do I do with all of this information, pray tell? And so there are really only two things that I want us to draw our attention to. The extravascular lung water and the global end-diastolic volume. So after the cold bolus through the venous line, you get a curve which is then log-transformed and then mathematically poked, prodded, and put through the black box to yield the global end-diastolic volume, which is really just the volume in the heart at the end of diastole. Is that a helpful number for anybody here? I agree completely. I have no idea what to do with this number. But the reason that we want to know where this number comes from is because it is subtracted from additional mathematical assumptions to calculate a value for extravascular lung water. And so for patients like, for example, with pulmonary edema, you expect for extravascular lung water to be much higher, suggesting perhaps a very clear fluid intolerance. So as you can imagine, the sources of error here are nearly infinite. As you can imagine, extravascular lung water will be completely underestimated by, for example, patients with pulmonary embolism, where the cold bolus doesn't quite reach that area of the lung, and therefore total amounts of lung volumes will be completely incorrect. So this leads us then to the OG way of monitoring patients. I guess OG is the wrong phrase here, because I guess the world existed before the 1970s. We use the pulmonary artery catheter in the CVICU in a very real and legally binding sense, and we're actually very excited about it. The way you float a pulmonary artery catheter hasn't changed very much over the past couple years. You get a relatively predictable waveform in the right atrium, and then you get into the RV, and you'll see a big jump in your systolic, then you pop into the PA, and then you'll see your diastolic step up. And then you wedge when it starts to look a little bit more like your CVP waveform. PA catheters actually are not a monolith. There are three generations of devices, and really the first generation was really where you just get your thermodilution signal. This was the original description, where you just see a quick up and then down, and then the second generation gives you a relatively continuous cardiac output monitor. And then the third generation integrates pulse wave analysis, or pulse contour analysis, to give you a little bit of inference regarding your cardiac output. All PA catheters can give you CVP, your PA pressures, and you can also draw samples from your proximal port in your RA and distal in your PA. You can add and subtract many features, which of course are available for cost. So this is from the ESCAPE trial. And so this was a trial that was meant to really help us figure out what to do with pulmonary artery catheters in people who are unwell. And what I noticed about these curves is that they look the same to me. When I saw this graph initially, I wasn't totally convinced there were two curves. So the authors I guess forgot to code for the second one, I'm not really sure. But what you see is that it didn't really save any lives. And the way that a lot of people interpreted this is that the swan is dead. That is it for the swan, it was the nail in the coffin, and this made a lot of people very upset. Because a lot of intensivists use pulmonary artery catheter to manage, to diagnose, and to make sure that you are not missing any harbingers of badness in the critically ill. So this has been looked at a couple of more times because we didn't like the results the first time. So in, the year was 2020, they looked at patients with cardiogenic shock. And across the really bad sky stages, which are the sicker patients with cardiogenic shock, patients managed with a pulmonary artery catheter had demonstrably better outcomes. And to make even more sure about this, there was a recent publication, which was quite nice this year, that looked at in-hospital mortality of patients managed with and without a pulmonary artery catheter. And what they found was a significant result, was that patients with pulmonary artery catheters probably did much better. So where exactly does all of this leave us? Your job at the bedside is to be the person to say what your patient needs to get better. And sometimes that's the device, and sometimes it's not. So use the device that answers your question. If you know that your patient has good cardiac output, I would argue you don't need a device that tells you your cardiac output. Your screen is a treasure trove of information, whether it is your arterial line that can give you a heart rate or a map or a DPDT, whether it's a pulmonary artery catheter where you can derive your pulmonary artery pulsatility index. Whatever it is, squeeze all information that you can out of the screen, because waveforms and numbers are a treasure trove of information. Avoiding really big lines may be really good. You don't want infections. You don't want procedural complications. But delaying diagnosis of a potentially devastating disease state can be really, really bad. And with that, I hope I did not induce general anesthesia here, and I want to thank everybody for their attention. Thank you.

minimally invasive cardiac output monitors

non-invasive cardiac output monitors

diagnosing shock

accuracy and preciseness

non-invasive methods

point-of-care ultrasound

invasive methods