Thank you very much and thanks Brandon for that excellent presentation. So now my job is to say, yes, Brandon, but no. So we'll see how that goes. Happy to be here. I have no financial disclosures, none of the physical examination companies have reached out to me ever, but if you know of any, let us know. But my job is to really talk about the physical exam and in order to make my point or present this side of the debate, I would like to first establish what are we disagreeing on. Let's talk about the disagreement first and maybe define some terms. Next I want to give you some arguments that I believe support my conclusions and I'm going to try to stay away from any of the data that Brandon has already shown very eloquently and try to show you some data that really looks at other aspects of hemodynamic monitoring related to the physical exam. And finally, what we want to always obviously reach is a conclusion and I will give you the conclusion that I have arrived after examining this subject and thinking a little bit about it. So what patients are we talking about? I think that when we talk about shock, obviously we all know that there's different types of shock, but I think that we heard from Brandon a lot of very, very pointed points regarding specific cases, a lot of them revolving around cardiogenic shock, which no doubt is a very important aspect of shock. But the reality is that if we're talking about shock in general, the vast majority of the patients that present to hospitals in shock are going to be distributive shock. And there is a subset, 16%, that probably would be cardiogenic shock, and that might be a specific population, but the patients that I'm going to try to make the point for are all comers, when we don't really know yet or we're monitoring for a large number of patients. The second, I think, point that we want to define is what constitutes physical exam and what is monitoring? I think that we would all agree that using our five senses at the bedside to get information is part of this physical exam. I think we would also agree that all sorts of devices, such as the pulmonary artery catheter and other devices that are more or less invasive that are attached to or inserted into our patients and give us readouts are monitors. Now what about in the middle? So if I were to add to my five senses a 200-year-old piece of technology and go to the bedside, I would think that all of you would still call that physical examination. What happens when I upgrade my technology and now I add something that's a little bit newer? And Brandon talked a lot about ultrasound, but when does that become physical exam? Now I'm going to be very generous with Brandon, and I am going to leave the echo outside, and that's probably a debate for another day, whether it's a monitor or a physical examination, but I'm really going to focus on the distinction of invasive monitoring versus true physical exam that all of us could do at the bedside. Finally, just to narrow that down, I also have to concede to Brandon that you could argue that basic vital signs, an arterial line could be a monitor, but we're going to lump those. I think a lot of us would agree that that's just part of the basic things that we have at the bedside. But what I'm really going to talk about is how do we utilize physical findings at the bedside to try to assess for perfusion in different key organs. The skin has become something that a lot of people have looked at lately, and there's a whole concept of obviously the microcirculation and the macrocirculation coherence and hemodynamic coherence when there's a disruption, especially in distributive shock, what happens and how we can utilize perhaps findings on the skin as a window to what's going on with our patients. And then obviously there's the basic labs, which I'm really not going to spend a lot of time in. And regardless of what you're using, I'm sure that our patients get more than enough labs that we need, and I'm not really going to go into that in much detail. So I think we could all agree, and I think Brandon would agree with me, that an ideal tool for shock should be available, which means that we have it in all our hospitals and we have it in hospitals around the world, should be accessible. So if it's locked behind the door and not everybody can use it at 3 o'clock in the morning, it might be available but not accessible. I think we all want it to be reliable, which means that not only it tells us the same thing over and over again, but it tells us something that is valuable. I think we want it to be safe, and obviously you could argue about safety if a diagnostic exam leads clinicians in the wrong direction. It could be unsafe. But what I'm talking about is the safety of applying that tool to the patient, right? And even within what was presented earlier, I think we would all agree that putting a probe with gel on a patient is exceedingly safer than sticking needles and trying to gain access. So they're both well done, well, probably meet the safety requirement that we would ask. More importantly, I think they should be associated with patient-important outcomes. I don't think Mrs. Johnson cares if she's fluid responsive. I think Mrs. Johnson cares if she's going to die. I think Mrs. Johnson cares if her kidney's going to stop working, right? So we focus a lot on surrogates, right, that are important, or we believe are important to us, but really are not patient-important, right? So again, we want to make sure that the tool we utilize is associated with patient-important outcomes. And finally, and obviously the holy grail, and the reason why we have this discussion is because we haven't found it, is what is the tool that clearly and inequivocally will improve patient outcomes when applied in our practice? So with that, my position is we don't need monitors. Physical examination and some blood tests are all you need for a broad range of patients with circulatory failure. The next part of my presentation is the arguments that I'll make in favor of the physical exam. So whether you are examining a patient, looking at a mottling of the skin, doing capillary refill, I think that everybody would agree that it's available, it's accessible, and it's safe. So at least those boxes I think are very easy to check, and probably we could move on. What about reliability? This becomes a little bit more complicated, and there's not a lot of great data on intra and inter-reliability on examination findings. Clearly they've demonstrated, and we'll talk about Andromeda shock and some of those more recent trials that have looked at physical examination, they've demonstrated that if you apply a very strict protocol and you protocolize how you do certain parts of the exam, you can improve both the intra and the inter-reliability, but that's not something that has been published enormously. Now, one of the big challenges we have also with a lot of our other monitoring devices, including ultrasound, is that we all agree that in the expert hands of somebody like Brandon, we believe that it's reliable. But as more and more people have access, is that intra and inter-reliability also a problem? So that is something that I think is hard to really elucidate finally. The next thing is that they have to have a correlation in terms of reliability with some parameters. Now, I said earlier that I'm more interested in patient important outcomes, but the reality is we always correlate anything new to either cardiac output or to a wedge pressure or to other findings that we have in the past, and we'd like to see that physical findings do correlate with some of the parameters that we talked about in the previous talk and that are important for clinicians. And ultimately, what I think is always our interest as clinicians is understanding the correlation of the tool we're applying to patient trajectory. Who's going to do well? Who's not going to do well? Who's responding to therapy? Who's not responding to therapy? So reliability, I think, is something that is difficult to fully establish, but I'll share with you just some interesting studies. And some of this literature obviously goes back some time. This is a study that I have always found fascinating just because I can't even imagine what an IRB would say if they tried to do it today, right? But Cohen and collaborators looked at the reliability of cuff pressures of arterial blood pressure and shock versus arterial line, and what they would do is they took a group of patients who were in shock on vasopressors, stabilized the blood pressure, and then turned off the vasopressors, waited for the blood pressure to drop, and then compared mean measurements with a cuff and mean measurements with an arterial line. And what they showed is that in shock states with high peripheral vascular resistance, there is significant discrepancy between the cuff and the arterial measurements. That seems to be a little bit less problematic in low PVR states, but it's just interesting that even there's very little literature on the basic measurements, right, of the vital sign that defines circulatory shock in these patients. In terms of our clinical skills and being reliable, Brandon did share with us, Dr. Wiley did share with us, some of the early data that suggested that perhaps PAPR and pulmonary recutterers were not associated with improved outcomes, but for many years, and for those of us who have crossed the line of 50 that Brandon has not reached yet, and we remember using PACs all the time and being trained with PACs, right, and all of a sudden the use of that in the general population has dramatically dropped, but this is back from that time, and you can see that in general ICU patients, clinicians' clinical assessment based on exam was not very good in predicting where the cardiac output would be high, normal, or decreased, or where the wedge pressure would be elevated, normal, or decreased. However, when you go into a subgroup of patients, like in the CCU population, which probably have a greater representation of cardiogenic shock, which is a lot of the things that we talked earlier were discussed around, that accuracy, let's call it, improves significantly. So obviously the physical exam does not correlate always with what we find with a monitor device such as the PAC. This is a very interesting study, and I think that we've talked a lot about Dr. Maxwell, and I think with the cafe, I had to include one of his slides, but this is, he was interested in looking at temperature, and this is a study he published over 50 years ago looking at the correlation of toe temperature with cardiac output, and obviously it doesn't seem like a bad correlation here. Now I go to a lot of ICUs as part of my job. I haven't seen a lot of people measuring the temperature in the toe, to be honest, but it's been studied, and I think a lot of people have taken this to other, in other directions with the idea that there might be some value in measuring temperature, and we'll talk a little bit about that. This is a much more recent study from HIFSTRA and collaborators where they looked at a host of clinical examination findings as they correlate with cardiac index measured by echocardiography, and what you can see here that they have four categories here, central circulation, markers or findings of cerebral and renal perfusion, skin perfusion, temperature, which a lot of this has really derived from the original studies that Maxwell did on toe temperature, and then skin perfusion looking at capillary refill time and motling, which I'll talk a little bit more about. And the bottom line is that some of these are more specific, less sensitive, but in general in different groups of patients, right, what you can find is that the accuracy hovers around 60%, so not super, and what he also found in this study is that of the large number of studies of interventions, physical examination interventions that they examined, only a small number actually had that correlation with cardiac index. We talked about whether an ideal tool would correlate not necessarily with indexes that we are interested in, but actually with some patient important outcomes or trajectory. This is more data from the toe temperature studies, and really what they showed is that the temperature does correlate with those who survive and die before they die, I guess. I mean, at one point, I mean, their temperature really is going to be a problem if you're dead, but as you get there, right, you can demonstrate a difference, but more importantly, and a lot of other groups have actually explored this further, is that gradients in temperature, either the toe to ambient or maybe a central to a peripheral temperature, can also be more predictive of which patients are going to do well and which will actually not survive the shock. A Motlin score has been something that has been published more recently about, and it's really looking at an objective scale of Motlin around the knees. This is from eight Alfella and collaborators, and what they showed here is that you can actually correlate Motlin scores to not only survival in a large number of patients with shock, but also to some other markers of perfusion, such as urinary output, arterial lactate, cardiac index, and SOFA scores or organ failure. So you would say that there is evidence and there are studies, and I'm not going to show you all of them, obviously, that do suggest that there are physical findings that do correlate with patients' trajectory. This just gives you a summary, it's from a more recent review article, of several different parameters, a lot of them looking at peripheral temperature, looking at capillary refill times, and looking at Motlin and some combinations, and how they correlate to organ failure or mortality, or even how they change and can be associated with the trajectory of patients who are responsive or not responsive to the interventions that we're doing. Again, suggesting that there is evidence supporting that physical findings do correlate with patients' trajectories. Finally, the real question is, does it improve patient outcomes? And that, perhaps, has been the key problem for all of us, that there's not a lot of good studies that have examined this question, and there's really no high-quality studies that have said that a hemodynamic tool, whether it's a device or a physical examination finding, is associated with better outcomes. But it also is an unresolved problem, because for years, the pulmonary decatheter has been the standard for hemodynamics, and obviously, it's very closely, I think, related to the growth of our specialty. And I think it probably does have a pendulum swinging, as Brandon suggested, that's important. But the reality is that every new device just shows that it can predict the same thing as a PAC. But nobody's really doing the studies that demonstrate that these are tools that actually improve patient outcomes. And that, obviously, is a bigger problem. So as we close, I do want to share with you, and you're always probably, a lot of you are aware of the Andromeda shock trial, where they actually tried to examine this in a randomized trial. Now, it wasn't the comparison that we're talking about today. It would have been nicer if it was peripheral perfusion versus PAC and ultrasound, but we don't have that study. But I still felt that there's some lessons to be learned here. And this was 424 patients randomized to either a peripheral perfusion, which was capillary refill time, that was standardized, basically, they would take a microscope slide, press the index finger for 10 seconds, and then with a timer, measure the refill time, versus lactate clearance, which is something that Jones, Nate Shapiro, Atreziak, and others showed back then was equivalent to other tools that we were using in terms of outcomes. And by lactate clearance, they meant that the lactate normalized or in subsequent checks, decreased by 20% or more. And this would guide a very protocolized approach to using vasopressors, fluid boluses with fluid challenges, et cetera. The primary outcome of this study was all-cause mortality at 28 days. The secondary outcomes included death within 90 days, SOFA scores at 72 hours, mechanical ventilation, vasopressor, renal replacement therapy, three days within 28 days, and other outcomes such as ICU and hospital length of stay. They defined septic shock as presumed or confirmed infection, plus a lactate greater or equal than 2 millimoles per liter, and the need for vasopressors to maintain an MAP at 65 or above. After a fluid load of at least 20 mLs per kg over 60 minutes, and really, the intervention period was like eight hours after enrollment, and they got these patients pretty early. Now, when you look at the primary outcomes, the mortality at 28 days for the peripheral perfusion targeted resuscitation was 34.9%. And the lactate level target resuscitation was 43.4%, and they basically missed their P-value by a little bit. It was 0.6, which raised the question, could this be a problem with this being underpowered, or truly, maybe it doesn't make any difference? We heard during this conference from the Clover Study Group, and it seemed that there, where you were within a ballpark of reason, it didn't make a difference how much fluid you gave in sepsis, so maybe there's something that we're missing here. But anyway, you can see that other outcomes are showed here. This is the Kaplan-Meier curve, and again, the peripheral perfusion was 34.9%, and lactate was 43.4%. I'm sure that if we had these numbers and the P-value was 0.05, people would say, oh, it's a positive study, right? So raises the question, again, about, are we using the right tools to figure out how to advance care at the bedside? That would be a debate, I guess, for a whole conference and another time, but we can talk about that offline. And finally, when you look at some other outcomes, they did show that the organ failure was significantly decreased by a SOFA score point of at least one by 72 hours, and that was one of their secondary outcomes that they had predefined. Looking at some of the subgroup analysis that was pre-specified, the group that here looks a bit different is the lower severity group. You can see that the one that's really different is the less than 10 SOFA score, and perhaps suggesting that for some patients who are going to circulatory failure, if they're responding and intervene early, you may really not need much more than the physical exam, but that is a point that I think is congruent with what Dr. Wiley shared, which when you see that ignition light coming on and you don't know what's going on, and things are not going as you expect, that might be the time to escalate and get more information. They also did a post hoc Bayesian reanalysis, which obviously is another method of trying to look at probabilities, something that more and more studies during the COVID period have used. And what's really interesting here is that when you look at different models, what they call the optimistic prior, neutral prior, or pessimistic prior, or also the null hypothesis prior, the probability that CRT is better than using lactate, for example, is significantly high. So when you look at here, is anything that's on the left of this line would be favoring the probability of capillary refill time being associated with improved mortality. And anything that's on the right of this, which is very small, you can see it here, depending on the model you use, would be the probability that it would be harmful or that lactate would be better. So interesting, obviously, it just raises more questions, questions that are going to be answered with Andromeda shock 2, which is currently enrolling, and maybe next year when we come back in Phoenix, we'll hear from the investigators what they found. Some additional considerations for the physical exam. Anything that brings us closer to the bedside, in my mind, is a positive. I think that we could all agree with that. I think there is something to be said about clinical intuition. When you're in a room and you're looking at patients, you are recognizing patterns that might be very valuable. We might not have the ability to articulate what we're recognizing, but there is value in being at the bedside and seeing how people are responding to treatment. The human connection, I think, is very important. I think the more time we spend away from devices and with our patients and their families, the better off we'll all be. But that is something that in our current practice becomes more and more difficult. There's also, obviously, advantages to filling some regulatory compliances that you actually re-examine the patient. I'm not going to go there. I'm sure I don't want to get in that hornet's nest. And finally, the Lindy effect. So the Lindy effect is something that comedians observed in New York many years ago, but that now has been mathematically expressed as the likelihood of an idea or a non-perishable to perdure in time is directly proportional for how long it's been in existence. So we've been talking about the physical exam since Hippocrates. That's a long time. We've been talking about the PAC for decades, and we've been talking about ultrasound for years. So just based on building the effect, I expect that in Congress of 2200, whatever it is, 2200, we will be talking about the physical exam still. So with that, I would like to close. We said that the ideal tool will be available, accessible, reliable, safe, associated with patient important outcomes, and lead to improved clinical outcomes. I do believe that there is clear evidence that the physical exam is available, accessible. We can argue about just reliability. I think we all agree that it's safe. It is associated with important patient outcomes. And there might be emerging studies that suggest that it might be associated with improved outcomes, more to come there. But what I would say is that we need monitors. The question is, which monitor? And that might be answered by what you have available, what is your expertise, and what patient are you dealing with? Hard for me to argue if some of the cases that Dr. Wiley presented that an echo would not be something that would be helpful. But maybe not for all patients. And that is something that also needs to be informed by future studies. What I would say is that the physical examination and some blood tests are always the starting point. And that we should be sure that we're checking that box off first, cuz we can do that anywhere. And if you don't believe me, these are two physicians that know a lot about life. And they told me once that you can't always get what you want, but if you examine your patient, you might just get what you need. Thank you very much.

physical examination

circulatory failure

accessible

safe

reliability

mortality