Computer generated transcript

Warning!
The following transcript was generated automatically from the content and has not been checked or corrected manually.

Hello, everyone and welcome to medical education today. Um We are going to be looking at making process e eg work in Perioperative Practice. Um Before I hand you over to Ashok, who is new with us today. I just wanted to let you know about medical education. We are now a year old. We have been running uh this program for a year. We, we started it for those in Sudan, but now we see 100 and 44 countries registering for our events. This event today, I think, oh, has 44 I think, I think I told Alice that we have about 44 countries registered. Um And over the past year, we have ran uh 100 events. Uh some of them have had to be canceled, but we ran 100 events and we've had 17,000 registrations for our events, which is just phenomenal. So I just wanna say a huge thank you for registering for coming along and being part of this program. Um You know, if you didn't come along, we wouldn't put it on basically. So I'm really glad that you come along. Uh and, and, and get this teaching from us. So I am actually going to now introduce you to AK who I'm hoping is going to be helping me a few times with these events and I'll let him introduce you and continue. Ok, thanks. Hi, everyone. Uh Good afternoon. Uh Welcome to the Middle All Education Platform and it's um one of the most relatively complex topics for uh general medical practice is E eg um Today we have doctor Humphreys um will be talking on making processed e eg work in preoperative practice. Um There will be a feedback link towards the end of the session, so please do complete them. So you're putting a certificate of attendance um and the session is recorded and will be available um or to learn again on the middle platform. Um If you do have any questions, I'm sure there will be a few, please pop them in the chat window and this will be answered at the end of the webinar. Um E EG is a very difficult topic and this is the first time that all is um doing this event and hopefully there's much more to come over to doctor Humphrey's now, right? Uh Thank you very much, both. So, er, hello everybody. Um My name's Alice Humphreys. Uh I'm a consultant here in the United Kingdom. Um That's a, er, attending anesthesiologist for those of a uh US type, er, bent. Um, and I'm here to have a chat about a subject of some personal uh, interest, um, which is used of process eg um in per practice. So, a depth of anesthesia monitoring for the anesthetized patient. Um, just before I start, um, I'd just like to declare a couple of, er, er, er, disclaimers. So one, I've got no financial contrast, er, conflicts of interest to declare in any of the, the products we're gonna be discussing. Um, and the views expressed herein are those of myself and the authors of the paper cited. Um And don't necessarily represent uh His Majesty's government, the National Health Service or any other body. So, um my plan for this er, er talk is to go through a couple of bits about process e eg so firstly, a little bit of the background and theory behind it, um we're going to er do a bit of an engineering breakdown and the most commonly used um processed E EG machine, the, the bis um and then um discuss um the use of uh more several of the common er pe eg monitors. Um We'll then break that down further and actually look at the eeg waveform under anesthesia, er, and in particular, look at density spectral arrays. Um And then uh a brief word at the end um about the placement options for electrodes, cos er, if your patient's having er, cranial or facial surgery, um it isn't necessarily straightforward as to where you're gonna put your dots. Um And the, the question is really why why are we talking about this specifically? Um, certainly in the United Kingdom, um is because it's a requirement. Um, the in Association Aist of Great Britain and Ireland and Society for intravenous anesthesia have guidelines for delivering total intravenous anesthesia, um which suggests that all anesthetists er, should be familiar with the principles, interpretation and limitations process, eeg monitoring. Um and the observation of the trace um is likely to improve the clinical utility. Um and that uh we should be using these machines um in er, TIVA cases where we've given the patient a neuromuscular blocking drug. Um And certainly my experience is that um whilst a lot of practitioners are using it, um they aren't necessarily able to um understand um some of the more nuanced bits of interpretation and particularly the limitations of the machine. Um There's several P EG monitors on the market at the moment. Um probably the, the most familiar and most commonly used. One, certainly the most cited is Bispectral index um which um was originally developed by a firm called Aspect, er, which is in turn bought by Cian, er, which is in turn by I bought Medtronic, um which is the current uh manufacturer. Um and they also license other modules, er, from a couple of other manufacturers, um, less commonly used certainly in the U K's Narcotrend um E Entropy and Z line. Um And then you have the, er, the kind which is normally seen as a purview entirely of ee nerves which is raw, um which is, er, produced by multiple manufacturers. Um And I like raw very much for reasons we'll get into later. Um So we're gonna go into bis in some engineering depth and I make no apologies for that um for several reasons. One, it was the first to market. Um So it was launched back in 1994 by aspects before the subsequent acquisitions that we've already discussed. Um And the idea is that it takes the incoming ee wave form deconstructs it via a fourier analysis. Um And then, um using a proprietary algorithm um produces a dimension this number between zero and 100. Um And, er, certainly in the UK, the nice, er, guidance DG six, recommends its use in general anesthesia. Um And the idea if you read the brochure, um is that the best, best value should, er, er, give you an idea of the depth of anesthesia of the patient. So 100 is hopefully all of us here today. Um, fully awake. Um 80 to 100 is moderate sedation. So conscious sedation, 60 to 80 is deep sedation. Uh 40 to 60 is allegedly, er, the deep hypnotic state that we're after for general anesthesia. Um 20 to 40 is you've overdone it a little bit. Um, and 0 to 20 is you've overdone it a lot. Uh And what you've done is given the patient a coma rather than anesthetic. Um the trouble is, this sounds very appealing but it's probably a little bit more complicated than that. Um And a question I often ask uh Juniors when I'm teaching um is what is consciousness um because there isn't really a particularly decent answer to that and philosophers, er, clinicians and theologians have been thinking about this for some time. Um And it's not just um a nice thought experiment because if we don't know what consciousness is, then we don't know what unconsciousness is. And if we don't know what unconsciousness is, then we don't know what anesthesia is, which does rather imply that we don't know what we're doing, um which is a slightly sobering thought um over the centuries, various few points um have been suggested um as to what consciousness might be. Um So in the middle ages, the thought that there was the ones inner life, the, the world of introspection of private thoughts, er, imagination and, and volition. Um these days, um it tends to include any kind of cognition, experience, feeling or perception that is our, our interaction with the world around us. Um And certainly, if you're into science fiction, it might be awareness, awareness of awareness, self awareness sentience, whether that's continuously changing or not. Um And the fact that um all of these theories have been brought up, none of them have improved um is, is er, of great interest to me personally and I'll happily have a, a chat over a, a coffee or a beer with anyone who wants to discuss that further. Um, certainly I think des cartes, er, put it most pithily when he said it was, er, um, Coto Ergo, some. Um, that is, I think, therefore I am or of course he never wrote it in Latin. He was French. Um, and point is that we know what it looks like on an E EG. Um, so this is a full 1012 E eg that a neurologist might use. Um And we can see um on this, we have high levels of alpha beta and gamma activity. Um all of which we'll discuss a little bit at, at length later um with those large deflections being blinks. Um and that's what er, any E eg would look like if put it on anyone who's awake. Um and er, in interpreting the world around them, um problem for clinical purposes is it looks exactly the same as rem sleep um with the er, added extra deflections of the rapid eye movement. So this is someone who's dreaming. Um And you can see that we have all these deflections, they're more numerous because the patients in rem, but they're the same sort of deflections as you'd see um from the blinks. Um and we still have alpha beta and gamma. And so strictly speaking, if you put an E eg A, a processed E EG machine on someone who's dreaming, it will say they're awake, which is a thought, isn't it? Um, and I think that's really interesting because, er, as a fan of science fiction myself, um, anyone who's watched the Matrix knows that cipher, er, would say that he knows the steak doesn't exist, he knows that when he puts it in his mouth, the matrix is telling his brain that it is juicy and delicious. And after nine years he realizes that ignorance is bliss and there in lies the problem, if we don't even know what dreaming looks like or rather the processed E EG machine doesn't look, look like, doesn't know what dreaming looks like. How could we really hope or to trust it um to tell us how awake our patient is. Um, and certainly they're prone to error. Um So this is a photo I took of a patient I had myself. Um if you just go off the best machine reading, uh it says 80 that means as we discussed before, but the patient's mildly sedated and if they're having major surgery, they're probably about to get up off the table. Um However, er, this patient was too deep. I actually lowered the anesthesia or, or seeing this reading. Um And what this interference is is diathermy from the surgeon's instruments, um which the para e argos is misinterpreting as cranial activity. Um And you know, if you over anesthetize a patient that certainly has implications for um prolonged wake up time at the end of the case which has issues for the efficiency. Um, but also increases their chances of postoperative cognitive decline or postoperative delirium. Conversely, um, many folks might have seen the, er, the famous study, um, where, um, her 10 of volunteers, I think they were all medical students um, in Australia, er, were given muscle relaxant and muscle relaxant only. Um, and then using the isolated forearm technique, the researchers were able to demonstrate them doing, er, er, arithmetic, they asked them to do sums and give the answer o on a hand. Um and the, er, the bi machine er, said that they were deeply anesthetized so, er, it can overread out, underread quite spectacularly, um which is a problem and, and this kind of comes to the nub, this is a proprietary black box system. Um It's always been known that it involve fourier transformations, but up until recently, we didn't actually know what it was that it did. Um It's always been controversial, um be aware trial and the unaware trials in the Lancet and New Engine Journal suggested that they might be useful for reducing accidental awareness under anesthesia, but that's an exceptionally rare outcome. Um and trying to work out rates of er, rarity of things um is pro quite a problem when you have such a small incidence. Cochrane reviews have suggested that they might be useful in the itu population to prevent over anesthesia in the periotic population to um help with uh depth of anesthesia monitoring and some benefits to postoperative co to decline and delirium. Um, but that's a lot of ifs buts and maybes, uh, and these things aren't cheap. So, um, obviously, er, hospital trusts will get block purchase discounts. Um, but the only published figure I could find on the internet after a fairly extensive search was that the bis strips which are single patient use are about $45 a pop. Um, which is, er, you know, in the grand scheme of things, not much, but soon adds up if you're having to use it on for every team of patient. Um Fortunately, it's not a, a black box anymore um because it's been broken down from an engineering point of view. Um And this is this um this is the work of AAA mit professor er called Chris Connor um who got hold of er a bis machine, extracted the machine code um from its chips, um reverse engineered the algorithm um as to what it is that er bis is doing. Um and this product which the manufacturers have been riding on quite happily er, since 1994 charging large amounts of money for is 64 lines of MATLAB code um which has a bug in it incidentally. Um And this, this is all it is. Um and I don't know about you but I'm not entirely certain that 64 lines of code um can er interpret the state of consciousness of the human being or indeed um do a better job than a trained physician. Um You just gave us some interesting things. One, it's not bispectral, it's monos spectral two, it's not one algorithm, it's several, three at least. Um which explains the lag, we see when we're inducing and, and waking patients up with massive drops in between um the various phases. Um And as I mentioned, there's a bug in it. Um So the way it calculates spectral edge frequency, which again, we'll go into with some detail later. Um means that if the pa the E eg is entirely flat for three minutes, um the machine will read an S EF of 30 Hertz um which would mean the patient's awake. Um So that's er again, got some fairly significant implications for the of anesthesia monitoring. Um So, the way it works is this, it divides the eeg waveform. So the density spectral array into various bands and what it does is then compare them. So for sedation, er which is one of the algorithms. It's comparing the very high band with the midband for general anesthesia. It's comparing the high band with the midband. Um And then for the burst suppression algorithm. So this is extremely deep anesthesia. It's looking at the low band and those are the different frequencies that we see on this Nancy spectral array here. Um for the low band, it's particularly um weighting er the proportion of flatline E EG um which is probably more reliable than the other two. but, er, yeah, it's curiously highly weighted towards gamma, um, which, er, is, as we've seen, associated with conscious, er, uh awareness of the world around us. Um So you can kind of see what they're going for here, but all it's doing is comparing strips which probably explains um why, um, it's able to come come up with such erroneous numbers because if any of those have interference from an external source that's gonna massively change the algorithm. Um it also demonstrates what that lag that we mentioned earlier. So when you initially anesthetize a patient, it will be going through the sedation algorithm looking at this band and then it'll only take, it takes a good 30 seconds for it to realize that the patient's deeper than that and swapped to the general anesthesia one. So that's why you get uh the waveform changing with no concurrent change in the best number and then suddenly you get a massive drop all of a sudden um which, er, clearly doesn't represent reality. Um So it's looked at these bands that we, as we've discussed. Um So general sedation and burst depression, um the burst suppression bit is looking at er, the amount of the eeg which is flat. So isoelectric as in burst suppressed. Um and um the proportion of that is given as the burst suppression ratio, um which is outputted as a separate number. Um The three scores are mixed and in general whoever's lowest wins, um, which is why we get those, those big drops, um, that's fairly detailed go through. Um, you don't need to have completely understood that to be able to utilize the rest of the lecture. Um, but if this is something you're interested in, I would strongly recommend you to, uh, take a look at, er, Connor's papers. Cos, er, the way he did it was an absolute wonderful demonstration of top quality Nary. Um And er, yeah, that's, er, that's er, some really interesting paperwork. Um That's just uh another way to kind of illustrate the, the burst suppression ratio. So, um if it's less than 40% it doesn't change the best number. Uh thereafter, it's a linear increase down to 100 which would be bad, don't have a birth suppression rate of 100. Um This is a demonstration from a said line of the directional delay. Um just the sheer um difference um between when it was swapping between the three algorithms. So we hear we see on induction um there's a, a decent, er 12th delay from awake to the, the general anesthesia one, but then on emergence, it's much bigger on the, the order of 20 seconds um from a burst suppression rate um which is very high to a decent general anesthesia uh is quite slow as well. Um But what's really astonishing is the, the difference when you're em emerging the patient from deeply anesthetized to fully awake. It takes ages for the numbers to correct. And that's, that is a directional problem. It takes, uh, huge amounts of time, um, to, er, to correct those. So now we've kind of looked a little bit about some of the limitations of these machines. Uh, what I hope to do is go through each of the main ones in turn, um, and kind of break down how you might be able to use them more efficiently um and more effectively in your clinical practice because just going off those numbers as we've seen has delayed, prone to in to er er interference from ex external sources. Um and doesn't really tell you everything that's going on with the patient. I don't know about everyone else but I personally like uh my patients er status to be presented to me, not hidden behind an algorithm. Um So anyone that's used bis er whether an external module or er integrated one will probably recognize um what or the other of these demonstrations of the display, um the integrated modules have less room on them, but they ultimately are showing the same data which we'll now go through. So the first one that a bi machine will generate a single quality index. Um So out of 100 ideal er is er greater than 80. So a higher number is better with the inequality index. And it's, as it's describes, it's the, the quality of the signal that the machine's, er, getting um, below 50 is when it will give the, er, er message that, er, it's got poor signal quality and below 20 it won't calculate, er, a, a bi score at all. Um, it tends to fall as the E MG rises. Um, so, um you can kind of use it as a, uh AAA poor man's um e er, um, monitor of muscle relaxant. Um, but, uh, that's certainly not how they advertise it in the brochure. Um, and it's affected by lots of things. So, ambient electromagnetic fields, surgical diathermy, the electrodes falling off or you're putting them in the wrong place. Um, so higher, the better for SQ I for E MG. Um, it's again 0 to 100 it uh shows muscle activity. Um, you want to target an E MG of less than 30 an increase in it. Um, even with a low bis score is supposed to indicate a lack of analgesia, which is probably the, the, the hub of the, um, no suction monitors that are now coming onto the market. Um, which is another topic of mine, uh, which uh I'd be happy to, uh, discuss at some length. Um, a, a fully paralyzed patient shouldn't have an EMG over 30. Um, uh, but the thing is, it's unrelated to level of consciousness. Um, so you don't need to think that if the E MG is going up, you need to over, uh you need to give the patient more anesthetic. It's just muscle relaxant. Um And for what it's worth on the Bistra, it's gen generally derived by electrode four. The suppression ratio is one of the more important ones. Um It's the proportion of eeg suppression within any given minute. Um And that's a maximum of 100. So when we over anesthetize a patient, we start to get flat isoelectric segments on our E eg um what that's called is burst suppression and it is a he uh it is an unhealthy state to be in what you generally get is a bit of an isoelectricity and then a burst of activity that's a healthy response to an unhealthy brain state. Um But really if you're seeing any isoelectricity, um then you've overdone it and you need to turn the anesthetic down. Um The suppression ratio should be zero at all times if it's more than zero, turn your anesthetic down and that applies equally to volatile as TVA. Um and also be very, very wary about misinterpreted E CG signals. So this is a photo, a colleague of mine sent round the department asking how they get rid of the E CG er from their bis er bis line er to which the answer is of course, turn down the anesthetic. So basically this patient was so over anesthetized that the bis was picking up their E CG instead of their E EG. Um So if you see some of the books like an E CG. It probably is and please turn your anesthetic down. Um spectral edge frequency is probably the the most useful number. But a bis generates, ironically, um it's um the frequency below which X percent of the E eg energy is found given in huts. Um S ef 90 or 95 normal and indeed by will generate you a 95. So frequency below which 95% of the E eg energy is found. Um And various studies have done have suggested that uh age 13 is probably the optimal spectral edge frequency to achieve adequate general anesthesia. Um I would urge you if you're uh when you're recording, if you've only got uh a processed eeg as well as wri writing down the output of that machine, please write down the spectral edge frequency as well. Um It's a really useful um uh marker of the state of the patient. Um And um although it's dependent on multiple factors including age, the agent you've got um and any opiates you've given um 10 to 15 Hertz seems to be a fairly pragmatic compromise uh in, in the field. Um And you'll, you'll note the S ef changing much more responsively than the bis does. Um cos it's AAA real time calculation. Um So just uh to, to go over those er again for summary. So if you're using this, um this is probably the start to take a screenshot of um, cos it's, er, got, er, a few take homes. So, um, the number itself 60 to 80 for 40 to 60 for anesthesia. Um E MG, it's a marker of muscle relaxation only. Um, please don't over interpret that. Aim for a spectral edge of 10 to 15. Um soon in quality index, the higher the better and suppression ratio should always be zero. Ok. Um Ensure that you've got the waveform displayed on your screen. Um because er, as we are gonna go over in uh later on in this er chat, it's er extremely useful clinically and tells you a lot about the patient. Um you can optimize the scale of that. Um So for Children, you want to change the scale to 100 to 200 microvolt um cos they've got such powerful brain activity and thinner skulls, uh 50 to 100 for adults and 25 to 50 microvolt for the elderly. Um turn this pressure ratio on and the filter off. Um But when you turn the filter on what you do is filter out the delta waves which again as we'll go over er, later on is the most important part of it. Um And for what it's worth if you're using e entropy, um these tips are the same for that. Um So these changes are, are good to help your interpretation. Um Narcotrend um uses a completely different er system. Um So it uses a classification of A two F as well as a numeric index. Um, the A two F classification is based on a system called Kugler, um, which was published by a West German anesthetist in the 19 eighties, um, which is for, it's worth ha, what I used to interpret my raw E eg and record it. Um, if you're using narcotrend E is the target that you want for general anesthesia. Um, and, er, if you're getting down into er deep FS, then again, you've probably er overdone it a little bit. Um B EG index is again a number from naught to 100. It's another proprietary algorithm. It's not been reversed, engineered, but it's probably quite similar to the bis. Um So look at the waveform, er, and look at the, the index letter um because we, we know how that works, what it does. Um And it's been shown to work nicely. Um It will give you a single channel waveform and it will give you a density spectral array. Um again, both of which will go over later. Um And um yeah, it's, it's quite, quite elegant um be getting similar numbers to, to a best one. So this is Kogler, um which rates the patient's er state of anesthesia from A to F, as we've mentioned. Um And what it's doing is looking at the, rather than comparing one band against another, what it's doing is the same fourier transformation, but then assigning a waiting to how much of a particular waveform there is comprising the waveform. Um, so a zero is fully alert. Hopefully everyone that's here, um, which is normal, we itself beer and gamma. Um, if you're starting to find my voice is a little bit dull and are interesting, you might be slipping into a one or two which is subvigilance, um, at which point the, er, alphas become more frontal, um, and become lower in amplitude and again, we'll go through all the various waveforms again. Um, as you become drowsy, um you start to get into, er, er, feeders. Um, and then as you go off to sleep you start to get deltas which are the nice, slow, gentle waves that we see um, on a patient that's, er, that's asleep. Um, and then depending on how much of the waveform energy is that is delta will determine whether as you go through CD and E. Um, and, er, yeah, I tend to aim for D2 to E one, for my patients certainly. If they're having something milder done, I'll be aiming more in the DS. Whereas if they're having, uh, say a um cerebral aneurysm clip, then you're gonna be wanting them on um E one to E two. you can take the things according to that, um, beyond E two, you're into an anesthetic coma. Er, so down into f, which is the point at which you start to get birth suppression and for what it's worth if you put uh an E EG on someone that's going into normal physiological sleep, um you then get rem um we don't see that in any size patients because anesthetic patients don't enter rem sleep. So for e entropy, if anyone uses this particular one, what we think it does is it measures irregularity in spontaneous brain and facial muscular activity. Reason to that as you will um again, it's a proprietary um E EG and E MG processing algorithm which produces the fast reacting er response entropy and the steady state state entropy. But then they noticed the big lag between um stages with some of their competitors and kind of tried to change it so that it was a bit quicker on the uptake. Um And what they're looking at is irregularity uh and variation in the way you form rather than steady state, which kind of makes sense, I guess. Um as, as the signals become more um regular and er less va variable, um the entropy will start to go down um which indicates suppression, which we might suspect given the fact that we get start to get isoelectricity if we overdo it, um then er it produces a number same as the others. Um again, 0 to 100 for the res er response entropy. Um but from 0 to 91 rather oddly for the steady state entropy, I guess because um if you're fully awake, then you wouldn't be reaching a steady state. So it'd be able to calculate it. Um, again, um, 40 to 60 is what the manufacturer recommends as a target. Um, and, er, the values near 40 will indicate a low probability of consciousness according to manufacturer. Um, again, it will give you, er, spectra frequency in a wave form. So make sure to have those turned on. Um, and, er, be happy with looking at the, uh, the wiggly line cos it'll tell you a lot more than the numbers will, er, sline pretty similar um to the others really, except for the fact that they give you four channels for E eg. Um again, you get a proprietary index, NATT to 100 they recommend 25 to 50 rather than 40 to 60. So slightly different. But you also get density spectral array, which is the, the colored er graph here, which will go through E MG birth suppression ratio and it will also uh give you er, indications of artifact. Um So that's similar to the see equality index we've seen on bis. So that's the four manufacturers. And er, if anyone wants are happy to go back over any of those particular ones in the Q and A, if any or if anyone uses any other machines that I've not gone through, then uh I'm happy to have a chat afterwards. Um So when we're using these machines as they're intended to be used with full understanding of their advantages, disadvantages and limitations. What's the utility of them? Um Well, they possibly reduced the rates of an awareness under general anesthesia. Um They're a good marker of trends. Um Certainly they're better over time than they are at rapid changes. They're a very useful marker of intrinsic brain vulnerability. Um So I tell about a case. I si did a, a one night shift where a lady had a, a one of those chronic on chronic subdural hemorrhages, um which the neurosurgeon was adamant needed to be done uh out of hours. Um The patient actually had a G CS of sort of 13 to 14. So the 13 weren't convinced but um we did go to theater um and I popped, um, it was the best chip onto the patient, gave her 100 mcg of fentaNYL and it dropped from er, 100 to about 10, um just on a bit of fentaNYL. Um which I think was an incredibly useful marker of how vulnerable her brain was. Um I ended up doing by a hemispheric. Um and it was quite interesting because the good half had much better numbers than the bad. Um And as the surgeon was hoovering out the um uh crude oil from within her skull, the bad one came up to meet the good one. So I was able to tell him before he even knew that it was working, that it was working. So that, that was quite handy. Um And the car to that is that probably the most regular use of pe that I do in my clinical practice is to make sure that the patient isn't too deeply anesthetized. Um It avoids delirium uh and leads to a clearer and more rapid wake up. Um And that's got several benefits. One from a point of view of theater efficiency, but also particularly in the field of new or anesthesia. It means the patient's able to wake up and answer questions that we can assess a neurological state faster um than they might otherwise do if we'd really overdone it. And you know, mercenary, it's a number to write down to the veins against claims of awareness, isn't it? Um shows you were doing something to make sure the patient wasn't the weight when they shouldn't be. So, um I would write it down. But again, if you were using propriety algorithms to write down that and the spectral edge, the reason it's useful in um TIVA in particular is that we're all encouraged these days to use target controlled infusions. So TCI pumps, um but all of those algorithms are models. Um all models are by definition wrong. Um And so all we can really say that the algorithm TC algorithms do is provide us with a gain switch. Um The real concentration of the agent in the patient's plasma is irrelevant for the titration because everyone responds to it differently. Um And it's not the blood that we're monitoring anyway, it's the effect site cos it's the site of effect. Um Even if we could measure the concentration in the brain, all patients from differently to different concentrations. Um And so that's not very helpful for monitoring the patient's state. Um When we're inducing, we need to be able to measure that without delay. Um Because if we can monitor the patient's depth of anesthesia and induction, and we can note the point of induction of anesthesia. And we can look to our TCI pump, see what number it's reading. And that allows us to calibrate across. Um But this is 15 to 30 seconds behind reality, we can't measure the ce and the scient of peak effect influence it. It, it, it depends on the TCI bolus varies between models and in particular, um I has implications for the elderly. And so looking at the waveform, the raw E eg is the only immediate measure of the patient's brain state. Um because all the algorithms are behind. And so what I like to do is I'll induce the patient, look for the signs that will now go over, monitor the point of in of induction. And then I can note what TCI values were being read and you can then calibrate where you are thereafter. Um Because otherwise the numbers are useless, they um all they are is are naught to 11 dial, which isn't very helpful in clinical practice. So, uh I know you've all been er, very much looking forward to this. So it's the er bit you've all come here for, we're gonna look at E eg waveforms under anesthesia now and how to interpret them. Um And we do this by dividing er, the waveform into spectral frequency bands. Um These are er from deep to awake, slow delta theta alpha beta gamma. Um there's annoying split in the middle um because when E EG was invented, er, the first pros were put on, on, er, a chap er, with his eyes closed and the first waves that were seen were named alphas. Um, he then opened his eyes and the waves he saw were then beaters and then he got really excited and the waves that came to the fall were gammas. And then as you drift off to see you go down through your theta deltas and finally into slow waves, um, when we're into uh deep anesthesia, um, and you see these different waves become present as you progress through deepening sedation. So if a patient is awake or lightly anesthetized, um, you start to get more beater and alpha. Um, as you become more deeply anesthetized the, er, er beta, um, to continue to decrease you get more alpha and delta and for what it's worth, if you, her full 1020 set on the alphas would become much more anterior. Um, as you become then deeper, um, the, you get more deltas and burst suppression and uh, deep coma, you get completely flat activity. So what you're wanting is for the intermediate ones. So we have very little beta but alpha and delta and that's the classic pattern that you see in propofol anesthesia. So, these are the stages you'll see as you induce a patient going past on your monitor, on your eeg waveform. So when you're awake with eyes open, you'll have beta and gamma oscillations, you can see they're very high frequency, very low amplitude. As you begin to anesthetize the patient, you'll get something called er paradoxical aci excitation where you get big beta oscillations. Um and those er look like this one here. Um as they enter the point of sedation, you'll start to get alpha and beta oscillations coming through very nicely like this. And then, um as they become un er unconscious, so generally anesthetized, you'll get these slow oscillations, um which you will see and we'll see in a moment on a video these then kind of broaden out to become the nice slow delta and alpha oscillations, which we see for unconsciousness at a surgical level which then flatten out to become burst suppressed and then eventually you'll get an isoelectric E eg and this is the point in which you start ce CG coming through um for what it's worth um theoretically, this is the point at which you could then intubate the patient, you probably find your muscle relaxant and have them worn on. Yeah. So I wouldn't do that. Um, but theoretically that's, that's the point at which you're good to go. So, hopefully the tech works and we're going to demonstrate beautifully going through all of those phases on an actual case. So, um, we've got a patient, um, who this lady was, er, in her 77 um, undergoing um, general procedure. We do what we have are the frequency beaters and gammas and then big reflections of the blinks that for, um, in the graph at the start. Um, uh nicely now. So, um, you can see she's closed your eyes. Um, I think at this point, we've given her an opiate, um, she's blinking a lot less with the peter low amplitude, very high frequency for is worth the, this is actually probably correct at this point. Uh as is the spectral edge frequency, which I don't know if you can see that, but it's, er, in the low twenties, right? So we've started the pumps now, as I recall. Um, and you can see we're starting to get some changes in the waveform. So we're getting a big bit of blinking activity there. Um, and then she's starting to breathe a little bit less as she becomes anesthetized, the opiates are wearing on. Um, and she's starting to drift off to sleep. Um, what you're starting to see is, um, the beaters are starting to disappear off. Slight, we're starting to get breakthroughs of alpha activity coming through completely, but um to nod off, there's a nice little bit of beer there. You can see it starting to come through very high gamma time to disappear off next to the first. Um I said, what we're starting to see is the um uh alpha activity coming to the full. Um And this is the deep at the point of induction. So this patient is generally anesthetized. You notice the bis says uh she's wide awake. Um The spectral edge is dropping um not completely and, but this is still saying she's awake. So those deep slow OS solutions are starting to broaden out a little bit. Now, you can see we're getting the deltas coming through, the spectral edge is now reading correctly. So we've got er quite deep. Now, um Bis will say she's awake, she's not. Um but we've got the delta activity. So the larger waves that are underlying the smaller ones. Um And uh as I recall, this is the point at which we um intubated of the patient. So you can see here the this is do m er is starting to recorrect you now, but she's been anesthetized for a good while uh spectra edge is five, which is probably correct. Um So we did overdo it a little bit in this particular instance, but hopefully you'll be able to see the different waveforms quite nicely um for the E eg spectrography then. Um So this is um what we see on a density spectral array. Um What we're looking at is the unprocessed waveform is then pulled down into different power segments. And then these are displayed um on, on as a colored graph. So the hotter the color, the more activity um there is um and then that's really er really useful um for visualizing the pro the progress of the, the anesthetic over time. Um They look different for each agent. Um So Propofol has this classic delta alpha um frequency split that we saw that we discussed earlier. So there's a lot of activity in the delta range and some in the alpha with a bit of a split between some of the others in between times. Um But all the agents have a particular one. So Propofol is probably the one that we would see most because we generally mostly use E eg um for TA. Um But certainly, if you've got a patient who's sick or rather elderly, I would encourage you to put on some ee eg if you're using volatile as well, um which you'll see um this particular waveform placea flaring. So this tends to be more delta and fetter. Um So you get uh the same uh tramline split, but it's lower down um for Ketamine, um which is a different er mode of anesthesia altogether. So, probes there, ultimately, both go for Gabaa Ketamine as we know, go through an MDA. Um And so you get a very different waveform with energies. Um more up towards, er, um, beta and delta as well. Dexmedetomidine is alpha two, inhibition and you just get delta's, um, it's very straightforward. Um, each volatile does have its own one but ultimately they're, they're similar, um, er, er, tramline type pattern. Um, and the point is that what you wanted to see is delta activity among any of the agents. And if you've got that delta activity, um, then, um, you're probably uh good to go. Um And just to illustrate how that's actually different, um We can see um in ace of fluorane er maintenance that you don't have the um alpha activity, what you've got instead is some lower frequency um fter and the waveform looks subtly different from how it did for um the prib form. Um and then just a reminder o of Kogler. So when you're looking at these waveforms, um certainly, if you're using raw as well as the waveform, you'll get um outputs of the frequency percentage of each of those bits um for er, er which particular waveform segment is most powerful, but you can get a general idea of where you are just by looking at the raw waveform. Um And I think that's, that's really handy. So, um Kogler Kogler is where it's at for recording um raw um just a little quicker, little bit about electro replacement. Um This is mostly for folks doing neuro or Ent Max Vax. Um because the traditional places to put your E eg dots aren't necessarily accessible during that kind of surgery. Um, but the thing is particularly for the processed ones, the proprietary, er, the, the manufacturers say you have to put it in a particular place. Um, but as we can see from the system used by the neurologist, so the 1020 system, there's all sorts of bits of the brain that you can, uh take in with the electrodes and largely for our purposes, we're not looking at particular areas to highlight issues. What we wanna be able to do is just look at the brain as a whole. And so as long as you've got sufficient amount of brain tissue in between the electrodes, you're probably gonna be OK. Um If you've got a multi channel system, um so, er, s line or bifrontal bit or something like that, obviously, that's gonna be a problem because the er, er electrodes are shaped in a particular way. Um But there have been various electrode placement positions described in the literature, most common ones, probably auricular around the ear, um which is very useful for bifrontal cranial surgery. Cos you're not in the surgeon's way. Um For what it's worth, it's much easier to do these other positions with E CG dots um which you can use with raw E EG or with narcotrend um or with needle electrodes um rather than the strips, but it can be done as demonstrated here. So the strip will go very nicely around the year and that seems to work pretty well. Um You can also put it along the mandible. Um, this paper compared readings from the forehead with the mandible, uh and found that they were broadly similar. Um, you could put them nasally, um particularly with the bifrontal one. So that's slightly less in the way of the surgeon than it might otherwise be, even if you can't put it any other particular place, um you could put it under the eye. So infraorbital um that seems to work quite nicely too um above the lip supralabial um or bilateral. Um So certainly if you're doing narrow and you want to look at both hemispheres separately, um then putting on bilateral ones is er is a, is definitely something you can fairly easily do. So, um in summary, then, um hopefully, what we've done is go through a bit of P EG background of theory. Um We've done an engineering breakdown of the B system um gone over a little bit about the common monitors. Um We've looked at the waveform and density spectral array and how that might be useful and a very brief can through the electrode placement options. Um take home messages, um treat the algorithms with suspicion, know the uses know the limitations, treat the patient in the round, understand the other generated values. So not just the preparatory algorithm number but also the spectral edge frequency diverse depression ratio and whatever indicator of artifact or interference that the system has optimize the monitor based on the patient that's in front of you. Um And understand what E eg waveform and density spectra you should see because if you're not seeing that, then you've probably got an issue. Um If this is something that interests you, um these are some quite useful resources um which are freely available. Um So the first two are two different e courses which are er will take you through interpretation of E eg um uh for the purposes of anesthesia. Um The third one is a paper in anesthesiology by Pin et al um which go which quite a lot of um my understanding of the wave forms came from. Um And I would urge you to go through that. Uh These are my references. I don't expect you to be able to read those and I'd be very happy to take any questions anyone has. Um Thank you very much for your time. Um And then I think I need to do that and we're good a shock over to you. This one question in the uh they say I've just flack it up on to you. OK. Uh Is the pre anesthesia evaluation for processed e eg similar to traditional simple? Um No, so um these particular systems are only really, really used in theater. Um You wouldn't expect to, you know, you wouldn't send a patient home with uh a monitor on to any to see what their normal physiological sleep looks like because it's not what you'd see in clinical practice. Um This is the kind of thing that really needs looking at at the time um and needs real time interpretation and this is why it's useful to be able to look at the waveform and interpret it yourself. Um Because you can see from that a little bit about the patient's brain fold ability and their state, um which the algorithms might not necessarily um uh interpret for you. Thank you. Any more questions. I stunned everyone into a steeper that's normally how it goes. No, I think they're just like, oh, that was too much information. There you go. There's a question there. I'll hand back up. OK. What have we got? Huh? That's a very good question. Um So I think, um uh thank you by, by the way. Um So how would you mean if they disagree with it as in, they disagree with you using it or they disagree with the number? Um Because it's not something they can really dispute at the time. Um I think the utility of it is that if you have a claim of accidental awareness under anesthesia, um post facto, then when you present your paperwork to the relevant authorities by writing down or otherwise recording the numbers, what you've shown is that you're a competent attentive anesthetist who's bothered to use the depth of anesthesia monitoring as per the guidelines Um, and so if there's still been an incidence of awareness, then you've been following best practice um, and done the best you can to try and minimize the risk of that. Does that answer your question? Oh. so the lady I mentioned is uh, so the, oh, um, so me, the, the numbers were probably about right for the waveform. I was seeing the reason they were so, er, er, different from what I expected was because her brain was so sick from the subdural hemorrhage. So she only needed very little amount of anesthesia um to then reach the state of general anesthesia if that makes sense. Um um For Eric um increasingly actually, um I think for status epilepticus or things like that, generally, the neuro neurology folks would come and do a formal 1020. Um but certainly for depth of sedation monitoring then yes, it is starting to come in a bit more um in the itu setting and, and the um Cochrane review suggested that it was useful for that. Um We would occasionally attempt to interpret it ourselves. Uh A bit of roar if the patient was in status, but in generally, I'd defer to the neurologists on that um rather than trying to interpret that myself. Um Amela. Um so the waveform shapes will change as you go through. Um And if you look at them as you anesthetize your patients, next time you'll see that character change, you get the different, um, er, shapes, um, coming through. Um, so BBBB, you will see these different patterns if I share that screen again. And right, so the wave forms look different for the particular, um, segment. So the wave type, so the alphabet, gamma, et cetera, um, you don't need to necessarily be able to tell where one ends and one or another begins. Certainly, I wouldn't, um, attempt to interpret the f proportion unless the machine was doing it. For me. As in, if you're using raw, even a raw one will then generate the amount of energy that each waveform type produces, which you can then interpret yourself rather than having a proprietary algorithm do it. But it's the, the machine is doing a fourier transform on the wave. What I'm hoping to encourage is you to have a vague idea of the general shape. Um So that, you know, that that's vaguely what you're looking for if that makes sense. Um Does that answer the question? I think that's that, I think it's a moment of waiting, isn't it? It's just like Mhm That moment of everyone doing this. Oh, quick, quick enough. Great, lovely answer to a question. Beautiful, brilliant. Any other questions out there as most of you know, I am not medical, that's why s on this call as well so that he can answer medical questions? Are there any other questions out there that we can put to Alice? I'll give you a couple more minutes to type away. Um As always, at the end of this event, your feedback form will be in your inbox. OK. Um Please fill it out. We really do want really good feedback. We'll be passing the feedback on to Alice. All right. Mel says excellent talk. Thank you. Thank you, Mel. Oh, you're mute. Alice. Do you know Mel? Alice? II know a couple of males. Are you, are you a Southwest based male? L? She must have, I used to be such a small world, right? Ok. Well, if we're done, I'll actually, I'll hand over to Asher cos he's, he's, he's hosting this one. Er Yes, metil. Er this talk will be shared to the colleagues. It'll be on meal platform um and this can be accessed once you complete the er feedback form. Um Thank you, Doctor Humphreys er for just opening a new file in our brain with regards to the expensive um process E eg um And thanks everyone for attending, please do complete the feedback form. Uh That will be um passed on to doctor Humphreys and if there's any further teaching that Doctor Humphreys could be um could maybe cover or even it's totally unrelated with regards to anesthesia, you can add this to the feedback form and we will be reviewing it. Aim to plan our upcoming sessions. Um uh Yeah, Estonia says the, the part where you showed the video is excellent. Bra thanks, sadly, that's the one bit of it, you probably can't share and I'm gonna ask the platform to expunge that from the video. Um I'm allowed to use it in presentations but not share onwards. So um er yeah, um um I've just sh shared the um feedback form um on the chat window. Um It will also be sent by email. Um Thank you so much. Thanks everyone, looking forward for doctor S for the next session. Thanks very much. Bye everybody. Thanks for coming.