Transcript Video The Effect of Fever on Outcome After Cardiac ArrestProf. Markus Skrifvars < Back to Boundaries of Temperature Session 3: Utility The Effect of Fever on Outcome After Cardiac Arrest Presented by Professor Markus Skrifvars we come to the next speaker, Professor Marcus Cree Falls. He's coming from Finland and he is a specialist in anesthesiology and intensive care and professor of pre hospital emergency medicine and he was a member of the steering group of the T. T. M. Three trial. Please. Thank you so much for that and thank you to the organizers. It's a great pleasure to be here to talk to you about fever and and outcome after cardiac arrest. And some of these things may be a bit controversial but I've got my bulletproof west on so I'm I'm ready to all your comments afterwards uh with regards to temperature. Please bear in mind I'm a Fin I come from Finland, you know far away And in Finland temperature is different. We when it's -5° in Europe, that's what it's like in Finland when it's -30°. We only start to get going and why is this? Well it could be due to these pharmacological adjuncts that we use. I'm sure some of you tried some of these adjuncts yesterday as well and you felt good. This is very useful for fins because you know the finished summer. It's it's typically very rainy and it's typically very cold and it's also very very short. So it's very useful to have this kind of attitude towards temperature. Uh and we had an Australian guest yesterday. I'm not sure if he's here today but you know summer in Finland is a bit like winter in Australia. However this summer may be a little bit different. You know the weather forecast is fantastic and you know santa claus. He lives in Finland and this is what it looked like A few years back when we had a forecast like this. So even Rudolph the reindeer had to do a bit of chilling in the in the in the in the river around the arctic circle because it was so warm and even the Norwegian guests were happy to join in this swimming. So despite this or thanks to this, I'm not sure Finland is still the happiest country in the world. So please come and visit in the summer in the winter if you like. So moving on to the real talk uh science. So I present to you a little bit about the guidelines, what they say about how to manage temperature. Uh and then some difficult issues with regards to association between temperature and outcome. Both if you don't do any form of T. T. M. So we have a nice new data set that we have looked at and then also about what if you do T. T. M. For example at 33 or 36 you then look at the association of fever, a curing after that an outcome. And then finally where to from here how can we get some more answers about how to manage these patients? Well we start by a clinical vignette this is a very typical kind of patient. I'm sure you've seen tons of these 48 year old male found unresponsive on a walking path. He has a P. E. A pulseless electrical activity as his initial rhythm. And he needs 15 minutes of Cpr before he has regional spontaneous circulation and is brought to your I see you and that's the situation 20 hours from the arrest. So here you see the the monitors. His heart rate is nice and stable. Blood pressure is fine. No nor adrenaline running. His saturation is fine, entitled C. 02 but his temperature is now 37.9. So the question is now what do we do about this? Should we be worried? Should we give paracetamol? Should we start exposing the patient or should we start using the device? Should we put in an intravascular device? Or should we use noninvasive cooling? And that's of course the clinical problem that we face with regards to these patients and the injuries that the patient has. And of course it comes down to the situation of cardiac arrest, ischemia, re perfusion injury, inflammation and activation of a very complex system which then will interact with your brain. Especially if you have increased I. C. P. The increase in temperature may be very deleterious. That will further aggravate this inflammatory response. And then of course you have the lungs and other organs. You may have pulmonary oedema, chest compression may cause a little bit of pulmonary contusions that will further aggravate this inflammatory response and all in all. You'll end up in this vicious circle. Of course, we all know that fever can be a cause of an infection, You know, bacterial infection, viral infection. But there are lots of non infectious causes as well. And many of these are very relevant in this cardiac arrest patient. I mean the brain and it's it's it's it's also very difficult to to really know because the brain injury in itself will cause inflammation and may cause an increase in temperature. But but there's of course the other question, does the increase in temperature. Is it really a cause or result from this brain injury or not? Myocardial infarction? Many of these patients have myocardial infarction and they that can influence the temperature of these patients. And then of course aspiration many patients will have aspirated, they have bacterial components perhaps to that as well, that will further increase temperature. So with regards to the guidelines. So the guidelines on post cardiac arrest management were published last year but then given an updated systematic review about the evidence and the publication of the T. T. M. To study this was then discussed last autumn and then published as an additional paper in february this year. And we already professor donatello nicely showed these guidelines already. So we should be monitoring temperature. We should actively prevent and treat fever defined as temperatures higher than 37.7 and we should do that for 72 hours. How should we do it? We can expose the patient, we can give antibiotic drugs but if that doesn't work we need to use a cooling device and targeted temperature 37.5 the guidelines leaves open the possibility to to use cold temperatures if you feel that that's a patient sub population where you feel that that's that could be beneficial. The problem is of course what and who are those patients? How can we define them? And I don't think that the literature really provides us the final answer on that and evidently will need more studies more other studies. Looking at this in these subgroups of patients uh then it also gives the recommendation that we should not use large volumes of cold fluid in these patients. So of course the sad thing with any recommendation like this is that it's based on very weak evidence. We don't have the large trials that will show us if this really works. The only thing where the evidence is a little bit better is that we shouldn't be giving lots of cold I. V. Fluid. The rest it's it's it's very poor quality data and I think we owe it to the patients to do do better. Do better and larger trials so that we have better level of evidence in order to provide treatment recommendations for these patients. And this is of course not unique to cardiac arrest. So this we discussed already neurology yesterday when we got experts in this room who may want to comment. But in 2015 this group published in international journal of stroke with regards to man judgment of fever in stroke patients. And they came with the conclusion that there simply isn't enough evidence there to provide any recommendation whether fever should be treated actively or not just because the evidence is so mixed. But then what is the evidence like in cardiac arrest patients? Well, Michael Holtz gave an excellent talk yesterday and he may have some additional comments on this paper here from the Vienna group. One of the first studies to look at this in 2001 prior to the Hakka study, they made this retrospective study about 250 patients very typical male middle aged patients and looked at their temperature trajectories over the 1st 48 hours in the I. C. U. The findings are shown here so it circles you see the patients who made a good functional recovery And with triangles the ones who had a a poor functional recovery. And interestingly if you look at these curves. So there's 37° and 38°. They're not really that different apart from what happens here at 48 hours there seems to be a persisting fever in the patients with who eventually have a porn ecological recovery compared to the patients who have a good neurological recovery. So of course the 100 billion question is can we do something here to influence what what happens here. We had the possibility to look at this because some colleagues of mine did the Fin Sooskee study. So an observational study done 10 years ago. So this was not randomized. This was simply looking at all cardiac risk patients. 500 cardiac risk patients treated at 200 different. I see us in Finland And we collected lots of data. All the electronic data monitoring systems were connected to a database. So we had lots of data on cardiac arrest, the initial rhythm, the later return of spontaneous circulation and also functional recovery at 12 months from the event. And because all this data was automatically collected. We had a very highly granulated database of these patients. And in this we kind of in this study that we've just done or this preliminary analysis that we've done this year, we looked at the ones who were not treated with T. T. M. So this was 2010 to 2011. So at that time T. T. M. Meant T. T. M. At 33. And most I see us in Finland were very strictly following the Haka trial inclusion criteria. So ventricular fibrillation and a certain delay to return of spontaneous circulation. That was the target group. And basically those who were not treated with T. T. M. We're basically not treated with any form of device. They may have had percy tamal, they may have been exposed. But they were definitely they they did not receive any form of feedback control device. Uh So we had a lot of temperature and I'll show you the results in just a second. We defined prior to this that fever was considered a temperature higher than 37.7. So all in all this is very busy. But showing you here the temperatures from 0 to 72 hours from cardiac arrest and each dot is one temperature in one individual patient. So this is data from 200 cardiac arrest patients not treated with T. T. M. And here are their mean temperatures. So the first striking thing that was very surprising to us was that really, really high fever wasn't really that common, that we had thought, and we instead had plenty of spontaneous hypothermia. So patients who were not cooled, but they were just cold. Uh And of course, some of this will be the patient with the horrible type of brain injury where the autonomic nervous system shuts down and the patient doesn't move, they don't breathe, they don't create, they don't use any energy. And that's why they're hypothermic. We've all seen that a brain dead patient is not hypothermic. They are hypothermic, but we didn't really see that much high fever that we had expected to see. So we looked at which factors predicted the occurrence of fever. So, looking at mainly factors prior to arriving, the I. C. Was the witness, the rest shock, herbal rhythm delayed return of spontaneous circulation. Bystander cpr. The one thing that came out very strongly in a multi variable model was non shock kable rhythm. So they were more common to have fever after after cardiac arrest. So then looking at outcome. So we included these factors in a multi variable model and tried to see whether hypothermia or fever higher than 37.7 whether that predicted outcome. And we didn't really find any any such finding in this analysis here here. So another way to show these findings would be like with a fit plot like this. So on the y axis you see an odds ratio for increase in mortality. So the higher the number, the worse it is for the patient and here you see temperature. So here The first seven for 24 hours and here's the 1st 72 hours. And basically what you'll see is that the odds ratio for worse outcome is very stable over quite a wide range of temperatures. So the other thing here is that you can see that yes, probably very high temperatures may be harmful, but also very low temperatures spontaneously low temperatures may be harmful. So, kind of the thing I want to show with this figure is that it's very very difficult. I I put two it's impossible to to to to be sure whether we can change outcome by doing something about these low temperatures and also doing something about these high temperatures just because of the nature of the data, Just because of the data is observational and we have these two different types of patient who spontaneously are either hypothermic or hypothermic. So to to be sure with an observational study is in my mind completely impossible. So we need randomized trials to make sure to be to be really sure about what really works. So how about T. T. M. At 33 degrees? What about fever after that? So this was shown by Professor donatello. So data from Denmark 400 patients, 270 treated with T. T. M. At 33. This was before the T. T. M. Study, 50 patients developed fever. They had clearly higher mortality than the ones who did not develop fever. And also they looked at which factors were more common in the patients who had a fever compared to those that did not. And you can see comparison of lots of factors, nothing really significant apart from the initial temperature. When the patient arrived in the ICU, which kind of makes sense. If you arrived with a slightly higher temperature, then you might be more likely to develop fever afterwards. So the other interesting finding that they were able to show that yes, there was a difference in outcomes of patients who had fever had higher mortality than the ones that did not. But when they looked a little bit further at what what type of fever was harmful. So here's your highest temperature and again, has had ratio. So the higher the number, the worse the outcome for the patient. And as you can see that again, it's quite a wide range of temperatures where there's no increase in mortality. But then with lower temperatures and also higher temperatures, this seems to change. But the other very interesting finding was that timing was very important. So how long do you have your fever for? So fever higher than 24 hours or 20 hours? Very sharp increase in your mortality. So it might not only be the peak temperature, it's more this that we should be avoiding long periods of treatment resistant and fever. So we also had a look at this in the t th 48 trial which is a trial comparing 24 48 hours of fever. And we looked at fever a curing after the intervention. And basically what we were able to show that the patients who had rebound fever were more likely to be young, more likely to be male and more likely to have received mechanical chest compressions. So perhaps it's the mechanical chest compressions can cause a little bit of more of organ injury, lung contusion and maybe that will associate as well with the development of fever in some of these patients. And also interestingly high crp was associated with fever. So the higher the worst the inflammatory response, the more likely that the patient has develops post rewarming post T. T. M. Fever. So we also looked at the duration and the magnitude. So we calculated areas above these different thresholds and then compared the good and the poor outcome patients. And if we look here we see that at 38.5 this difference becomes statistically significant. So higher times and magnitude over these different thresholds were more common in the in the poor outcome patients. But overall this was not very common in the poor outcome patients either. So the opposite has actually also been shown. This is some data from Korea. Very provocative study can be discussed, but interestingly they showed exactly the opposite. So here are the patients who did not develop fever that many had a poor outcome and and that many had a favorable outcome. And these are the ones who developed fever. So, again, giving some more insight that this is incredibly complex and very difficult to study in observational studies because of these two phenotype of patients hypothermic and spontaneously hypothermic, which will be typical for your very severe brain injury patients. So, my conclusions, I think it's it's very conflicting this evidence. The more you look at it, the more you read about it, the the more more you realize that that we know so little about about this problem and I don't think that's surprising. It's it's a very complex syndrome, post cardiac arrest the disease fever peaks may be less harmful than prolonged periods of fever may be the case. And to our surprise, very high fever during the 1st 24-48 hours was in fact less common that we had thought guidelines are based on low levels of evidence but we should follow them at the moment. So my recommendation might take home to you is to follow the guidelines treat fever actively. If your conservative measures don't work you should use a device to control temperature. But definitely I think we need more trials on this topic. We need to know whether which which are the targets that we should be treating and which are the methods. So the question is is fever management the parachute of medicine something that we can't study that we can't do a trial on. Well I would put to you that it's it's not. Uh And I also think that one problem with all these interventions about T. T. M. Is that we simply have done too small trials if we think about this primary injury, this duration of ischemia, this complex thing that this patient goes through through, it's probably likely that the things that we can do in the I. C. U. The magnitude the effect size of what we can do is probably much much smaller than what we have thought. Hocker trial TC 48 these trials were powered for effect sizes of 15%. I think that's probably unrealistic and we should be looking at much much smaller effect sizes. Indeed a Beijing meet analysis that we did probably said that if there's an effect there at the moment it's probably around 34%. So to to find that out, we need much larger studies. So we propose a factorial randomized control trial that we call the step care trial where we include include these different interventions. So continuing or stopping sedation in patients managing fever very actively with the device or more conservative measures and then targeting a high or a very standard blood pressure. And to do this we will aim to randomize 3100 patients all commerce. So not only the only group will exclude are the non traumatic cardiac arrest. We have some funding, plenty of funding. So we will start this trial this year. It will take many years to complete and we're able to detect the 6% absolute difference in outcome, which is not ideal. But it's it's it's what we can do with the funding available. Thanks to the organizers. I'm happy to take any comments via email and we can we will discuss the these and the other presentations in the discussion and yes, these are some nice pictures from Finland in the summer. No, just kidding in the winter from a very talented young photographer and with that I conclude Created by
