Dr Jean-Louis Vincent, professor of intensive care medicine at Université libre de Bruxelles, Belgium, past president of the World Federation of Societies of Intensive and Critical Care Medicine, and accomplished author and researcher, joins Dr Roman Jaeschke in the newest McMaster Perspective episode to scrutinize the latest Surviving Sepsis Campaign 2021 and shares his clinical experience.
For part 2 of the interview, click here.
Roman Jaeschke, MD, MSc: Good morning. Welcome to another edition of McMaster Perspective with Professor Jean-Louis Vincent, our star performer. Recent practice guidelines in critical care are making waves, and Professor Vincent essentially started this process of critical care practice guidelines 20 years ago or so. Could we start by exploring what you think about the role of practice guidelines in today’s critical care?
Jean-Louis Vincent, MD, PhD: Well, I think guidelines are fundamentally important, and as you say, that’s the reason why we started the guidelines for the management of sepsis, actually here, in Brussels, back in 2000s. Our first guidelines were published in 2001. But an important point was that we were only 9 people in the group, and we wrote the guidelines in hardly >2 days, maybe 2 days and a half, in that room nearby Brussels. Of course, the scientific societies wanted to join us, and that’s very good. One issue is that the number of participants increased to 60 people now. It’s very difficult to get a consensus with 60 people in the room when the evidence is not very strong. And the problem in the field of sepsis is that virtually all prospective randomized controlled trials (RCTs) have been negative, [showed] no difference in outcome. So, you could end up with guidelines saying “don’t do this,” “don’t do that,” “don’t do this,” but people would like to know what to do, and at some point the people who are perhaps more experienced and [who are] perhaps “more experts” should tell the others how to treat these patients. Exactly as I like to read guidelines about the management of lung cancer. I am not a specialist in lung cancer, and in half an hour I can read carefully a good article on the management of lung cancer. And I feel it’s really very useful, very useful, the “do this,” “do that,” “do this,” “don’t do that,” “do this,” “do that”... And you know how to manage these patients.
The problem with the Surviving Sepsis Campaign is that they wanted to go too broadly and to make some simplistic rules. So, doctors are no longer necessary. Just bring blood pressure to [the] 65 [mm Hg] mean value, just give 30 cc per 1 kg of fluids in 3 hours... A machine can do that. So, you don’t need a doctor at the bedside? What I would like to read is more about how you assess the need for fluids. How do you do that? How do you do a fluid challenge? How do you do a passive leg raising test if you want to use that test? How do you evaluate the appropriate blood pressure level? Instead of giving antibiotics within 1 hour in every septic patient, try to tell people: “When the situation is very bad, start very, very early.” It’s a matter of minutes in meningococcemia or in severe septic shock. If the patient is somewhat less ill, yeah, you may have 1 hour. You may have time to collect more samples, maybe to quickly call the infectious disease consultant, because this is a complicated casus with possibly resistant organisms. So, why can’t you personalize your attitude? We can give guidance for this. Maybe not 60 people in the room, but it’s possible to give guidance.
The other thing is that now they want to be so broad that they insist on having these guidelines applicable in low-income countries. And then they write: “Oh, lactate levels may not be available everywhere. Oh...” Put in the guidelines that we need to monitor lactate levels, and people in low-income countries can go to their management and say: “Look, these are the current guidelines, we need to have the instrument to measure lactate [levels]!” instead of showing a paper where it’s written: “If you cannot measure lactate levels, don’t do it.” Oh, it’s too easy for the management to say: “Oh, they acknowledged that we have other priorities, my friend. You know, we cannot afford everything in this country, so you cannot measure lactate levels.” No, let’s move forward. Let’s try to be better. Let’s try to provide better therapies. I’ll stop here.
Roman Jaeschke: Well, that was quite a tirade. I want to remind everybody that your first sentence was that practice guidelines are instrumental and have their use, and do some important work.
Jean-Louis Vincent: Oh, yeah. They’re very important. I love guidelines.
Roman Jaeschke: OK. At the same time, let’s go then into this other issue. Let’s talk about what you do, and I will ask you a series of questions on different recommendations. I’m less interested in hearing, you know, what the meta-analysis or a particular RCT is showing. I’m interested in what you are doing.
Jean-Louis Vincent: Whenever there is sepsis—which is infection plus organ dysfunction, as simple as that—I consider the 3 components in therapy: infection control (antibiotics, source control when possible), hemodynamic management (fluids, vasoactivation when indicated), and the host’s response. Corticosteroids, vasopressin, maybe gammaglobulins, maybe other therapies like thrombomodulin and other substances. These are the 3 aspects. And I go through the 3 of them at the same time but separately.
So, I consider the infectious disease aspect. Let’s start with the antibiotics. Let’s give the antibiotics that will cover for all possible microorganisms involved in this septic case, and let’s give them without any delay and actually very, very urgently if the situation is very severe, life-threatening. Source control, whenever possible. Don’t waste time. Tell the team that we need to control the source as soon as possible. If indeed there is a need for laparotomy for peritonitis, do not waste time. Let’s go for it.
Hemodynamic management... First of all, the current guidelines, despite my comments, have not included anything about endotracheal intubation. When to intubate the trachea? Not too early, of course, but not too late. And I see everywhere in the world septic patients who develop respiratory arrest and even cardiac arrest because they were not intubated on time. Let’s be careful that we give enough oxygen and intubate the trachea when needed. Fluid administration... If you follow the 4 phases of salvage, optimization, stabilization, and de-escalation—as we published in the New England Journal of Medicine (NEJM)—the salvage phase can be very short, maybe half an hour. Give 1 liter, quickly. But by that time don’t tell me you don’t have the time to have a monitoring technique. You could at least have an echo. Let’s start with an echo. Let’s evaluate the heart. How big is the heart? How contractile is it? It’s very easy. So, don’t plan to give fluids for 3 hours. Plan to give fluids for just half an hour, and then you individualize your fluid administration. This, today, is based on fluid challenges. We cannot have any static measurement telling you that this patient still needs fluids. You need to try and test. So, you give 200 cc in 10 minutes, not more. Don’t give 1 liter in half an hour. That’s too long and then, after half an hour, you don’t know what the effects of fluids were. Give a small amount of fluid. In an operating room (OR), it could be in 1 minute, which they call the mini-fluid challenge. OK, we are not in the OR for most of the cases, so let’s give a [small] amount of fluid over a short period of time and reevaluate cardiac output on the one hand—that’s a potential benefit—and a possible increase in cardiac filling pressures on the other hand—it could be a simple central venous pressure. Not a single measure, no, no, but delta central venous pressure. Did it increase much or not? If it did not increase much, if you feel that there may be some benefit, you give a fluid bolus again. Passive leg raising—it’s OK, physiologically it makes sense, but it’s not easy. I find it difficult because it’s so transient. The volume effect will take only a few seconds, so you need to have a very precise measurement of stroke volume, and the patient should not be stressed by the leg raising. But it’s very stressful, so I have difficulties with using that test. It sounds simple, but it’s not so simple.
OK. Vasoactivation... If there is profound hypotension, we need to start vasopressor agents right away. And, of course, it’s noradrenaline (norepinephrine), which is our first vasopressor to be given. You may recall our own prospective RCT, published in the NEJM, on norepinephrine versus dopamine. And clearly, dopamine is the loser. Don’t use dopamine any longer. Epinephrine is dangerous: it can create arrhythmias, it can vasoconstrict, it can increase lactate levels... Don’t use epinephrine. So, norepinephrine is the best agent, because it vasoconstricts, but it has some inotropic effects too. It has some beta-adrenergic properties. OK, when to start vasopressin? It’s already in the modulation of the host’s response. I wrote a paper, about 20 years ago, saying that vasopressin administration is part of endocrine support, like hydrocortisone administration, like insulin for hyperglycemia. To me, vasopressin is a hormone, and we should not use it as a vasopressor. We should use only low doses, up to 0.03 units per minute, not more, 0.03 units per minute, when we are sure that there is a hyperkinetic state, a high cardiac output state. I hear people saying “You need to give vasopressin for hypotension.” No, not for hypotension. For a situation where there is a low blood pressure associated with a high cardiac output to restore a vascular tone. With the Brazilian group, we showed an improvement in survival after cardiac surgery in hyperkinetic states when you give vasopressin early. So, it can save lives if you use it at limited doses in the right patient, but do not say “It’s just another vasopressor, why not?” No, no, no, no, it’s much more complex than that: it can decrease blood flow to the skin; it can decrease blood flow to the gut; it can decrease blood flow to the coronary circulation. So, vasopressin is not miraculous. It’s an excellent molecule for those who clearly need it.
If the patient is vasoconstricted, if there is myocardial depression, poor tolerance to fluids, check mixed venous oxygen saturation (SvO2). If SvO2, or [sometimes] central venous oxygen saturation (ScvO2), is low, it’s a clue to indicate that oxygen delivery is not adequate. So, what can you do? You can give more fluids or you can give a little bit of dobutamine. A little bit of dobutamine—it doesn’t hurt, a few micrograms per kilogram per minute. Heart rate will go up. Oh yeah, of course, heart rate will go up, but cardiac output will go up too, and therefore blood flow to the organs, to the kidney, to the brain, etc. And you can often see clinically the effects of adding a little bit of dobutamine.
The third option could be a transfusion. And now we see everywhere “Oh, no, no, no, no, don’t give transfusions unless hemoglobin is below 7!” Why? Seven or 70 maybe in Canada or in other countries if it’s in gram per liter. Why 7? “Because studies have shown that there is no improvement in survival.” Wait a minute. What are the studies having shown an improved survival with anything? More fluids? With what? Blood pressure >65 [mm Hg]? With what? Endotracheal intubation? With what? No, no, no, no, no. The problem in these large prospective RCTs is that the patient populations are very heterogenous. So, you may have some patients who received more transfusions when they didn’t need it in these trials, and other patients who were not transfused when they needed this. So, look at SvO2 [or, if not available, ScvO2], look at the risk of coronary artery disease, look at signs of altered tissue perfusion that persists. These are elements that should be taken into account in addition to hemoglobin levels to decide whether to give blood or not. Of course I would not give blood if there is no anemia. But if there is anemia, let’s ask more questions about the clinical background, clinical context, and the SvO2 [or, if not available, ScvO2]. Yes, indeed. So, individualize this.
And finally, you can do the same with lactate levels. Lactate levels should be measured repeatedly. Not to guide therapy, [as] the changes are quite slow, but to make sure that we are [heading] in the right direction. Check lactate levels every hour, maybe every 2 hours when the situation is less acute, to make sure that they go down nicely and they do not go up. This is not lactate clearance, as some people say. No, we are not studying the clearance. Actually, we should administer hypertonic lactate or radiolabeled lactate to study the clearance. No, no, it’s just lactate kinetics. It’s production more than elimination. Lactate levels will reflect—quite grossly, I agree—but they will reflect the oxygen metabolism of the cell. If the cell does not improve, the lactate levels will not go down and may even go up. If the situation improves, you will see the lactate levels going down.
So, all of this must be integrated. We should not look only at this, only at that. “I look only at blood pressure”—oh, terrible. “I look only at cardiac output”—oh, terrible. “I look only at SvO2”—no! “Only at lactate [levels]”—no, too slow. So, we need to combine these and other elements: urine outputs, skin perfusion, mentation, etc.
I didn’t mention it but incidentally sedation should be avoided in these patients. And the nurse often wants to sedate these patients, saying “Oh, it must be uncomfortable for the patient.” But why? Actually, we all note that when the patient is endotracheally intubated, after perhaps a little bit of dyspnea, the patient may fall asleep by himself or by herself. So, why would you add sedation? For comfort? No, no, no, I don’t need sedative agents to sleep at night, so it’s not for comfort. Pain should be controlled. Of course, always, always! Don’t allow the patient to have pain. Don’t accept it, of course. But sedative agents... They add insult to injury. All sedative agents will alter a vascular tone and myocardial contractility. So, if the patient is not so good hemodynamically, avoid sedative agents. Well, I can continue for a while, if you like but...
Roman Jaeschke: I think I will ask some more precise questions. I am glad that we now have both the Surviving Sepsis 2021 guidelines and Professor Jean-Louis Vincent’s guidelines, and they should be looked [at] in parallel, and both have their great role. I think this will be the end of our first part. I hope, in the second part of our interview, to talk to you shortly about several consecutive individual areas. Some of them you’ve already addressed, so work is already done. For the moment I want to thank you very much.