Recorded at the 6th McMaster International Review Course in Internal Medicine.
All lectures available at the MIRCIM virtual platform, click here to buy access.
Professor Yaseen Arabi: We have a list of great questions, actually reflecting the quality of the talks that were given in this session. So I think [in the] interest of time, I am going to get started on these questions. The first question is for Professor Jean-Louis on hypovolemia in patients with edema. In the talk you illustrated that we should not assume that a patient with edema could possibly be hypervolemic. The question is how [can we] differentiate hypovolemia in this setting from hypervolemia?
Professor Jean-Louis Vincent: Yes, that’s a very good question and a complex one. But as I indicated in my talk, the static measurements do not help very much – just measuring urine output, blood pressure, heart rate, ScvO2, lactate, or central venous pressure – all these variables taken in isolation will not help very much. So what we need to do is to have a dynamic test and that’s what we call the fluid challenge. We described it in the critical care, a few months ago, 4 months ago. It’s open-access, so everybody can look at it, but the idea in the acutely ill is to give a small amount of fluid quite quickly and see how the patient responds. And if the patient tolerates the fluid well and you see a reduction in heart rate, for instance, that would be a sign that the patient was indeed hypovolemic. If the heart rate does not go down, if the patient is still tachycardic – without any change in your fluid administration – it is unlikely that the tachycardia was due to hypovolemia. Now, the same is true for blood pressure in purely hypovolemic states. We all know that in trauma when you see some degree of hypotension, you increase the rate of infusion, you give a big bolus and you see [that] the blood pressure comes back. In bleeding patients, that works very well. If there is pure hypovolemia, like in severe diarrhea, for instance, that will work, too. However, in septic patients who are vasodilated, the blood pressure may not go up and yet the cardiac output may increase. That’s [why] in septic states, it is always better to have a measurement of cardiac output, or at least stroke volume. And that’s how we could use echocardiography to monitor the VTI during the administration of a bolus of fluid. That’s an easy way to go if we do not have a more sophisticated cardiac output measurement, like PiCCO or even the Swan-Ganz catheter. So, that’s the principle. By looking at the patient, you cannot assume that the patient is hypovolemic unless the history is clear. Of course, if the patient comes in the emergency department [and] has hypotension, tachycardia, and recent diarrhea – OK, you can assume that the patient is hypovolemic. And the patient may have edema associated with hypo-oncocity, low albumin level, or significant alterations in permeability related to sepsis, for instance. So it’s a very complex situation; there is no easy answer to that question and we usually need to combine different variables [in order] to decide that hypovolemia is likely and therefore we will do fluid challenge test.
Yaseen Arabi: Thank you. Maybe a follow-up question for Dr. Rochwerg, related to fluid also. You illustrated nicely the current state of evidence regarding different types of crystalloids. So the question from the audience was, “What [are] the current practice guidelines regarding crystalloids and especially balanced fluids?”
Professor Bram Rochwerg: Yes. So the last surviving sepsis recommendations didn’t differentiate between the crystalloids. You know, that there are new guidelines coming out soon. I would caution to say that, as I presented, although perhaps the needle is moving towards the fact that balanced crystalloids have benefits, I think that equipoise does still persist. And there is still uncertainty, especially when we look at larger volumes of fluid resuscitation and in the most sick of ICU patients, whether there is benefit or harm in using one approach or the other. And thankfully, there are a number of randomized controlled trials, as I mentioned, that are going to be released soon and will give us further direction in terms of optimal crystalloid. I think for now, I presume the majority of people have been steering towards using balanced crystalloid, but as I mentioned, I don’t think we’re yet in a position where we can say [that] using saline is the wrong thing to do. And perhaps maybe even a mixture of two – and I know Jean-Louis feels strongly about, you know, a policy of one fluid is probably unrealistic – and maybe there is an opportunity to mix based on physiology and to individualize. So I think for the time being, perhaps an approach in keeping with that is reasonable, but as I mentioned, knowing that BaSICS will be published soon, knowing that PLUS is continuing to enroll in Australia and New Zealand, and I’m leading a fluid trial specifically in patients with septic shock in Canadian and international centers, I think we’re going to have a lot more data addressing optimal crystalloid available to us in the near future.
Jean-Louis Vincent: If I may, I would like to say that we have enough data, we just know what’s in the solution, we just know the composition, we just know the content of each of these solutions. So we know what we give to the patient; we don’t need a prospective randomized controlled trial to see what it does when you give sodium and chloride and perhaps lactate and perhaps some other substances to the patient. When we need to know is that saline solutions with 1.54 milliequivalents per liter of chloride will increase the chloride levels in the blood, and so they may result in hyperchloremia if you give too much. And we know that that’s bad for the kidneys – it has been known for many, many years. So we don’t need a prospective randomized controlled trial to know that hyperchloremia is bad. We know hyperchloremia is bad, hypernatremia is bad, hyperkalemia is bad, and hypermagnesemia is bad. Hyper-something is always bad. So why would it be an exception for hyperchloremia when we know it’s bad for the kidneys? So if you do prospective randomized controlled trial, if you give a lot of saline, you will show harm – of course. Unless you measure chloride levels and when they go up, you decide that you should not give more saline. That’s what I do, that’s what we do every day in our department, exactly as we do for the potassium. So why would it be an exception? Now, when we go for balanced fluid, we must remember that the Ringer’s lactate solutions are a bit hypotonic. In Europe we call Ringer’s lactate solution Hartmann’s solution because Hartmann developed it. Hartmann was a pediatrician treating dehydrated children with diarrhea; so it is somewhat hypotonic. If you give it to a patient with major brain damage, you are doing what I will call malpractice. It is just unacceptable to give a Ringer’s lactate solution to a patient with some degree of brain edema. But people don’t say that. They would like to see a prospective randomized… We know that hypotonic solutions can be bad for the brain. So you could use PLASMA-LYTE, but you must know that in the PLASMA-LYTE solutions there are some electrolytes which may not be so good for the body, including acetate, which is a bit vasodilating. And gluconate, we don’t know really the fate of gluconate. So we must be a little bit more careful with PLASMA-LYTE, although once again – when we give a limited amount, it cannot harm. You may know this famous sentence from one of our colleagues from Australia who was saying that you could give cats’ piss via the veins. If you give a small amount, it will not hurt. But if you give a lot of cat’s piss, that could be indeed very harmful. So, if you give a little bit of a fluid, in a prospective randomized controlled trial you will not find any difference. If you give too much of any fluid, you will find harm. So you don’t need to do the trial because we already know the results before we start. It’s basic physiology but people tend to forget basic physiology.
Yaseen Arabi: Fantastic. I think this is a great discussion reflecting the debate in this area. I have to say that we have a list of great questions and I hate to miss some of them, so if we can try to keep some of the responses brief, if possible, so we can try to catch up on the questions. Next one is another interesting area about qSOFA for Dr. Alhazzani. And the question is: “You mentioned qSOFA. Could you remind us what it includes? How objective and how reproducible is qSOFA?” Waleed?
Professor Waleed Alhazzani: OK. So Quick SOFA has 4 components. These components include, 3 components that include the GCS [Glasgow Coma Scale score] of the patient, the blood pressure… [problem with the connection]
Jean-Louis Vincent: …the respiratory rate. Actually, during our consensus, I was part of that consent to our friend. Is he there? Hello?
Yaseen Arabi: We lost the sound. So if you want to comment on this, Jean-Louis. Briefly, what’s the component?
Jean-Louis Vincent: Yes. So of course, sepsis is infection plus some degree of organ dysfunction. And I wanted to keep it very simple and say, “OK, any clinician knows what some degree of organ dysfunction means.” But the others said we need to have objective criteria. I accept that. So that’s how we decided that at least for a scientific publication, it should be an increase in SOFA score. OK, but then some people said, you know, it may be a little bit too complicated and that’s how the qSOFA was developed. Waleed is back, so maybe he can continue. So the 3 items, as you say, the blood pressure, the respiratory rate, and the Glasgow Coma Score investigating the degree of confusion of the patient. Go ahead.
Waleed Alhazzani: OK, sorry guys. My laptop’s GCS dropped once we mentioned GCS. Thank you Prof. Jean-Louis for highlighting the components. I think the main issue with Quick SOFA are 2 things. First, some people use it as a screening tool for sepsis, while in my opinion it should not be used as a screening tool. It should rather be used as a predictor. So there is an association between, you know, patients meeting the qSOFA and having worst outcome. Obviously, once your GCS or your blood pressure drops or you become tachypneic, in the context of sepsis, you know, you probably have a serious condition. So it’s pretty specific or more specific than, you know, the source criteria in terms of predicting bad outcomes in patients with sepsis, but I think using it as a screening tool, you will end up missing some cases, especially early-on cases. So it has to be applied within the same context that the Sepsis-3 definition has proposed it and not use it for, you know… “OK, does the patient meet any of those criteria? If not, my patient does not have sepsis, I’m OK.” You know, so I think this is the main area of confusion with clinicians. I personally, when it comes to looking at patients, I use Gestalt or clinical decision-making more often than using any predictive scores. I think predictive scores are also helpful for standardizing inclusion criteria in randomized trials or research [making it] easier to present to people. Having said that, there isn’t a perfect tool yet to screen patients for sepsis or even, you know, to make the diagnosis early on. Once we have the field advance in this specific area, I think more work is needed because early recognition of sepsis is absolutely important and, you know, so far we have the clinical judgement, maybe some biochemical support, but I think we are far behind in the field and there is more work to be done in this area.
Yaseen Arabi: Thank you, Waleed.
Jean-Louis Vincent: What we want to identify is the patient who needs immediate attention. And the qSOFA is not specific for sepsis, of course. Think of the patient with severe heart failure and cardiogenic shock. Of course, the patient will be a bit confused, will be a bit hypotensive, and will have hyperpnea to compensate for the lactic acidosis, and perhaps hypoxemia associated with acute lung edema. So qSOFA is of course not specific at all for sepsis; it’s more specific for shock, yes. The trauma patient, a patient with pulmonary embolism, when it’s severe, will have a positive qSOFA. So it comes down to the alarm systems that we have developed for use on the regular floors, the MEWS [Modified Early Warning Score], for instance, that the English colleagues developed – which is very good. So you know, you need to identify the patient who is in trouble and it could be the heart rate as well. The heart rate is not in the qSOFA, but it’s important. And if you can add the lactate, add the lactate. So I’m not a big fan of qSOFA. I don’t think it’s specific for sepsis and I think what you need to do is to identify the patient who is in trouble and needs immediate attention, whether it is sepsis or something else. We need to intervene quickly for those patients who are very sick.
Yaseen Arabi: Thank you very much. The next question is related to my talk, actually, about the anticoagulation in COVID-19 – another very complex question. And the question is: “More than half of the patients end up getting anticoagulated. What’s the explanation that studies comparing full anticoagulation with prophylactic doses have knock-on benefit?” And whether we can speculate on this question. So, I think there is significant interplay between the coagulation and inflammation in COVID-19. And the trial – the multi-platform trial that we only have in pre-print so far – showed that full anticoagulation in ICU patients was not associated with lower mortality in ICU patients, but in hospitalized patients it did, actually, reduce mortality. But this is still not peer-reviewed or published, [though] it’s [been] released. I think there is a lot more to learn, so it’s hard to speculate at present. It might be that actually the coagulation part [would be] better controlled with anti-inflammation. So maybe it’s like a vasculitis kind of thing, where you treat the inflammatory part and the coagulation improves, but I think [there is] still a lot to learn about this before we speculate. And also we don’t know the right dose. The INSPIRATION trial of intermediate dose versus prophylactic dose has not shown a difference in outcome, but there are other trials ongoing, so maybe we should just wait for this. The next question, I think, is for all panelists; maybe I will start with Roman Jaeschke, who is filling in for Deborah Cook. Very interesting and very relevant question about withdrawal of life support, and the question is: “Active withdrawal of life support could be considered not acceptable in certain cultures or legal environments… Do we have observations from different countries, especially outside North America?” So there is clearly a lot of variation of withdrawal versus withholding life support. So maybe, Roman, you want to reflect on this?
Professor Roman Jaeschke: Well, I want to apologize [on behalf] of Dr. Deborah Cook. I just have spoken to her and she is truly attending a major family emergency, but she has provided me with answers to questions directed to her. The short answer is, yes, we behave differently in different parts of the world. And we had a great example during the last hour, in which we learned that we treat remdesivir, ivermectin, convalescent plasma, and tocilizumab in completely different ways, not to mention dextran and albumin. And we behave differently in [such] important things as end-of-life care. There are clearly jurisdictions in which withdrawal rather than withholding would not be considered an option. The world is changing around it. I think part of the change is related to communicating with each other and recognizing that what we consider an absolute standard of behavior may not be the standard of behavior somewhere else. And it gives us some pause and the ability to reflect on what we are doing. Deborah added to it, that what we have to do is to be extremely sensitive to cultural differences and religious differences, which influence legal differences. She commented that quite a lot of the wishes are actually faith-related. They may [range] from bringing some artifacts to having the geographic location of the bed [face] towards the east or towards the west. And that wherever we do it, especially in societies which are mixed, it’s a very sensitive topic which has to be recognized and addressed. So the short answer is, yes, there are such jurisdictions. The world is changing around us. It’s important that we know how we behave somewhere else.
Yaseen Arabi: Dr. Vincent, do you want to comment on this? Jean-Louis Vincent: Oh definitely, because the standard should be that we should withdraw life support in patients who do not benefit from it. And that should be accepted all over the world. And if you are a very religious person, I’m sorry but there is no God who ever said that humans should [be opposed] to an inescapable, unavoidable death. So that’s what we tell our Muslim community in Brussels and that’s how they can understand and accept that we withdraw life support in some cases. Because it is part of modern medicine, it’s because we have invented the respirators, you know, we put patients on respirators, breathing instead of themselves. So at some point, if it doesn’t help, we must stop it. There is no way around it. And if we could not do it, there would be 2 consequences. The first one is that we would sometimes withhold therapies in patients who may perhaps still benefit from it, but we would be concerned about the fact that more than 9 out of 10 patients would not benefit if we start mechanical ventilation in very extreme conditions. So we would not dare to start it; we would not offer the best treatment to the [1] patient out of 10 who could benefit from it. And the second element is that our ICUs would be very quickly occupied by patients who do not benefit, who should not be there. They are just in the ICU, they occupy a bed, they occupy our personnel, they take resources, it costs a lot of money, etc., for a treatment which is useless. In English they would [call it] “futile therapy.” In French we say “l'acharnement thérapeutique” – therapeutic stubbornness. And this is prohibited in Belgium and rightly so. And in Belgium, we have published it in Journal of Critical Care, if you’d like to look at it. We openly say it that in Belgium we sometimes accompany the patients at the end of their lives by increasing the doses of sedatives. This is not euthanasia, no, no; it’s not requested by the patient. The patient is no longer capable of asking for it. It is to accompany the patient to die peacefully with the family around. And of course, the family is well-aware of the fact that we increase the dose of sedatives and very often they ask themselves, “can’t you increase it a bit faster because there is no point here, the patient is comatose now, the patient cannot even recognize us. What are you doing? Doctor, please, try to bring it to an end.” So we need to cope with the modern medicine and allow people to die without suffering and with dignity. Roman Jaeschke: Talking to Deborah, she mentioned – and I’ve actually mentioned –situations where this sensitivity is crucial, and one of them was actually Israel and the other one was my own country, Poland. It sounds like monotheistic religions may have something to do with concepts of dying. But that’s a very sensitive topic by itself, so I will leave it at the moment. Jean-Louis Vincent: No, no. I mean, it’s entirely compatible with what I said. Poland is very Catholic indeed. And in the past, in Belgium too, we had difficulties with very religious persons – primarily Catholics in Belgium – and we had to tell them that God has never requested to continue what you may call “futile therapy,” even though the term may not be the best one. And now, the priests are with us. They say, “yes, what is the point?” You know? So people now really understand it and we speak about it on TV and the media speak about it. And people really accept it. When you speak about Israel, that’s a very complex situation, and as you know, as withdrawing is not allowed – withholding is allowed of course – but withdrawing is not allowed. They have a law, as you probably know, Roman, allowing people to put timers on respirators. So you decide in advance that the respirator will work for 3 days, for instance. And after 3 days it will stop by itself without any intervention from humans. It’s not a rumor, it’s a law; you can check it. It was published in The Lancet, you know, and the Israeli leaders will recognize that they have that law to just go around the very religious oppositions to what we have to do when the treatment doesn’t make sense any longer. Yaseen Arabi: Thank you. Waleed, do you want to comment on this one? Waleed Alhazzani: I don’t think I have much to add. I just would say that I agree with what has been mentioned, that you know, what is ideal sometimes is very difficult to apply in practice and changes have to come from the law itself to support physicians and families in decision-making. As long as the law allows, you know, family members to make decisions, even extreme ones, then what we can do as physicians to advocate for our patients is sometimes limited. Yaseen Arabi: Thank you. Bram?Bram Rochwerg: I think it’s easy in the situations that Jean-Louis discusses, where it’s very objective, you know, what is futile care and what’s not. It’s much easier. Because I agree. I think that there need to be processes in place where we cannot persist in these situations and the legality, as Waleed says, needs to support physicians in making those decisions. The challenging part, obviously is that it’s often not black and white, but grey and subjective, in terms of what is quality of life. Is this the quality of life that they would be happy with? And the family says “yes,” you know, and the physician team says “no.” And how do you align that interaction and assessment? You know, you want to do what’s best for the patient and not use futile care and resources, but also not project your own values and preferences onto the patient and the family in terms of what is the quality of life they would tolerate. I don’t know. I have young children and if I was… if I could interact but was on a ventilator, would that be a quality of life that I would be happy with? It might be, but other people might look at that and say “no.” So I think the challenging part gets into these more subjective situations, where it’s not clear and then there is a lot more uncertainty about what to do.
Yaseen Arabi: Certainly a very complicated topic. So religion probably is part of it, but I think also culture. There was an interactive survey in New England few years ago where people were asked about the scenario [of] terminal patients with neurologic problems, whether they would withhold or withdraw. And it’s very amazing. So, North America, Europe, and Australia – predominantly people [would] withdraw. But China, India, Japan, [the] Middle East – more people [would withhold] life support. Many of these countries do not share the same religion, obviously. There are a lot of things related to our culture, that are Eastern maybe versus Western. But it’s a complicated topic, and the decision that makes us go for this or that are an amazing area for discussion. Perhaps we can move to the next question, again to resuscitation, to Jean-Louis, about fluid challenge. You suggested fluid challenge… 200 ml in 10 minutes with the measurement of global blood flow response. How do you do it? How do you measure global blood flow response?
Jean-Louis Vincent: As I indicated, if it’s purely hypovolemia, the blood pressure may suffice. If there is tachycardia that may be due to hypovolemia, just look at the heart rate. But in most cases, and especially in septic shock, you need to have a measurement of flow. So what you need to do is to use your favorite technique to measure stroke volume. Now, some techniques are not very reliable. Let’s say, if you go bioimpedance, bioreactance, you know, end-tidal CO2 – that’s not very reliable. But you can use a PiCCO or a LiDCO to measure cardiac output and in complex cases, you may even use the Swan-Ganz catheter, but we use less and less the Swan-Ganz catheters, of course. You could use echo, just looking at VTI. You look at Velocity Time Integral before you give the fluid bolus and at the end of the fluid bolus. And you give the fluid bolus over a relatively short period of time, not too short. In the operating room, it could be in 1 minute. Because things move so much in the operating room, we have no stable period. But in the ICU, you could often afford to ask people not to touch the patient for 5 to 10 minutes. Nobody touches the patient. No physiotherapy, no movement, no sectioning of the tracker. All things remain constant. And so you can see whether this fluid administration is associated with an increase in cardiac output or not. And if you see some increase in cardiac output after, let’s say, 200 cc of fluid, you repeat it and you re-assess the patient. And if there is a positive response, you repeat it and you re-assess. But the fluid challenge in its concept does not encourage you to give more fluid. It rather encourages you to stop giving fluids when the patient does not benefit. How often do we hear in some ICUs people – sometimes over the phone – [say,] “well, try another 500 cc of fluid and we will see”? What will we see? “Well, we will see.” Come on, 500 cc – that’s a lot, and giving it like this? And then you don’t know whether the patient benefits or not. That’s not good medicine. We need to really appreciate the effects of what we do. If you start some dobutamine, you want to make sure that the cardiac output increases. If you increase the FIO2, you want to make sure that the oxygenation improved. That’s what we do all the time. Why would it be different for fluids and for other interventions? So if the patient does not benefit from what you [are doing], stop it. If you give dobutamine [and] the patient doesn’t seem to improve, you don’t see that the blood flow improves – stop the dobutamine administration. And you know, that’s what we do regularly at the bedside in the Intensive Care Unit. And that’s what we should do. So, let’s evaluate the effects of fluid administration. And to do that, the fluid administration should be pretty fast. So we need to give a relatively small amount, but over a limited period of time. If you give 200 cc in 20 minutes, you may not see any hemodynamic changes, or some other things will happen around the patient. And then you will not know whether there was a positive response to your fluid administration.
Yaseen Arabi: Thank you very much. Dr. Rochwerg, a question about the albumin, [which reads] that one of the talks mentioned that in patients with dengue fever stressed the use of colloids in the resuscitation of patients with sepsis syndrome. So the question is, “are there certain conditions or indications where you prefer to use colloids?”
Bram Rochwerg: Yes. Thankfully, practicing in Canada, North America, I don’t treat a lot of dengue, so I don’t speak from a position of expert perspective here. I am aware that the sepsis and the SIRS response [systemic inflammatory response syndrome] to dengue is relatively different compared to distributive shock from sepsis, with a loss of oncotic pressure and leading to hypotension. And there have been differences shown in some of the small dengue fluid trials in terms of the optimal approach. So it’s possible that the lessons that we have [learned] in sepsis in general are slightly different. That being said, I think that the fundamentals of fluid replacement probably should stay the same. And you know, the data that we have suggest, as I mentioned, that starches are harmful and are associated with increased acute kidney injury and mortality. And I heard from the previous talks that those that treat dengue might still use these agents specific to it, but I would have a hard time administering dextrans and other colloid starches to this population, given the totality of the data that we have for general critical care populations. Now, the role for colloids versus crystalloids, you know, I still think that this is an ongoing debate. And the role for albumin… unfortunately a lot of the studies that have used albumin, as we discussed in the talks, sort of looked at an approach of albumin-only versus a crystalloid-only approach, which is not how most of us use colloids. If we do use colloids, we rather give some crystalloid if patients are found to be fluid-responsive based on the parameters that Jean-Louis has discussed. Perhaps then we mix in a little bit of colloid when we see the serum albumin drop, thinking that there is a benefit to drawing fluid into the intravascular space. So I think, you know, whether an approach that does make sense on colloid or whether an approach that uses a predominantly crystalloid approach, I think either is reasonable and I think, like I say, acknowledging that I don’t treat a lot of patients with dengue, I would use a similar approach in those. I don’t know if anybody else on the panel has more experience treating patients with dengue?
Jean-Louis Vincent: Well, I have a limited experience of course; there are not many in Belgium, but through a lot of international contacts, you know… in dengue sometimes you need to give a very large amount of fluids. And indeed, if you give only crystalloids, you may end up with tremendous edema and perhaps what we could call “fluid overload,” although I hate to use the term because I’m not sure we know exactly what we mean when we speak about fluid overload. I wrote an editorial with Michael Pinsky in Critical Care on this: we should avoid the terms “fluid overload.” But… so, albumin, as you say, Bram, would be the colloid of reference, of course, yes. But we know that in some countries it’s considered very costly and not widely available. So yes, I think dextrans are somewhat outdated. But hydroxy-ethyl starch [HES] – I’m not so sure they are so bad, you know? The dispute has never taken place in the US because in the US, hydroxy-ethyl starch solutions are virtually as costly as albumin, so why would you give an HES solution in the US? If I still worked in the US, I would not use HES solutions over there because albumin solutions are about the same price. But in many places in the world, HES solutions are much cheaper than albumin solutions. Personally I’m not so convinced that HES solutions are so nephrotoxic. May I remind you that in the big CHEST trial from Australia, published in the New England Journal of Medicine, the authors ultimately concluded that HES may harm the kidneys, but according to the RIFLE criteria, there was an improvement in kidney function. And the nephrologists [who] work in the ICU love the RIFLE criteria, or now the KDIGO criteria. So, but there they change the rules and say “no, no, no, we have no confidence in the RIFLE criteria, so we will say that HES solutions are harmful for the kidneys.” And there was, as you know, a German champion who really was very excited about this and was saying that we were killing patients with HES solutions and ultimately, we no longer use them. In many countries we use albumin, and OK, we can afford to give albumin, but in some countries in the world, you know, not much, but some HES solutions, in my opinion, could still be allowed. But giving some colloids could make sense when the patient requires a very large amount of fluid, because edema is never good and we know that colloid solutions decrease the edema formation, decrease the need for intravenous fluid. This has been very well shown and this is, again, basic physiology. So, adding some colloids when we have to give large amounts of fluid – and you alluded to that Bram, I don’t think you are against the idea – I think that would be reasonable.
Yaseen Arabi: Thank you very much. A question for Dr. Alhazzani about the choice of antibiotics in patients with sepsis. It’s a very big question, but maybe some very general concepts. How do you choose antibiotics in septic patients without detecting the source of infection?
Waleed Alhazzani: OK, so I assume the question means [that] you don’t know the source of infection. Ideally, when you make the diagnosis of sepsis or septic shock, there has to be at least, you know, a presumed source of infection most of the time. That does not mean you know what bug you’re dealing with. But I believe that antibiotic choice depends on a few things: how sick the patient is, the microbiologic profile of organisms [at] your institution or [in] your country, because that’s hugely different. When I practiced briefly overseas, in some hospitals we had a 70% prevalence of multi-drug-resistant organisms. It was completely different, for instance; if I applied what I learned in Canada, for example, I would be, you know, completely malpracticing medicine there. So absolutely, you have to take into consideration the microbiologic sensitivity or profile of bacteria. And also, the source of infection. So there isn’t a straightforward answer. I can’t tell you, you know, “when you have a patient where you don’t know the focus of infection, just start them on antibiotic X or antibiotic Y.” It’s just not going to happen. You have to use your clinical judgement and combine these 3 parameters together, and then make the best decision possible. Some people use an approach where they say, “OK, you know what? I don’t want to think, I just want to throw in big guns,” I start with, you know, carbetamide and, you know, vancomycin and I just go to bed. Maybe antifungal sometimes. You know, that’s not medicine – that’s dangerous; doing this routinely in a persistent way has consequences. Antibiotics have side effects to patients. They can increase the resistance of organisms down the road. So you know, unfortunately sometimes you have to squeeze your brain and clinical judgement and combine these 3 parameters together and make the best decision possible. And always re-evaluate your patients. Hopefully in the future, we may hopefully in the next few years, have more rapid diagnostics in terms of identifying organisms with their resistance profile as well. So once this breakthrough happens, then you know, practice of medicine will be much easier.
Yaseen Arabi: Thank you, Waleed. Yes, great, I think this really summarizes the main concepts and I fully agree. A question for Roman regarding the Three Wishes program. This is an amazing program for supporting patients at end of life. The question from one of the attendees was, “Which of the wishes were most often forgotten?”
Roman Jaeschke: Well, it’s interesting. Obviously, I asked this question of Dr. Cook. And her answer was [that] the most commonly forgotten wish is to ask people what they wish. Essentially, if she were to choose one answer to it, she would say, “please ask the person, please ask their family or the patient, what is most important to them at that stage.” And I asked her then, “so what are the things?” And she gave me a list of things which to some degree surprised them at the beginning of the process. So, people were asking to bring items from home, to bring favorite pieces of clothing, to bring a favorite blanket, to bring a favorite photograph, a picture, to bring some food, to have some favorite drink – and those drinks ranged anywhere from orange juice to pure vodka – to bring a blanket to cover their bed, to make it feel a little bit more like at home, to have a favorite piece of music played. So a lot of those things which we take for granted, which we don’t think [of] unless we ask. So the most important thing is to ask what’s important to them, what’s important to their family. That’s how she answered.
Yaseen Arabi: Thank you very much, Roman. A question for Prof. Vincent about passive leg raising test. The question is: Should we begin this test within a semi-recumbent position? How long should it be kept? What [are the] criteria for positivity?
Jean-Louis Vincent: Yes. So physiologically, it’s very good: it’s the internal fluid challenge. So, you take blood from the lower half of the body – from the legs and the abdomen – into the central compartment, into the chest, and you see if there is an increase in stroke volume and cardiac output. So it’s exactly the same as the fluid challenge, but without giving fluids, using the fluid which is already in the body. And the concept of course makes sense, physiologically. The problem is that the effects will be very transient and you may actually miss the effect of the change in position if you do not use the right tool properly. When I make rounds in various units around the globe, and when someone speaks about the passive leg raising, I love to tell them, “OK, show me how do you do it.” They look at the bed and then they look at the blood pressure and the heart rate. Wait a minute, that’s totally wrong. Wrong, wrong, wrong. And of course, if the patient is not sedated – and today we try to avoid sedation – then the patient is, of course, under stress and there is an increase in heart rate and in blood pressure. Of course, it’s expected. So you need to do it in a very smooth way, in a quiet environment, and you need to have the tool to measure a very transient change in stroke volume. You could use echo, OK, if you are very good at echography, because [you don’t want to] miss it, and otherwise you could perhaps use a PiCCO, if you have a PiCCO, but be careful – it will be very transient. So personally, I am not sure I do it very, very well. And I don’t know many people who do it very well. Yes, I know the champions of this, Xavier Monnet and Jean-Louis Teboul in Paris. And so I have seen them doing it in their department. It’s not that easy. I can tell you, it’s not an easy procedure. And if you like, you can go to [the] website of ISICEM and you will hear an e-chat – it just takes 10 minutes – asking Xavier Monnet how to do it precisely. And he explains it very, very well. It’s free, it’s you know, you just go there and you have access to is. So I think he could answer the question much better than I could. I saw the question passing: “What is PiCCO?” PiCCO is actually a commercial name. It’s semi-invasive cardiac output measurement, where you insert a modified artery line and you use a central venous catheter and it’s transpulmonary thermodilution, allowing you to have a relatively precise cardiac output measurement. But of course, if you use it routinely in your department, you can use it. If you use it exceptionally, I’m afraid you may not use it very well. It’s always the same thing. When you do things very rarely, you may not do it very well; you need to do it quite regularly to do it well. So, the passive leg raising test makes sense, it may work, but you need to know how to do it and it’s not so easy.
Yaseen Arabi: Thank you very much. A question for Dr. Waleed about sepsis and phenotypes. And the question reads: “Could you explain phenotypes and the related mortality?” Maybe I will slightly add to this: You summarized nicely some of the recent studies on phenotypes. How much of this has clinical implication on my practice at the bedside, at present at least? Maybe in the future it will be different.
Waleed Alhazzani: Yes. I would like to compliment colleagues from Pittsburgh for doing this because I think, you know, we need to recognize that sepsis in not one size that fits all. Sepsis to me at least is like cancer 20 years ago. You know? You say “cancer” and you refer to all malignancies as cancer. So sepsis could be different even between organisms, between, you know, bacteria, viruses, etc. So it’s not surprising, because we already know that some patients behave differently than others, whether it’s related to genetic factors, to immune factors, [or] to the infection itself. So it’s not surprising to know that there are phenotypes; it’s just that nobody has, you know, attempted to do that on a larger scale. And I think the value of this JAMA paper is, you know, two-fold. A: they use artificial intelligence to help them with, you know, finding patterns. And B: they were able to group certain clinical and biochemical profiles together that could be validated in future studies, could help advance the field forward and make us understand better. You know, a simple example – patients who respond to steroids. Some responds to steroids, some don’t. You know, is it because there are certain phenotypes that respond differently to immunomodulators? Are there any specific patterns? This is I think the fascinating part about sepsis and I think, as I mentioned, when we focus on diagnostics and understanding the different phenotypes, I think we will treat our patients better. So, I really think it’s an interesting paper. As for implications in clinical practice, it did not you know, affect my clinical practice yet because I don’t know what to do with this information, but I think it definitely made me think, you know, how to think about sepsis in a different way and hopefully help advance the field in future research.
Yaseen Arabi: Absolutely agree. Thank you very much, Waleed. A question for Roman. If you could expand on dying at home versus in the hospital and the Three Wishes project.
Roman Jaeschke: Well, this project generally centers on the question at the end of life: Should we communicate with a dying person, or other very close people who are equally or even sometimes more affected, about the wishes which they may have at that time? And the answer to this from Dr. Cook was “absolutely”; it could be applied at home, it could be applied in hospice, it could be applied in hospital. It should be applied at home, it should be applied in hospice, and it should be applied in hospital by whoever can influence the process and make it more humane and more… I wanted to say “pleasant” which is no way, but somehow more tolerable and more memorable in a positive sense. So the short answer is “absolutely yes.” It may be the family doctor who is involved, it may be a social worker who is involved, it may be a nurse who is involved, it may be a palliative care doctor who is involved. It may be a family member who is involved.
Yaseen Arabi: Thank you. I think we can have a couple more questions before the end. A question for Prof. Vincent about how other conditions and diseases such as hypertension, heart failure, renal disease, etc. affect the diagnosis of hypovolemia? I will add to this [that] sometimes we’ve seen patients come [who have been] seen by a medicine resident or etc. who have some heart problem, but they end up very much under-resuscitated because they are so worried about fluids. What’s your take on this one?
Jean-Louis Vincent: Yes, it’s very difficult to guess the volume status based on the history of the patient. Sometimes – and that refers to my talk on oliguria – sometimes people may be afraid to give fluids because the patient may go into renal failure, but if you miss the opportunity to correct some degree of kidney hypoperfusion the patient will definitely go into renal failure. And then you will definitely have to withhold fluids or rapidly go for renal replacement therapy to eliminate the excess water. So let’s be careful – and we should not guess too much in medicine, especially in critical care medicine, where we can monitor our patients. So we are there to evaluate what the patient may need and try and see how the patient responds. That’s why I became interested in critical care medicine myself; I was an internist because I like to look at the patient as a whole, but in internal medicine when you prescribe something, you need to wait for one week to see some effect, whereas in the ICU, you turn a button and after 5 minutes, 5, you can already see an effect. It’s not always like this, but very often at least in hemodynamics, in respiratory failure, at least that’s how it works. So don’t guess too much. I hate when at rounds people say, “no, it won’t work, no, it won’t work.” But try and see. You know? Even if there is only 1 chance out of 10 that the patient may benefit from your intervention – just 1 out of 10 – it’s fantastic if we could win 1 out of 10 times when we [play the] lottery, we would [play the] lottery much more commonly. I mean, it’s fantastic. So don’t miss that opportunity. So, you know, patients with heart failure – as I tell my students – let’s imagine that now, I am very healthy, let’s imagine that now I have my myocardial infarction. Right now, right here. And I fall into cardiogenic shock, with massive lung edema, with severe hypoxemia. You take me to the hospital. You look at the chest X-ray. There is diffuse edema. What do you do? Do you give diuretics? That would be a mistake. Where does the water come from – the water in the lungs? Where does it come from? It comes from my vascular space, because there was a very abrupt increase in hydrostatic pressures associated with my massive myocardial infarction. So my teacher, Dr. Wyle wrote many years ago: try a fluid challenge in cardiogenic shock. And of course, now it is accepted, in cardiogenic shock we should try a fluid challenge. Not fluid administration, I open the bottles, I give 500 cc. No, no, no. Try a fluid challenge and see if the patient benefits or not. And here, since there is no decrease in vascular resistance, no decrease in vascular tone, you can just look at heart rate and blood pressure. You don’t need any sophisticated measure; you just look at that. And if the patient is in acute pulmonary edema, it’s not a perfect word, but people still speak about acute lung edema – OK, or cardiogenic lung edema, if you like – in these conditions there can be hypovolemia. But people don’t buy that when they see the chest X-ray. But in COVID-19, it’s the same thing. Initially people were focusing on the lungs and the respiratory failure, it may go into ARDS [acute respiratory distress syndrome]. Look at the chest X-ray, there is edema! Don’t give fluids, give diuretics! And these patients went into multi-organ failure with renal failure, which was usually the path to death in COVID-19, because you know, it adds insult to injury when there is renal failure added to the respiratory failure. But it was because the patients were kept hypovolemic. So, try to get the big picture: that’s the major message. And that’s what we should tell our junior people. Don’t try to look for easy recipes. The computer is not ready to replace the doctor. Maybe one day it will come with artificial intelligence, I think it will come, but it’s because our brain will be replaced by the computer brain. But we are not there yet.
Yaseen Arabi: Thank you. I think, Roman, we can take 2 more minutes. So just to finalize the last 2 questions. There is a question about IVC filters in the ICU. And there is a question about convalescent plasma. So I will do the IVC filter indication and then I will ask Waleed to comment on the convalescent plasma, to provide comment on this. So the IVC filter: in my talk, I summarized where the guidelines now stand on the use of an IVC filter. In patients who have venous thromboembolism and can be anticoagulated, at present IVC filters are not recommended in addition to the anticoagulation. In patients for primary prophylaxis, there is a randomized controlled trial in patients with multi-trauma; the patients were randomized to IVC filter versus non-; the primary end-point was not different, which was the combination of venous thromboembolism and death was not different between the 2 groups. However, there was a difference in one of the secondary end-points, which is PE in patients who could not be anticoagulated for the first 7 days. But at present, I think even for primary prophylaxis, the indication for an IVC filter is kind of questionable. The only place where an IVC filter might be useful is in patients who have a diagnosis of VTE [venous thromboembolism] but cannot be anticoagulated, have clear contra-indication to an IVC filter. We don’t have many randomized controlled trials on this one, but we know that an IVC filter actually captures clots – and that was shown in the RCT. And the question on convalescent plasma and COVID-19. Waleed, if you have a brief update of where the evidence stands now?
Waleed Alhazzani: Yes, I was waiting for the question. OK, so the question is about… So I assume we’re still talking about critically ill patients. You know, the idea of convalescent plasma is taking plasma from those recovering from COVID-19, so you have some sort of you know, transferring some sort of immunity to patients. I think even hypothetically speaking, if this were to work, it probably should happen early in the course of disease, rather than in somebody who [has] already developed organ failure and is suffering from severe consequences from, you know, inflammation. And the evidence so far, at least from 10 randomized trials that were summarized in a systematic review that included a mixture of a population who were hospitalized, the pooled estimate [of] relative risk sits at 1 – I’m talking about death – indicating no difference. So, I personally, if I was a patient or treating a patient with COVID-19, I wouldn’t want to receive convalescent plasma if I was intubated or in the ICU. I think it is very unlikely to have any role in that space.
Yaseen Arabi: Thank you, Waleed. I think this has been a great session with a lot of discussion and great questions from the audience as well. But certainly, I really enjoyed and learned a lot from all the speakers, it’s been fascinating. I’d also like to than Roman Jaeschke for inviting us to this course – a really amazing group of speakers and topics. So we’re grateful for this opportunity. Roman, do you have any concluding comments?
Roman Jaeschke: I just want to invite everybody who is listening to this session to the previous session, where we had a lot of COVID-19 covered. And I’m pretty sure you will find it fascinating from the critical care point of view as well. I am amazed by the quality of this discussion – it exceeded any of my or our organizers’ expectations and I want to thank you from the bottom of my heart. Thank you all.
Yaseen Arabi: Thank you very much. Bye.
