Ep 21 – Target-Controlled Infusion (TCI) Models: Marsh, Schneider & Eleveld

What are the differences between the Marsh, Schneider, and Eleveld TCI models used for propofol delivery in anaesthesia?

The Marsh, Schneider, and Eleveld models are pharmacokinetic frameworks used in target-controlled infusion (TCI) systems to deliver propofol in anaesthesia. The Marsh model uses a weight-based model without age adjustment. The Schneider model adjusts for lean body mass and includes effect-site targeting. The Eleveld model integrates data across age groups and incorporates covariates such as weight, age, and sex, offering broad applicability and improved accuracy in diverse populations. These models are crucial in anaesthetic pharmacology and form a core topic in FRCA exam preparation.

What Is TCI?

Target-Controlled Infusion (TCI) delivers intravenous hypnotics (like propofol) using pharmacokinetic (PK) and sometimes pharmacodynamic (PD) models to maintain a constant plasma or effect-site concentration.

The TCI system:

• Accepts patient demographics (age, weight, height).
• Uses a PK model to calculate drug distribution and clearance.
• Delivers an infusion rate to achieve a desired plasma or effect-site concentration.
• Uses bolus + maintenance infusion, sometimes applying "overpressure" to rapidly achieve the effect site target.

⠀TCI allows more precise control than manual infusion, but it’s not flawless. The model is only as good as its assumptions—and the data it's based on.

Why TCI Models Matter

Clinical variation in:

• Age
• Weight interpretation (particularly obesity)
• Body composition (lean vs fat mass)
• Paediatric vs elderly physiology

⠀...means that a “one-size-fits-all” model doesn’t cut it. Each TCI model has different assumptions, compartments, and equations. The result? Possibilities of different dosing behaviour in practice.

Comparing the Big Three: Marsh, Schneider, and Eleveld

Marsh Model (1991)

Core Idea: Based on healthy adults and scaled to body weight. No effect-site targeting initially. Simpler and front-loaded.

• 3-compartment model
• V1, V2, V3 scale with weight
• Fixed intercompartmental rate constants
• No age or LBM adjustments
• Best for average, healthy adults
• Overestimates in obese patients → under-dosing risk
• Can overdose elderly

⠀Pros: Simple, quick induction.
Cons: Fails at extremes. No dynamic adaptation.

Schneider Model (1993)

Core Idea: Introduced LBM and age as covariates. Uses effect-site targeting via ke₀. Designed to be safer in the elderly.

• Fixed V1 (4.27 L); V2 changes with age; V3 fixed
• Uses James formula for LBM
• ke₀ = 0.456 min⁻¹
• Good for elderly, leaner patients
• Struggles in morbid obesity due to paradoxical LBM overestimation
• Effect-site targeting smooths bolus peaks

⠀Pros: Better in the frail; avoids early overshoot
Cons: Guardrails limit BMI inputs (~42 for men); may lag in induction

Eleveld Model (2018)

Core Idea: A universal PK/PD model that incorporates allometric scaling for weight and age. Based on massive population data (neonates to elderly, 1,000+ subjects).

• Unified across age and weight ranges
• Scales V1, V2, V3 non-linearly
• Adapts for obese, elderly, paediatric
• Faster induction than Schneider
• Potentially "too much" in the frail unless dose is titrated

Pros: Most robust and precise across populations
Cons: Not available in all TCI pumps; less familiar to some

Clinical Pearls

When to Use Which Model?

Patient Type	Recommended Model
Fit, young adults	Marsh
Elderly / frail	Schneider
Obese, children, mixed use	Eleveld

Key Concepts Recapped

• Plasma vs Effect-Site Targeting: Plasma is easier to calculate but lags behind effect; effect-site targeting uses overpressure to reach desired effect quicker.
• Overpressure Principle: A deliberate overshoot of plasma concentration to rapidly achieve effect-site equilibrium.
• LBM vs TBM: LBM is better for induction; TBM becomes more relevant for maintenance as redistribution dominates.
• Model Limitations: All models fudge reality. TCI ≠ autopilot. Always verify age, weight, and settings.

⠀

Pitfalls & Practical Tips

• Watch for the wrong drug in the wrong syringe—Etomidate lipid looks like 2% propofol.
• Guardrails in Schneider avoid paradoxical overdosing in high BMI patients—don’t override them blindly.
• BIS monitoring is mandated in paralysed patients (per AAGBI), but be aware of its limitations—especially in un-paralysed patients with EMG activity.
• Standardise your lines and pumps: e.g., Propofol on clear, Remi on blue, Metaraminol on pink.

References

1. Struys MM, De Smet T, Glen J, Laudiano-Dray MP, Martinez V, Mortier EP. The History of Target-Controlled Infusion. Anaesthesia. 2016;71(1):S12–S18. doi:10.1111/anae.13262
2. Marsh B, White M, Morton N, Kenny GN. Pharmacokinetic model driven infusion of propofol in children. Br J Anaesth. 1991;67(1):41–8. Available from: Pharmacokinetic model driven infusion of propofol in children - PubMed
3. Schnider TW, Minto CF, Gambus PL, et al. The influence of method of administration and covariates on the pharmacokinetics of propofol in adult volunteers. Anesthesiology. 1998;88(3):1170–82. Available from: The influence of method of administration and covariates on the pharmacokinetics of propofol in adult volunteers - PubMed
4. Schnider TW, Minto CF, Shafer SL, et al. The influence of age on propofol pharmacodynamics. *Anesthesiology.*1999;90(5):1502–16. Available from: The influence of age on propofol pharmacodynamics - PubMed
5. Absalom AR, Mani V, De Smet T, Struys MM. Pharmacokinetic models for propofol—defining and illuminating the devil in the detail. Br J Anaesth. 2009;103(1):26–37. Available from: https://academic.oup.com/bja/article/103/1/26/478960
6. Eleveld DJ, Proost JH, Cortinez LI, Absalom AR, Struys MM. A general purpose pharmacokinetic model for propofol. Anesth Analg. 2014;118(4):1221–37. Available from: https://journals.lww.com/anesthesia-analgesia/Fulltext/2014/10000/A_General_Purpose_Pharmacokinetic_Model_for_Propofol.27.aspx
7. General purpose models for intravenous anesthetics: The next generation for target-controlled infusion and total intravenous anesthesia? Curr Opin Anesthesiol. 2023. Available from: https://journals.lww.com/co-anesthesiology/fulltext/2023/10000/general_purpose_models_for_intravenous.20.aspx
8. Principles of total intravenous anaesthesia: Basic pharmacokinetics. BJA Educ. 2017. Available from: https://www.bjaed.org/article/S2058-5349(17)30085-9/fulltext
9. Absalom AR. Target-controlled infusion of propofol: A review of the past, present, and future. *Anesth Analg.*2016. Available from: Target-controlled infusion - Past, present, and future - PubMed

Other Resources

SOBA One-Stop TIVA Guideline: https://www.sobauk.co.uk/_files/ugd/373d41_eebe369c3c6b4021bff6f3da059aa796.pdf

SimTiva App: SimTIVA | Free TIVA Simulation App to Simulate Target-Controlled Infusion | Simple TCI Calculator

Society for Intravenous Anaesthesia (SIVA): https://siva.ac.uk

AAGBI Safe TIVA Guidelines: https://associationofanaesthetists-publications.onlinelibrary.wiley.com/doi/10.1111/anae.14428

---

"Thanks for listening guys... Every day you are getting better at this. Take it day by day, don't overcook yourself, don't freak out, and keep studying!"

Podcast Information & Support

Support the Show

• Donations: Help us keep the lights on and click the Support link here!

Contact & Feedback

• Comments: Share your clinical experiences and ask questions!
• Corrections: Help us improve accuracy and clarity

Follow GasGasGas On

• BlueSky:Gas Gas Gas (@gasgasgaspodcast.bsky.social)
• X / Twitter: GasGasGasFRCA (@GasGasGasFRCA) / X
• FaceBook: Facebook - Gas Gas Gas
• InstaGram: GasGasGas

Transcript

Episode 21 - TCI Models: Marsh, Schneider and Eleveld

Introduction and Welcome

00:00-00:31

Please listen carefully. Hello, and welcome to Gas, Gas, Gas. This is the best podcast for the FRCA primary exam. Our goal is to fill your brain with all this highly useful information. You might be in the gym right now, commuting, or ironing your scrubs. Regardless, the revision is eventually going to end, but for now expect facts, concepts, model answers and the odd tangent. Make sure to check out gasgasgas.uk. There's show notes there, there's loads more detail. Make sure to like and subscribe. Anyway buckle up, get ready for your mind to be bent into a new shape. And let's get on with the show.

Episode Overview and Topic Introduction

00:33-01:36

Hi, everyone. This is James at Gas Gas Gas, the podcast where we break down complex anaesthetic topics into digestible lumps. Today we are diving into the technical aspects of the different TCI target controlled infusion models used in TIVA, total intravenous anaesthesia. Specifically, we're going to chat Marsh, Schneider and Eleveld models. Interesting side note here is that actually these aren't licensed for use in the US currently. So the US guys and gals are just setting infusion rates and titrating, not actually allowed to use the models, so they have to do more work. Anyway, we're going to explore how these models calculate the propofol pharmacokinetics and then compare their strengths and weaknesses and think about how we might use these models in the real world. So if you're interested in this, and you're thinking, hmm, I'd like to know a bit more about TCI, I've just been guessing and picking a model and cracking on, then this is for you. Let's get on with it.

Pharmacokinetic Background and Historical Context

01:36-04:13

So we've talked about pharmacokinetics at length in other episodes, particularly the multi-compartmental or poly-compartmental madness episodes, as well as touching upon this a number of other times when we think about volumes of distribution, etc. And naturally we apply these concepts heuristically on a daily basis. But if we're actually trying to be truly accurate and say, well, I've given this much drug and the plasma concentration is this, and it's this concentration in their brain - well, one, we can't really do that with the current scientific data we have. And two, to get anywhere near close to that would take a lot of maths, a lot of paper, a lot of tables, which is untenably complicated for real time anaesthetising of people.

Hence some clever boffin thought, why not we get a computer to do such a clever thing. So they developed pharmacokinetic models that would provide projections of concentrations in a human over time. You would then titrate your syringe pumps to these projections. Then came syringe pumps with computers attached to them - very clever, then it would actually control that flow a little bit. And nowadays we have syringe pumps with the computers built into them, which is where we are today.

The other important differentiating factor between today and back way then was that today's systems are open, i.e. the Schneider model could be found on any particular pharmacokinetics enabled syringe pump and you could use any syringe you so fancied. Whereas once upon a time, and you might have found a syringe or two of this in the cupboard, was Diprifusor, which was a proprietary closed propofol infusion system courtesy of AstraZeneca. Naturally it cost quite a lot. Naturally once you'd emptied one fifty-mil syringe of propofol, you'd have to go and get another one out of the cupboard. Came in nice cardboard box that looked all very swish, but was somewhat prohibitively expensive compared to what we can do nowadays.

There is a cracking paper about the history of target control infusions by Struys et al. And it was published in 2016. There'll be a link to it in the show notes in the references. That is if you have a peculiar urge to find out more about the history of such things.

What is TCI? - Definition and Core Concepts

04:13-06:02

What is TCI? So we would define that in an exam as a method of delivering an intravenous hypnotic drug using pharmacokinetic models, where we have a set target concentration we are trying to achieve. There are three models. There's the Marsh model debuting in 1991, Schneider model coming out in 1993, and then the Eleveld model which came out in 2018.

These models are not all the same by any stretch of the imagination. They all differ in how they approach calculating drug distribution, elimination of propofol, and the effect site dynamics. You should go on to suggest that a TCI system agrees with yourself a target plasma or effect site concentration, and then it handles delivery of an infusion to achieve this within acceptable time parameters. The idea being that you have predictable anaesthesia with minimal overshoot or undershoot target concentrations.

Naturally, doing this is more precise than humans. We have a tendency to get distracted, think we know better, or perhaps choose not to care as much. A computer is going to adhere to the rules you set it to work within and will beaver away. It's important to remember, everyone, that doesn't mean it can think for itself, and you still have to apply thought. If you're trying to do the same anaesthetic to loads of different people with the same settings every time, you will find that some of your patients tip off of that normal distribution curve and are either too anaesthetised, not anaesthetised enough, or take too long to go to sleep, in fact.

Model Development Background and Human Variability

06:02-08:15

So before we get into the meat of all of these different models, let's just apprise ourselves of the facts. We know that propofol behaves differently in a predictable manner in the young versus the old, in the underweight or the overweight, as well as the healthy versus the critically unwell. And some of these variables also influence the onset and offset of propofol. However, if we try and reduce a human down to a mathematical model that predicts with perfect accuracy, we would be there for quite some time.

So the beginnings of these models were actually achieved by taking a group of people, administering propofol at set rates and then subsequently measuring their concentrations of propofol in the plasma and how asleep they were. They then used this data to almost sort of backwards work a model out to try and achieve something that predicts the plasma concentration based on the rates infused in a human of certain height and age, weight, etc.

They then thought, Well, you know, how about we try and make something for children? And so they started doing infusions in children. Measuring the plasma concentration they were achieving versus the one they predicted they would achieve and then back work to fudge factor to say, well, we know that children chew for propofol at this rate and they are far more regularly at this predicted plasma concentration. Therefore, let's administer a fudge factor, which counters that and achieves a true-ish plasma concentration of propofol.

So it was all a bit approximate. And you can see how you could sit down and try and work it all out and it would be mathematically beautiful, and then you apply it to real life and it doesn't quite work. I suppose a bit like Newtonian mechanics. It broadly works, but it actually falls apart when you get too big and too planetary and then you need something else. Is that difference big enough to care? Well, who knows. It's probably quicker to do it the first way round than trying to work it all out and seeing if you've got it right.

Plasma vs Effect Site Targeting

08:15-11:23

Thinking more about TCI, there are two target options: a plasma concentration target and an effect site concentration target. Now a plasma concentration target is an easier thing to get right because we can measure propofol concentrations in plasma after a period of infusion time that would achieve equilibration. The trouble is, and I know we've mentioned it before, is that plasma concentration for most drugs does not reflect the actual amount of drug in the site where the drug works. So out came effect site concentration targeting.

This theorised in mathematical terms another space or another compartment where the receptors existed, and decided upon how big that compartment might be, and the rate of shift into and out of that compartment and then seeks to achieve a concentration there. This means that there's a plasma concentration overshoot, that over pressure concept we've talked about with sevoflurane. In order to push drug into that effect site and achieve your drug action.

Now it certainly is not cricket to dose someone with propofol, wait for them to equilibrate, and then take a biopsy at their effect site to see how much propofol is truly there. Because we can't do that really very easily. This overpressuring event - if you watch your syringe pump, it'll deliver a bolus to achieve an effect site concentration, and then it'll stop infusing and it will wait a calculated period of time before reinitiating infusion to achieve an equilibration of the plasma concentration with that effect site. Because we all know eventually the plasma and the effect site will equilibrate, but with an effect sight concentration target, it will over pressure to get there faster. Otherwise, you'll be twiddling your thumbs for quite some time before you actually get enough drug where it needs to be.

Interestingly, the dose to achieve a sufficient effect site concentration correlates better with lean body mass. Whereas your maintenance infusion rate correlates better with total body mass. Now this makes sense because when you give someone a dose of propofol in their plasma as a bolus, it doesn't just immediately ooze out into the fat, it tends to go to the places that are highly perfused, like your brain. Whereas over time when you're maintaining a concentration that has started equilibriating out, the total mass of that individual naturally bears more relevance.

The Marsh Model (1991)

11:46-16:17

Now if you're in an exam and someone's asking you to compare and contrast Marsh, Schneider and Eleveld, then you're probably doing well, or the curriculum has shifted a bit. So let's get on with it. So I mentioned earlier that Marsh came out in 1991. It is based on a previous model called the Gepts model. This was derived from eighteen patients having different fixed rate propofol infusions with their plasma propofol concentrations being measured.

The Marsh model uses a three-compartmental model, describing that central compartment, plasma, the fast peripheral and the slow peripheral compartments. It falls down in that it does not account for lean body mass or age effects, and that its compartments are all derived from body weight, i.e. V1, V2, V3 are all proportional to body weight, so they vary quite a bit. And the clearance rate is weight proportional, i.e., how much propofol is exiting the V one, central compartment, out of the system. However, the K rates, i.e. the rates of interchange between these compartments, are fixed.

Therefore, a high BMI patient or a high weight patient has really large compartments. And you think, ah, well, so what? Who cares? They're quite big. But if you have a really large compartment and you asked it to achieve a concentration of five mics per mil of propofol, you can imagine that that's going to be quite a large dose of propofol. So what? Well, we mentioned earlier that the dose required to knock someone out is different to the dose required to maintain their sleepiness over time. So a big cracking dose based off of their whole body mass is actually going to overdose them. So Marsh overdoses big people. Therefore, we would expect instability at extremes of weight using the Marsh model.

However, the benefits of Marsh, it is simple, and it does work well for average weight patients. And simple means fewer variables, which means less likely to put something into the computer wrong, and we know that the computer likes rules, and it'll follow those rules, and it'll follow the information you've given it. Your pump cannot eyeball your patient. You can imagine that Marsh, in a healthy fit adult, probably gets them off to sleep quite quickly because it's quite a large dose compared to some other models we're going to talk about in a moment. But it doesn't take into account age either. So actually, you're going to be giving an excessive dose of propofol to an elderly patient who's overweight and you might find yourself in quite a significant pickle.

Marsh is an early model. It was developed using plasma concentrations, and whilst some models have a fudge factor that create an effect site, it is generally not used, and should be used in plasma targeting only. It is a generally effective all rounder for average patients, requiring anaesthesia where you've chosen to use TCI. But we all need to bear in mind that all those volumes in the patient are based on the weight of the patient. It doesn't take into account age, so it ends up overdosing the elderly population. Also in the obese population, it tends to overestimate the plasma concentration over time, and as you get further into your anaesthetic with your large patient, you may find yourself running them drier or leaner than the machine suggests you are. You might be set at four, but actually your somewhat overweight patient might have a plasma concentration of two and a half or three.

So you end up flying unexpectedly close to the wind and you might end up increasing what you think your target is later in your operation. Someone else might walk in and decide that you are overdosing the patient because they're used to seeing numbers of 2.4 on the propofol if they haven't spotted that you're using Marsh. It's always important to be wary of just looking at that effect site target number and saying, well, that number is too high or too low. You need to think about your patient when using TCI.

So the rule of thumb when we now compare Marsh to Schneider, is that Marsh on average will dose more propofol than Schneider for the same weight patient with the same targets over the same duration.

The Schneider Model (1993)

16:33-23:36

So hop, skip and a jump. We're walking down the red carpet. It's 1993 and the Schneider model makes its debut. So why did they think they wanted to develop Schneider? Why wasn't Marsh good enough? Well, we've spoken that Marsh overdoses old folk. So the intent was to develop a model that did not do that. We're trying to be safer.

The Schneider model study volunteers ranged from forty-four to one hundred and twenty-three kilos and from twenty-five to eighty-one years of age. This is a bit better. More volunteers. And we're doing some old folk. And it utilises some different variables. So it will calculate your lean body mass based on your height as well as just relying on weight. Therefore, it achieves a better estimate of effect site concentrations. And from the off, it introduced an effect site to the maths it was using to consider what dose of propofol to give.

However, what about how it works out all the volumes of compartments? So Schneider has a fixed V one compartment. So it is always exactly the same, and that is 4.27 litres. So if your patient's 150 kilos or 50 kilos, it always presumes that your plasma compartment is 4.27 litres. Unless the patient is so scrawny, and then it will down measure the V1. So if someone's really very thin, it will reduce the V1 size, but it will never increase the V1 size.

Your V2, your fast peripheral compartment, alters with age. So that space where your propofol is first going to theoretically get into your heart, your brain, other highly perfused organs. They alter that with age in the Schneider model. Why is this useful? Well, because your V2 is the first step to effect site, isn't it? Because your effect site ain't in the plasma and it isn't in the parenchyma of the brain. Next to a cell or a neuron, we should say, in the brain. Therefore, you can imagine that the effect site shifts somewhat with age.

And your V three is again fixed, so your really big patient might have a fat compartment that is larger than the perceived fixed size of it within the Schneider model. So you can see here that it's much less clear cut. The Marsh model is simpler. The Schneider model is more complex. Now, some of these intercompartmental coefficients are varying and some of them are fixed. So the rate of flow into your slow compartment, into it and out of it, are fixed. So K1 into 3 and K3 into 1 - fixed rates of transfer. So that propofol eking out into fat and propofol eking back from fat into plasma, fixed rate.

Your shift from plasma to your fast compartment, your brainy, hearty compartment, these alter. And your clearance rate from central compartment, your K10 or zero, utilises the total weight of the patient and the lean body mass and the height to figure it out. I've just said three things that influence the K10, and it is a complex fudge which I don't know anything more about, but it does take into account three variables to decide on the clearance rate.

Should you ever choose to use Schneider in plasma concentration mode, I can tell you that because of the fixed and small V1, and if you're just targeting a concentration 4 in the plasma as opposed to a concentration 4 at your effect site, it'll only give a teeny bolus, and it'll always be more or less exactly the same bolus, unless the patient's very slim. So it'll take quite a while to achieve an effect site concentration by just using your plasma target.

Generally speaking, everyone suggests for this to use Schneider in effect site. If you're doing conscious sedation for a really long case, or someone's terribly frail, then you might choose to use Schneider in plasma concentrations, but you'll still be really twiddling your thumbs for quite some time.

There is another thing to note with Schneider is that it falls down a little bit with its lean body mass calculation. This tends to paradoxically cause quite a lot of bother with clearance in patients who have a high BMI, particularly 42 in men and 37 in women. So Schneider won't let you put in a patient's data that would yield a higher than 42 or 37 BMI. Now you could just max it out and then give them a bit more of a dose. But that's imperfect.

And I'm sure we all know that there's quite a few people who we anaesthetise every day that are morbidly obese and have BMIs in the 42 to 37 range and higher. Just for everyone's interest here, what happens is if you were to not have these guardrails in place and actually decided to tell the model that you had a bloke who had a BMI of fifty, the clearance parameter calculation that I mentioned earlier, that was a fudge of several different factors, would paradoxically yield a humongous clearance rate in that particular instance. Therefore, the propofol infusion rate would just escalate and escalate and escalate to try and maintain this theoretical plasma concentration, when in fact we all know that we're just ending up overdosing the patient in propofol.

Therefore, guardrails in place that stop you hitting this paradoxical part of their that curve they've calculated for the clearance of propofol. And that's why there's a limit on the Schneider Max BMI. Interesting, isn't it?

Broadly speaking, is Schneider any good? Yeah, it probably is going to not overdose your elderly folk quite so much. It might take longer to get someone off to sleep because remember we've got that fixed V1. And it's not terribly good at patients with really high BMIs, where you have to accept that it only thinks the patient has a BMI of 42 when they might have a BMI of 52, and you would have to titrate accordingly. And that was in nineteen ninety three that came out.

The Eleveld Model (2018)

23:36-26:38

And then actually, it seems like everyone sat on their hands for well over a decade and Eleveld started rocking onto the block, you know, clicking its fingers and looking all sharp and swanky. Eleveld rocked up in 2018. So why do we care?

Well, Eleveld has done what the other models didn't do, and actually it soaked up a humongous data set in which to decide upon the effect site concentrations achieved based on height, weight, age, etc. And they included really small babies, adults of a normal weight, obese people, really thin people, elderly people. It uses something called allometric scaling for weight and age and it means that you don't have to think quite as hard about which model to pick depending on your patient in front of you, because it'll cover both.

It adapts to body weight and age, so it doesn't massively overdose obese patients like Marsh would, but also isn't going to leave you twiddling your thumbs with an obese patient when you're using Schneider. When we say allometric scaling, it also considers patients who are younger and smaller and waterier versus the older, drier, and fattier patients.

So we all know that there's a greater amount of total body water in young people compared to the elderly. So Eleveld, because it doesn't have this fixed compartment, nicely induces most people. It doesn't take as long but it still may end up being a little bit more heavy-handed compared to Schneider in your older and frailer patients, in which case you might still be better off picking Schneider or starting at Eleveld at a lower rate and titrating up.

You might be asking, gosh, how many patients did they use for this study? Well, actually they took a lot of information from many different studies in order to develop the Eleveld model. They used about fifteen thousand propofol concentrations and also about twenty-eight thousand BIS measurements from a thousand and thirty-three different patients. The age range was from twenty-seven weeks to eighty-eight years and they were ranging from six hundred and eighty grams to one hundred and sixty kilos.

The real difference talking point between Marsh and Schneider versus Eleveld is that Eleveld is a so-called pharmacokinetic and pharmacodynamic model, PK and PD, because it's used such a broad range of data. And it scales the compartments it uses allometrically, i.e. these characteristics change in relation to its size instead of using a linear calculation.

TCI as a Tool and Safety Considerations

26:51-28:32

So we've had a chat about these three models that deliver propofol as an infusion using maths to aim for a sufficient rate of infusion to maintain anaesthesia. But it's important to also step back and just imagine this contraption as a black box, because it's far too much to really truly understand exactly what it is doing. We are not the designers of said models. If you just assume that it knows best, that's a mistake. And if you assume that it's completely useless, probably also a mistake.

TCI is a tool to deliver a safe rate of propofol infusion when you use it well. And there is no substitution for applying your brain and being vigilant when monitoring patients' depth of anaesthesia. It is recommended by the AAGBI that any paralysed patient has BIS attached. Why? Because the NAP5 study demonstrated that the odds of awareness in TIVA anaesthesia increased by an order of magnitude if the patient is paralysed.

This is why practice is evolving, and in fact avoidance of paralysis unless absolutely necessary often comes about with TIVA. If you give someone a sufficiently big dose of remifentanil, you can often intubate them sans paralysis and this means you are in a cohort of patients who are more likely to wriggle before becoming aware, and you can find out that that cannula has tissued. This is not a podcast about the pitfalls of TIVA and how to do that safely. That is something that's going to come much, much later.

BIS Monitoring Considerations

28:32-29:31

Now in practice, almost all patients who are TIVA anaesthetised in the hospitals I've worked in have BIS attached. The BIS algorithm itself is proprietary and until it was reverse engineered by some boffin, we didn't actually really know how it worked. I can tell you that it does rely quite heavily on the absence of the electrical signals from muscle movement.

I'm sure we've all seen the neuromuscular blockade wear off and the BIS go up, and we're sat there thinking, Well, they were deep, and now they're not deep. Is it something I've done? You then reparalise them and the BIS goes back down. Kind of makes it a bit useless. So whilst the BIS is a useful additional data point, it is absolutely not enough to rely on it or to completely disregard it.

Obese Patients and Practical Resources

29:31-30:33

Anyway, so let's move forward. And we've mentioned about obese patients. Like what on earth do we do with them? What if they've got a BMI of sixty or seventy? And we know that sevoflurane soaks in quite well, as does propofol. Are they going to wake up better with propofol than sevoflurane? Quite probably.

This is not a podcast for me telling you exactly how to TCI an overweight person, but I will tell you that SOBA, the Society of Bariatric Anaesthesia, is very, very useful as a data point perspective. There'll be a link in the show notes. There's an app which will help you to calculate appropriate drug doses for overweight patients when you're anaesthetising them. I have it, it's very handy, and it'll guide you as to whether or not using total body weight, adjusted body weight 40, adjusted body weight 20, lean body mass, etc. Dosing agents for anaesthesia and as to what dosing weight you might choose, depending on the model you're using, to TCI patient.

Practical Safety Tips and Pump Programming

30:33-31:53

So, overall, in a nutshell, TCI models are dumb. They're only as smart as they're ever going to get, Eleveld is going to be Eleveld, and it will be there in black and white if you get it wrong, and bedlam will ensue. So always check the height and weight as you put it in. And just think, does the patient really look fifty-five kilos, or has someone written this down wrong?

Confirm the model and the agent, and avoid mixing up syringes. I make a point of always prepping syringes prior to a patient arriving and trying to program the pumps if I have the data ahead of the patient arriving. If I am being distracted whilst I am programming pumps, I stop and I will start over. It's also just important to make sure you don't accidentally use 2% propofol in a 1% propofol model. Most TCI models assume you're only using 1% propofol, so don't cock that up. Equally etomidate in lipid emulsion form looks a lot like propofol.

Now, everyone might have different styles. I generally have three pumps. I'll have propofol at the top and then I'll have remifentanil underneath, and then I'll have a metaraminol infusion below that. The voices in my head tell me that is what I should do, and I will never alter that, because then I'm always going to know where the propofol is and if I want to bolus it. And the voices in my head are very good at telling me to do what they want me to do. Terrifying.

Summary and Conclusion

31:53-33:57

Anyway, so in summary, TCI has been an iterative thing coming on from a computer telling you what to do, to a computer telling a pump what to do, to a pump telling itself what to do. Marsh, Schneider, and Eleveld all probably have a role, but Eleveld is probably the future, if you have access to it. It generally seems to fly a bit closer to the wind. So an effect site of four in Schneider versus an effect site of four in Eleveld. To me at least, the patient seems a little bit lighter on Eleveld. So you maybe need to run them a little bit wetter, hotter, less lean. There's probably some sort of carburettor reference in there.

Anyway. Thank you very much for listening. I hope this has been useful. Check out the show notes for more detail. There are a bunch of linked papers, and I've read a lot of papers to try and distil this information out. But all the papers I've used will be in the show notes and there's loads more reading if you are super duper excited by all of this. There's also the Sintiva app. I've plugged it before, I'm plugging it again. It will let you simulate propofol dosing in patient sizes, volumes, weights, etc., and let you see how those plasma concentrations versus effect site concentrations alter over time, it really does help you visualise what's going on.

Anyway, if you've enjoyed that, make sure you do go ahead and subscribe and listen out for future episodes. We're going to get back into the long grass of pharmacology. Thanks for listening guys. I hope you found it useful, but if you found it awful, do let me know. Please like and subscribe, register with whichever podcast platform you find yourself using and leave a comment if you think I need to square something away.

I just wanna make sure that you guys know that every day you are getting better at this. There is a bucket of content to try and consume and it is like drinking from a fire hose. Take it day by day, don't overcook yourself, don't freak out and keep studying.