Transcript
Speaker: Welcome to Critical Matters, a sound critical care podcast covering a broad range of topics related to the practice of intensive care medicine.
Speaker: And now your host, Dr. Sergio Zanotti.
Speaker: Welcome to another episode of Critical Matters.
Speaker: Today we will discuss seizures in the ICU.
Speaker: It's a pleasure and an honor to have as our guest Dr. Thomas Bleck.
Speaker: Dr. Bleck is professor of neurological sciences, neurosurgery, internal medicine, and anesthesiology at Rush University Medical Center, where he's also a practicing neurointensivist and director of the clinical neurophysiology.
Speaker: Dr. Bleck is board certified in internal medicine with subspecialty certification in critical care medicine,
Speaker: also certified in neurology with subspecialty certifications in vascular neurology and epilepsy, clinical neurophysiology, and in neurocritical care.
Speaker: Dr. Bleck has been elected as a Master of Critical Care Medicine by the American College of Critical Care Medicine and Society of Critical Care Medicine and has received numerous awards for his contributions to critical care and to neurocritical care.
Speaker: He's one of the founding fathers of neurocritical care as a specialty,
Speaker: and is an ex-official member of the board of directors of the Neurocritical Care Society, of which he was the founding president.
Speaker: It's truly a pleasure to have him today to discuss seizures.
Speaker: Tom, welcome to Critical Matters.
Speaker: Thank you.
Speaker: So I think that a good place to start would be, I tend to notice that when a patient has a seizure in the hospital, where it's in the emergency department, on the floors, or even in the ICU,
Speaker: And perhaps this doesn't happen in a neuro ICU, but at least in general ICUs, it seems that nobody really knows what to do.
Speaker: It's one of these very traumatic events, both for families, but also, I think, for the staff.
Speaker: And it always catches my attention that nobody really seems how to deal with these seizures.
Speaker: Could you tell us, to begin with, Tom, what should we do when somebody sees it in front of us?
Speaker: I think the first thing to remember is that almost all seizures are going to end spontaneously.
Speaker: So there is no need to immediately rush in with some sort of therapy.
Speaker: Rather, observe how the seizure starts and mostly try to protect the patient from any harm.
Speaker: You don't want the patient, for example, to get smothered by the pillow or the sheets, which still happens from time to time.
Speaker: In the hospital, most of the things that are harmful out in the community won't be present, like the ability to
Speaker: bump into a radiator or put your hand on an open glass flame.
Speaker: But people can still be injured in the hospital, especially if they're seizing down on the floor where they can collide with objects.
Speaker: So when the seizure is over, which in the vast majority will end spontaneously, turn the patient to the side in the recovery position to try to prevent them from aspirating.
Speaker: Of course, if they're already intubated, that's not an issue.
Speaker: Secretions in the mouth can be a problem, but don't stick anything in the mouth except perhaps to separate the teeth if they're chewing on their tongue.
Speaker: Now in the small number of cases where the seizure doesn't end spontaneously, one should then proceed to use pharmacologic therapy to terminate it.
Speaker: It's usually not necessary to immediately intubate patients who are not intubated in a
Speaker: A judicious dose of a benzodiazepine in most patients will stop the seizure.
Speaker: If the patient already has an IV, then 0.1 milligram per kilo dose of lorazepam is appropriate for a convulsion.
Speaker: If the patient happens not to have an IV, then probably 10 to 20 milligrams of midazolam IM, based on the Rampart trial, would be superior to taking the time to start an IV.
Speaker: That's one of the things we have
Speaker: learned from randomized study most recently.
Speaker: About 65% of patients who require benzodiazepines to stop their seizure will stop seizing.
Speaker: Some percentage of them will go from convulsive seizure activity to non-convulsive status epilepticus.
Speaker: So if they continue to seize or they don't begin to wake up, then one would move on to second-line therapy.
Speaker: We're in the midst of a trial called ESET, the established status epilepticus treatment trial, that's comparing three of the usual anti-seizure drugs to see whether there's any advantage to one of them.
Speaker: So the drugs being tried are Phosphonatoin, 20 milligrams per kilo, Velproate, 40 milligrams per kilo, and Levataracetam, 60 milligrams per kilo.
Speaker: At this point, near the end of the study, we've stopped randomizing adults at the request of the DSMB because they believe we will not be able to establish the superiority of one of these agents over another.
Speaker: We're still randomizing children.
Speaker: The important thing to take away from that, though, is that a 60 milligram dose of levotiracetam, 60 milligrams per kilo, is a dose that is
Speaker: at least as safe as the others and at least as efficacious.
Speaker: So if you're going to use levotiracetam, you really need to use an adequate dose, giving 500 milligrams or a thousand milligrams.
Speaker: And considering that that's a failed dose because the patient continues to seize, really hasn't given the drug an adequate try.
Speaker: Within a year or so, we hope to have the
Speaker: results to that study ready for publication to see if there are any nuances that will be useful or to see if one of the drugs is superior when treating children.
Speaker: Tom, since you mentioned – Can I interrupt and ask you a quick question?
Speaker: You mentioned the underdosing of levteracetam, and I think that that's not an uncommon theme with other antiepileptics, including benzos.
Speaker: Could you comment on that a little bit?
Speaker: Because I think that a lot of
Speaker: intensivists who are not neuro-intensivists or neurologists tend to underdose patients as they're treating the seizures.
Speaker: Absolutely.
Speaker: And we know this actually from data being collected that were done subsequently focusing on children that people are not giving adequate initial doses of abenzodiazepine, which use
Speaker: 0.1 milligrams per kilo of lorazepam as their dose had again 65% or so effectiveness giving larger doses appears not to be more useful but giving less is not an adequate trial so focusing on lorazepam since most of the patients seen by Intensivus will already have an IV in place you need to make sure you've given an adequate dose
Speaker: If the patient's had a behavior change and your question is whether this could be a non-convulsive seizure and you want to give a few milligrams of lorazepam, I guess that's all right.
Speaker: But there's probably no downside to giving the whole dose other than some sedation.
Speaker: And if you haven't given it an adequate dose in convulsive status, then you've really missed the boat.
Speaker: And I think that there's also good evidence to suggest that
Speaker: intervention after the seizure has stopped in terms of preventing patients to go into status as a time-sensitive intervention and that we should be loading these patients as soon as possible, like you say, to take care of the patient, make sure they're safe.
Speaker: Is that correct?
Speaker: Yeah, so it depends in part on why you think the patient had a seizure.
Speaker: So if someone is having one or a cluster of alcohol withdrawal seizures, for example, there's probably no advantage to putting them on a
Speaker: chronic anti-seizure drug.
Speaker: But if you think that they're seizing because they've had an acute brain injury or they have some intoxication, they have a CNS or a systemic infection, then yeah, go ahead and give an adequate loading dose of a more long-acting anti-seizure drug.
Speaker: And as we're talking about the initial treatment, could you mention some tips for us, for our listeners, in terms of identifying differential diagnosis that can be sometimes confused with seizures, including pseudo-seizures?
Speaker: Right.
Speaker: So I would have to start by a caveat, which is that 20 years ago, we thought we knew how to identify pseudo-seizures pretty well.
Speaker: And as we've gone to do more recording to prove or disprove that contention, there are a fair number of movements that we used to think of as diagnostic of pseudo seizures that can be part of real seizures or epileptic seizures, I should say.
Speaker: So I'm still a little reticent to label things purely on clinical grounds as being a pseudo seizure, with the exception of something that I can stop by distracting the patient.
Speaker: People used to think that bicycling movements of the legs were characteristic of pseudo seizures, but some frontal lobe seizures will also always produce that.
Speaker: Most of the time, if someone is moving both sides of their body or complaining of bilateral sensory phenomena while they're awake and responding to you, then almost all the time that will end up being a non-epileptic seizure.
Speaker: But one has to be pretty careful to be sure that you don't label somebody as having paroxysmal non-epileptic seizures without having EEG recording to back you up.
Speaker: So clearly, the role of the EEG in making a definitive diagnosis is still, I mean, the gold standard and correlating what you're seeing clinically.
Speaker: That's correct.
Speaker: Is there, you mentioned the ISET trial, it's ongoing.
Speaker: So as of now, there's no clear evidence to suggest that one antiepileptic is superior to another, is that correct?
Speaker: Right, as the second-line drug.
Speaker: We're pretty sure that lorazepam is the best first-line drug, or if you don't have an IV, midazolam, based on the VA cooperative trial.
Speaker: But of the second-line drugs,
Speaker: Adequate doses of phosphatidine, valproate, or levotiracetam appear to be most likely equivalent.
Speaker: We won't know for sure in terms of efficacy versus safety issues that led the DSMB to stop randomizing adults.
Speaker: That trial does not include other agents.
Speaker: So for example, licosamide was not included because it didn't have an indication for pediatric use at the time that we were designing the trial.
Speaker: But licosamide is probably also a reasonable choice at that juncture.
Speaker: People have tended to use 200 to 400 milligrams as an initial dose, but
Speaker: The group at the Ochsner Clinic in New Orleans has suggested that a loading dose of 10 to 15 milligrams per kilo is actually what would be required and is more efficacious, but that remains to be confirmed by other studies.
Speaker: And it seems that one of the important points that you made earlier regarding dosing of
Speaker: levotiracetam is very important to reemphasize because I do see that this drug, commercially known as Keppra, seems to be utilized more and more, especially in the community hospitals.
Speaker: But it would seem, from what you mentioned, that people are consistently underdosing this drug.
Speaker: That's correct.
Speaker: One of the first things we do is to finish the load when we get the patient in transfer.
Speaker: There is a related drug newly released that is available for intravenous use called rivaracetam, but at this point we don't know where it fits into the armamentarium for status.
Speaker: And is there still a role for patients with status for other drugs like phenobarbital and other drugs that we've used in the past?
Speaker: So we do use a lot of phenobarbital, but generally
Speaker: after we've gotten the seizures under control since it takes a while to load.
Speaker: I think the 20 milligram per kilogram loading dose can be quite effective.
Speaker: We used to use more pentabarb in a third-line drug, but pentabarb has gotten very expensive, and I think the efficacy of phenobarbital is probably at least as good, and it seems to have less hypotension than pentobarbital does.
Speaker: Is there any role for propofol?
Speaker: So I would think of propofol as I failed the second line and now I'm going to go on to general anesthesia.
Speaker: So whether there's any difference between using propofol or using high-dose midazolam at a 0.2 milligram per kilo load and then starting with 0.2 milligrams per kilogram per hour as an infusion,
Speaker: compared to propofol at about 80 mics per kilo per minute as a starting dose, remains to be determined.
Speaker: There's no good comparative study in a large group of patients.
Speaker: If you have an anaconda available and can give an inhalational agent, one might also consider using isoflurane or desflurane.
Speaker: But again, these are getting into small numbers of reports as to what's efficacious.
Speaker: Our practice has been
Speaker: in adults to usually go to propofol if the second line drug has failed and we think we need to get immediate control of the patient and status or to use midazolam.
Speaker: I've always been a fan of midazolam, but you have to be willing to start with a large loading dose and then continue to increase the dose.
Speaker: So partly it'll depend on who's going to be watching the EEG.
Speaker: not able to get a full-time observation of the EEG by someone who knows what they're looking at, I may opt to use propofol.
Speaker: Otherwise, I'll use midazolam.
Speaker: And I think that we mentioned status and in medicine, I think it's important to be very clear on what we're referring to.
Speaker: And it's not uncommon, at least when I get calls, when I'm working clinical, to hear of somebody who had a seizure, got intubated, and they're being admitted for status.
Speaker: Could you define for us how we really understand status epilepticus today, Tom?
Speaker: Sure.
Speaker: So in terms of an operational definition, the
Speaker: International League Against Epilepsy has proposed a different time for what should be considered status in order to start treatment or what should be considered the point at which status is established or refractory.
Speaker: So in most circumstances that would come to the attention of the intensivist or the emergency physician, five minutes of seizure activity should be considered status enough to start the benzodiazepine treatment.
Speaker: If the benzodiazepine succeeds, then you want to try to figure out from an etiologic standpoint what you're dealing with.
Speaker: So the person who goes into status because they've become abruptly hyponatremic should get their osmolality repaired in a slight degree.
Speaker: Whereas the person who has a brain abscess and goes into status will need anti-seizure drugs, often several of them.
Speaker: Once you have given that benzodiazepine, if the patient's still seizing, then you're thinking of the second line established status.
Speaker: And there you should try to get control of the patient within 20 to 30 minutes.
Speaker: That's how long it will usually take to get the drug that you need available and administered.
Speaker: After that, then moving on to general anesthesia.
Speaker: five minutes meets the definition of status as far as starting treatment, the time at which the uncontrolled seizures by themselves start to produce brain damage is probably closer to 30 minutes.
Speaker: So one of the reasons that initial five minutes was chosen was to have a reasonable chance of the initial treatment getting the patient under control.
Speaker: And because of those five minutes, I would imagine that the vast majority of seizures will last less than three minutes, two minutes.
Speaker: Usually, like you said, they're usually resolved by themselves.
Speaker: So most people who have a convulsion, it would last less than a couple minutes, correct?
Speaker: Right.
Speaker: It always seems like it lasts a long time, but when you time them out, they're usually somewhere in the vicinity of 90 seconds to two minutes.
Speaker: And that was also one of the...
Speaker: I guess, pearls that I had learned regarding pseudosezures.
Speaker: People would say that if seizures are very prolonged, they're more likely to be pseudoseizures than seizures.
Speaker: But like you said, without ENG confirmation, that would probably not be a very smart thing to put all your bet on today.
Speaker: Right.
Speaker: If I'm not sure, I'd say if the patient is protecting their airway,
Speaker: then I will try to get the EEG before I would go on to a general anesthetic regimen.
Speaker: I might have given the initial benzo and a second line agent.
Speaker: But I have seen patients who were very convincing and got to the stage of general anesthesia where it became clear with subsequent spells that their seizures were not epileptic.
Speaker: So one shouldn't feel bad for trying to do your best until the EEG arrives.
Speaker: And you mentioned in a couple of times already the airway management of these patients.
Speaker: My impression is that people tend to have a knee-jerk reflex to intubate somebody who's seized.
Speaker: And I think that often with close observation, not everybody needs to be intubated.
Speaker: What are your thoughts on this, Tom?
Speaker: So I think since most of the time, at least the 65% of the time, you'll get control with your first drug.
Speaker: that's often the best way to basically avoid the need for intubation to go ahead and give the lorazepam or medazolam.
Speaker: If the seizure is not controlled with that but the patient is still protecting their airway then I think it's fair to load them with one of the second line agents.
Speaker: If they are convulsing it's hard to keep them in the recovery position but if you can
Speaker: keep them turned to the side to try to prevent secretions from going back in the airway.
Speaker: On the other hand, if in that circumstance they're starting to desaturate or you're having trouble keeping the airway cleared short of intubation, then I wouldn't be shy about intubating them.
Speaker: The choice of drugs at that point I've always found somewhat interesting.
Speaker: If you know the patient was ambulatory and normal before the seizures started, then
Speaker: the usual rapid sequence induction is probably okay, but if there's any question in your mind about preexisting neurologic or neuromuscular disease, should probably stay away from succinylcholine because of the risk of hyperkalemia and use rocuronium or some other rapid-acting agent, if necessary, reversing it with sugamidex.
Speaker: The sedative drug that you give along with that to intubate the patient, say you're giving propylacinol,
Speaker: will probably stop most of the residual seizures but often for a shorter period of time in a single dose than the neuromuscular junction blocker lasts.
Speaker: So you don't really know that whether the patient is seizing or not until the neuromuscular junction blocker wears off in which case you might have been seizing for quite some time if it lasted 40 minutes and the propofol effect was 20 minutes.
Speaker: So I would say
Speaker: to first go ahead and try and get the EEG hooked up as quickly as possible, but also consider reversing the neuromuscular junction blocker in order to get a better exam.
Speaker: And in terms of treating these patients once they're intubated, if they're not waking up, should this be a concern for us?
Speaker: And at what point should we be concerned about potential non-convulsive status?
Speaker: The patient who got the 0.1 milligram per kilo dose of lorazepam will probably be somewhat sedated for a half hour or so from that, although some of them do wake up pretty dramatically.
Speaker: But having gone through the extra sedation of intubating them, I would usually be watching for any sign of recovery in terms of even response to noxious stimuli.
Speaker: If they're not responding to noxious stimuli, once the neuromuscular junction blockers are burst or worn off, I would go ahead and load them with a second line agent.
Speaker: Once the EEG is hooked up, then you'll have an answer to your question.
Speaker: If you're in a circumstance where you can't get the EEG, you've loaded with the second line agent and they're not waking up, I really don't think you can manage them without transferring them somewhere to get the EEG done.
Speaker: know of a way to do that.
Speaker: And I think that that is a problem that a lot of community hospitals face.
Speaker: And I have seen community hospitals try to initiate programs of continuous EEG so they don't transfer the patient.
Speaker: But then my question is, who's reading the EEG?
Speaker: Because that seems to be
Speaker: as difficult as just getting the logistics of the EEG connected.
Speaker: Could you comment in terms of ideally when would you think that so the EEG really should be done within the first 12 hours or initiated within the first 12 hours of this whole sequence is that correct?
Speaker: Oh I would want to do it sooner than that.
Speaker: I'd say the future of this field now there are a couple of devices that are sort of quick connect limited montage EEGs
Speaker: that I think people should consider if they can't get a standard EEG hooked up 24-7.
Speaker: We have yet to actually study these devices in status, but there are already some published studies using them in emergency departments and other parts of the hospital to show that they are quite efficient at
Speaker: both being attached by non-trained technologists, it can be attached by physicians, nurses, respiratory therapists fairly quickly, and they can also send the EEG data to a central reader over the internet so that if you acquire one of these devices and have the reader available off-site, you could probably manage the patient without the transfer.
Speaker: And these are rapid sequence terms that they have less number of channels, is that correct?
Speaker: And they're easier to place?
Speaker: Right.
Speaker: So one of them, for example, actually two of them have just eight contacts, so you get four channels on each side.
Speaker: when people have studied what happens when you reduce the montages, you miss some of the information, but if your question is whether the patient's gone into non-convulsive status, they are, in my mind, quite useful and can at least answer the question in the short term so that you know whether you need to proceed or not.
Speaker: So it's much more like, I guess, the analogy would be when you're doing a point-of-care ultrasound to answer a specific question, yes or no.
Speaker: Do they have a pericardial diffusion, yes or no?
Speaker: I mean, obviously, there's a lot of detail that might not be there, but for a dichotomous question or binary question, it can be very helpful in terms of managing the patient at that point.
Speaker: Exactly.
Speaker: Exactly.
Speaker: And in terms of continuous EEG, how long, I guess it depends, but what would be the minimum that you would continue this for 48 hours once the seizures have stopped?
Speaker: I mean, what's the usual recommendation, Tom, these days?
Speaker: Right, so again, the data that have been collected aren't specifically for patients who are in status, but if you look at studies of patients in ICUs with various brain injuries,
Speaker: of the ones who will have seizures you picked up about 90% of them with 24 hours of monitoring and about 95% with 48 hours.
Speaker: Going from 48 to 72 you only pick up another 3% or so.
Speaker: So you can basically look at the data and decide if you're content with 90 or 95 or 98%.
Speaker: Of course if the patient is improving and you can follow them clinically then
Speaker: At that point, you can disconnect the EEG.
Speaker: These are for people who remain poorly responsive.
Speaker: So another area that I think creates a lot of confusion amongst intensivists outside of the neuro ICU has to do with potential seizures, non-convulsive seizures, and other abnormal waveforms that identify an EEG post-anoxic brain injury after cardiac arrest and after patients have undergone targeted temperature management.
Speaker: So the first question I have for you, Tom, in this area is, should every patient that has targeted temperature management
Speaker: be followed with continuous EEG?
Speaker: So first let me say that I do that.
Speaker: I think that is the best way to manage them, to know if they're having any seizures even if they're getting neuromuscular junction blockade for shivering.
Speaker: If you are not giving a neuromuscular junction blockade, you may miss some non-convulsive seizures as they're warming up if you don't do EEG monitoring
Speaker: My own bias is that the patient who has seizures should be treated for them, but that we shouldn't be giving prophylactic anti-seizure drugs in the absence of seizures.
Speaker: There's a very good study from the EEG group at Northwestern that looked at
Speaker: the myoclonic activity that is sometimes seen when patients are either being cooled or more likely when they're being rewarmed.
Speaker: And based on the EEG recordings they've done and other people have done, we really don't think that the myoclonic activity in most cases represents seizure activity.
Speaker: It's coming from some lower center.
Speaker: And the group from Northwestern has also published a nice study looking at the phenotype of these myoclonic movements and providing guidance on which movements are likely to portend a poor outcome versus those in which some of the patients have made good recoveries.
Speaker: Last author of that is Stefan Shula at Northwestern, and it's worth consulting if you're dealing with a patient who has myoclonus after a cardiac arrest.
Speaker: And I think that's a very – we'll put the reference in the show notes.
Speaker: But I was going to actually ask you this specifically because I often see intensivists mention that the presence or when they see myoclonic movement in post-arrest patients, no matter where it is in the time frame,
Speaker: to immediately associate that with a very poor prognosis.
Speaker: And my understanding has always been that what has been shown to really be a very bad prognosis before the area of hypothermia was myoclonic status epilepticus, which is an EEG diagnosis.
Speaker: And I've never seen that.
Speaker: Could you comment on that?
Speaker: Yeah, so very few post-arrest patients have real status epilepticus.
Speaker: They may have repetitive myoclonus, but it doesn't respond to anti-seizure drugs,
Speaker: It doesn't have the typical evolution of a seizure.
Speaker: And unfortunately, the American Academy of Neurology's guidelines and practice parameters in the past didn't distinguish these very clearly.
Speaker: I think the ones that are currently under review will do a better job of distinguishing that.
Speaker: If somebody has real seizures,
Speaker: based on EEG and by the evolution of their clinical picture, I would certainly treat that.
Speaker: There are recent publications suggesting that having seizures is bad, but not completely damning in your prognosis.
Speaker: By the way, we mentioned that we're really coming around to sort of regardless of any clinical findings, one should wait for about three or more days after
Speaker: rewarming before prognosticating that people who remain unresponsive with poor cranial nerve function have still begun to awaken in seven days and had a good outcome.
Speaker: So I think there's been a lot of premature termination of aggressive treatment for these patients.
Speaker: David Greer, who's now at Boston University, has done some very good work in this area, and there are more publications coming out every month.
Speaker: Absolutely.
Speaker: I think that with the advent of target temperature management, we've changed the old natural evolution or course of this disease of anoxic brain injury.
Speaker: And I do believe that today we have to wait more time.
Speaker: And I think a lot of people attribute findings as being prognosticator when they really aren't in that timeframe.
Speaker: And that's something that we've talked in a previous podcast episode with Fred Rincon, which I think is very important.
Speaker: So, Tom, one of the things that always comes to mind when I think of these patients is, are seizures just a byproduct of the damage and they just represent how poorly this patient would do?
Speaker: Or are there something that are treatable, especially when these are non-convulsive seizures or when we see PLEDS or BIPLEDS on the EEG, post-anoxic brain damage?
Speaker: Are these things that just represent a poor outcome or should they be treated aggressively?
Speaker: Well, so...
Speaker: Having either seizures or the repetitive EEG complexes that you've just mentioned carry a worse prognosis than not having them.
Speaker: But the thing on the EEG that is most useful in terms of prognostication actually is whether the background activity changes when you give sensory stimulation.
Speaker: So that whether it gets
Speaker: desynchronized meaning gets faster and lower in amplitude or shows what people have always called paradoxical responses where the sensory stimulation results in big slow waves.
Speaker: Any stimulus given either auditory or somatosensory that causes a change in the EEG means that signal is getting to the cortex and the cortex is making a response.
Speaker: And that looks to be the best
Speaker: single predictor of a good outcome in the patient who's been cooled after cardiac arrest.
Speaker: In the pre-hypothermia era, the presence or absence of the N20 response to somatosensory evoked potential testing was thought to be the single best predictor.
Speaker: But there have now been numbers of patients who lacked that response early on who went on to make good recoveries.
Speaker: So right now we would say that EEG reactivity is the single best.
Speaker: There have been a couple of studies in recent years that look at combinations of variables.
Speaker: The EEG responsiveness, the N20, perhaps adding either neuron-specific enolase or tau measurements in serum to try and improve that.
Speaker: The tau is still fairly experimental, but the TTM group
Speaker: in Scandinavia just published a large number of such patients.
Speaker: So I think we're going to be improving our prognostication.
Speaker: But right now I'd say EEG reactivity is the single most important thing to look for.
Speaker: Just having periodic complexes or seizures by themselves don't mean you can't recover, though.
Speaker: So let me ask you, in terms of how you utilize EEG during your treatment of patients with anoxic brain injury,
Speaker: So I presume from what you told me earlier that as you initiate the treatment, while you get them to target temperature and during the time that you are at target temperature and rewarming, you have continuous EEG.
Speaker: Is that correct?
Speaker: Yes.
Speaker: Once they are rewarmed, if the EEG is not showing you any seizure activity, do you stop the EEG at that point and give them some time to see what happens?
Speaker: Yes.
Speaker: I think if there's no seizure activity going on,
Speaker: Other than looking to see whether there's reactivity, say once they're rewarmed, I don't feel compelled to continue the EEG recording.
Speaker: If they have phenomena that nobody can explain easily, though sometimes seizures are just producing autonomic changes.
Speaker: So I'd say I haven't done this formally, but probably half of the patients
Speaker: do stay on EEG for one or more days after they're rewarmed just for us to watch what's going on.
Speaker: And I think it's a very common practice in many hospitals that as people are trying to prognosticate, like you mentioned earlier, we wait a little bit longer or there's a tendency to wait longer now after hypothermia.
Speaker: But at one point, it almost seems that nobody can be pronounced as a poor outcome without an EEG.
Speaker: And you did mention what are the findings on the EEG,
Speaker: that are most useful in that perspective.
Speaker: But at what time would you do an EEG again?
Speaker: Well, I would say if I'm getting ready to counsel the family that this seems like it's a very poor prognosis, I might look again for reactivity.
Speaker: Other than that, if I'm not suspicious of seizures, I probably wouldn't look it up again.
Speaker: I think there's a...
Speaker: One useful thing on the horizon, again from David Greer's group, suggesting that a good long-term prognostic study will actually be a quantitative apparent diffusion coefficient measurement on the MRI.
Speaker: And I think he's published one paper looking at this that establishes some thresholds other people need to look at it.
Speaker: But if I were looking toward the future,
Speaker: Where we're going with this, I'd say that the quantitative ADC of the MR will probably be the best single radiographic parameter and something that you could do once you've rewarmed somebody and now you're in the situation of trying to figure out what should I do.
Speaker: Absolutely.
Speaker: And I think that this is obviously an area that has evolved significantly since we started doing hypothermia.
Speaker: And there's a lot still that we don't know and a lot of, I don't want to say controversy, but maybe variability in how people are approaching this.
Speaker: But it is important to be able to give that information to families.
Speaker: And I do think that a point that our audience should take home is the use of EEG and the reactivity in the EEG as being a positive sign
Speaker: potential cortical function as opposed to what we used to look for before which were very omnious or negative signs that pretend of a no chance of recovery.
Speaker: I agree.
Speaker: So I think that this is all been very super interesting and I think be very useful for for our clinicians.
Speaker: We talked a little bit about how to first deal with seizures, what are the first line, second line, and some third line agents and
Speaker: talked about what status really constitutes and how to deal with it.
Speaker: I think this conversation on EEG is a fascinating one, especially outside of the confines of academic centers and neurointensive care units.
Speaker: I think that there's a lot of discussion of whether we should have EEG, who's going to read it.
Speaker: Hopefully, the role of telemedicine and, like you said, these new technologies will make it available to more patients sooner.
Speaker: But I do agree, Tom, I think an important point for our clinicians is that
Speaker: If you don't have capacity to do EEG and read them and you have a patient who you're treating for status, they probably should not stay in that hospital, should go to a place where that can be provided as soon as possible.
Speaker: And I think that's an important point as well.
Speaker: I really want to be respectful of your time.
Speaker: And one of the things that we usually do as we wrap up is we try to ask a couple of questions that really explore the wisdom and other aspects of practicing critical care from our guests.
Speaker: Would that be okay?
Speaker: Absolutely.
Speaker: So the first question, Tom, is what book or books have influenced you the most or what book have you gifted most often to others?
Speaker: Well, clearly the book I have gifted most often is not medical.
Speaker: It's the spy novel Tinker Tailor Soldier Spy by John Le Carre, which is not only a great mystery read but has a
Speaker: all sorts of insights into human behavior that I think are quite prescient and I end up quoting much of the time.
Speaker: And I have not read that, but we'll definitely put it in the show notes and I'll put it on my list.
Speaker: And I think that one of the things that I've learned from asking this question, Tom, is that a lot of us undervalue the amount that we can learn of human behavior from fiction.
Speaker: And I think it's a wonderful venue to, like you said, learn about what moves people and how people behave.
Speaker: So definitely this is something that we'll want to explore.
Speaker: The second question has to do with, is there anything that you believe to be true that most other people don't believe?
Speaker: Well, I've not done a formal poll
Speaker: among my friends and colleagues, but I'm getting more and more skeptical of the ability of randomized trials to answer questions in the intensive care unit because the control group doesn't just sit there and wait for them to either die or recover.
Speaker: In each case, we're trying to do our best for all the other things we do for the patient, even if they're in the, of course, we don't know if they're in the placebo arm.
Speaker: studies for example in subarachnoid hemorrhage looking at drugs like nicartipine have come to the conclusion that the drug is not useful.
Speaker: Well if you look at the quote unquote control group in the two intravenous nicartipine trials in subarachnoid hemorrhage, those patients got lots more rescue therapy than the ones who got nicartipine.
Speaker: So yes there was no difference in outcome but that doesn't mean
Speaker: the drug didn't work.
Speaker: And then if you start to look at almost everything else we do, patients in the fluid and catheter treatment trial, for example, well, yeah, there was a difference in their time to wean from the ventilator, but there were still many differences in the way the patients were treated based simply on what did they need at the moment.
Speaker: Yeah.
Speaker: And I think, go ahead.
Speaker: In that trial, I see in that trial,
Speaker: Mark Mickelson at Penn did a follow-up, a neuropsychological study showing that the patients in the liberal fluid arm had better neuropsychological function a year later, even though we think that it was good to be in the conservative arm and get off the ventilator faster.
Speaker: So I'm starting to think that we are gonna need to have smaller, very focused trials that the populations we're dealing with are really not amenable to large randomized trials.
Speaker: And I think that you mentioned this topic.
Speaker: It's something that I've been thinking about in terms of every year we go to national conferences and
Speaker: either one of two things happens.
Speaker: Either they present negative studies that make you wonder, like you said, or even worse, they present a study that doesn't change anybody's mind because of confirmation bias, right?
Speaker: So people who believe something works, no matter what you show them, believes it still works.
Speaker: And people who didn't believe in it, no matter what you show them, still don't believe.
Speaker: And I think that not only the scientific value of really answering questions, but the ability that trials have
Speaker: the few times they are positive, to change behavior I think is also very interesting.
Speaker: The final question relates, Tom, to what would you want every intensivist who's listening to this podcast to know?
Speaker: So I've been a full-time neurointensivist since about 1989.
Speaker: And when I look at what we did then compared to now and
Speaker: not just the survival but the functional outcome of patients.
Speaker: Patients with devastating neurologic injuries of various sorts really do so much better now than they did back in the 1980s that, you know, it's a miracle to me to see them.
Speaker: When you're going to work every day, you don't see that incremental change because, you know, no day has a breakthrough in it.
Speaker: simply paying better attention to these patients and not assuming that, well, they're devastated, they're going to do poorly, and working hard to try and improve their outcomes is really worthwhile.
Speaker: And I have to say, I've really been proud to see the growth of the field and also the acceptance of neurocritical care principles by intensivists of other fields.
Speaker: We now have very good interactions
Speaker: across the Society of Critical Care Medicine, the American Thoracic Society, Society for Neurosurgical Anesthesia and Critical Care, even the American Academy of Neurologic Surgeons.
Speaker: So some days seem like they're very depressing, but
Speaker: things really are a lot better and they'll continue to get better.
Speaker: And I think that what you mentioned, Tom, which is very, very important, speaks to the fact that we tend to overestimate what we expect on the short term, but turn to underestimate what can happen over a long period of time.
Speaker: And like you said, I mean, from 1989 to 2018, the advances in how we care for these patients and for the field of neurocritical care are really remarkable.
Speaker: And I think that we're going to learn.
Speaker: There's a lot of exciting studies that are ongoing and hopefully we'll continue to learn how to treat the brain because at the end of the day, it's the only organ that really matters, right?
Speaker: Well, if it doesn't work, the others don't matter at all.
Speaker: Okay.
Speaker: Absolutely.
Speaker: Well, Tom, it was a real pleasure talking with you about these topics.
Speaker: I wanted to thank you for sharing your time and your knowledge with us.
Speaker: Hopefully, we'll be able to have you back on the podcast soon.
Speaker: And again, thank you very much for your time.
Speaker: I appreciate it.
Speaker: Thank you for the invitation.
Speaker: Thanks again for listening to Critical Matters.
Speaker: Make sure to subscribe to this podcast on iTunes or Google Play.
