Tasty Morsels of Critical Care

Welcome back to the tasty morsels of critical care podcast.

We’re going to cover a bit of an environmental/tox topic today and look at carbon monoxide poisoning from Oh’s manual chapter 83 on burns. I have previously covered this on the old tasty morsels of EM series back when i was doing my EM fellowship exams.

As you no doubt remember from school chemistry classes, carbon monoxide is a colourless, odourless, tasteless gas produced when combustion occurs with insufficient oxygen.

We’re likely to see this in a couple of contexts.

1) the house fire victim, pulled from the fire unconscious and sick

2) the sub acute or chronic poisoning in a patient presenting with headaches and flu symptoms that seem to get better when they leave the problem environment. The classic EM example is the whole family who present with flu symptoms and no fever and even the dog is sick. We’re much less likely to see this cohort in the critical care side of things.

How does it make people sick? Haemoglobin is a fickle little protein, while evolved to carry oxygen to needy tissue beds it actually has a distinct preference not for our beloved oxygen but for carbon monoxide. Introduce some carbon monoxide at the alveolus and the haemoglobin molecule will bind to CO with an affinity 240 times that than for oxygen. I take that number of 240 somewhat at face value but I presume someone got a PhD from working that out. In visual form my preferred means of explanation for this would be the distracted boyfriend meme where the haemoglobin boyfriend looks longingly over his shoulder at the carobon monoxide while his oxygen girlfriend looks on in horror. Hopefully you get the idea.

So instead of having lots of circulating oxyhaemoglobin we’re instead left with lots of not especially useful carboxyhaemoglobin. Let’s imagine 50% of our Hb is now carboxyHb and 50% is OxyHb we’re left with a sort of severe fucntional anaemia where half of our Hb is out of action. One might be inclined to think that this is the major cause of morbidity and mortality in CO poisoning but in fact this is only a small portion of the problem. CoHb actually has a direct cytotoxic effect on things cytochrome oxidase and myoglobin function. As such it interrupts the whole process of oxidative metabolism and life as we know it.

We can measure the level of CO fairly easily, any blood gas machine worth its salt should be able to give you a break down of the types of Hb present in the sample. This is co-oximetry and typically it’ll show you oxy, deoxy, carboxy and met haemoglobins. All these different forms of Hb absorb different wavelengths of light. The lowly pulse oximeter does not have the subtlety to distinguish the different wavelengths as it only functions at wavelengths of 940 and 660nm. Indeed the pulse ox often demonstrates a non diagnostic number somewhere in the 80s rather than a true reflection of the CarboxyHb or OxyHb present.

Severe CO poisoning resulting in obtundation is going to have high level of COHb on our cooximeter. >10% is quoted but it’s more often over 30%. Patients are going to be pretty sick often from multiple pathologies but COHb on its own is enough to produce severe neurological injury, shock and even cardiac injury is also quite prevalent. Expect a high lactate given the disruption of oxidative metabolism. Resuscitate and investigate as you would any sick patient.

Treatment is nice and simple in that we just give loads of oxygen. Oxygen reduces the half life of CO in the blood quite dramatically, commonly quoted numbers are

There is a substantial rationale and literature on the use of hyperbaric oxygen as a means of accelerated clearance of COHb. But the RCTs that have been done don’t seem (to me at least) to give a clear benefit. The Lindell Weaver NEJM RCT in 2002 did suggest a neuro benefit but only 8% of the patients in this trial were intubated. A follow up trial in 2011 by ICU steroid guru Djilalli Annane did not find a benefit . So if anyone should get this it might be the non intubated isolated COHb poisoining. This is not really our cohort. Our cohort is likely to be tubed, shocked, with multiple injuruies and not someone you want to transport cross county to put in a single person hyperbaric chamber for hours at a time.

Oh Manual Chapter 83

Welcome back to the tasty morsels of critical care podcast.

We’ve been talking about pulmonary hypertension, last time we had a pretty broad overview with a focus on group 1 or pulmonary arterial hypertension. This time we’re going to go through some management strategies that might keep you between the hedges on a night on call or a fellowship exam viva.

We briefly mentioned the PH specific drugs that someone might be on. The evidence base for these is almost exclusively in group 1 PH. But what should we do with these meds in someone with group 1 PH who has just arrived back from theater after a laparotomy and a hartmans and they’re on a bit of noradrenaline? The simple answer is continue them. The more complicated answer is you should usually continue them. For example there will be the very rare patient whose pulmonary vascular resistance is kept low in the community with a PICC line and an epoprostenol pump. They are critically dependent on this drug with a very short half life and it should be continued at all costs. Think about it like an adrenaline infusion running at 10mcg/min, not something you can tolerate a break in.

A recurring message from the review papers on critically ill patients with PH is to focus on treating PVR not PA pressures. This is a somewhat philosophical approach that reminds us that the PA pressures themselves don’t prognosticate especially well but a failure of flow from right to left will result in cardiogenic shock and death.

We have a lot of vasoactives to choose from in helping with this, most of which have varying impacts on the PVR. Vasopressin has some animal data suggesting it causes less rise in PVR than our beloved noradrenaline but take that with an appropriately loosely defined portion of salt given that animal data is not ICU patients. Milrinone seems like a great idea as an inotrope that is easy on the PVR but the often dramatic drop in SVR is often a disaster. Dobutamine has the benefit of at least having substantial clinical experience in PH patients even if the tachycardia and even worse the a fib is less than desirable.

The ventilator is a bit of a poisoned chalice. Not only do you have to tolerate a significant risk of peri-intubation cardiac arrest even once you get them on the vent you have to deal with the adverse effects of positive pressure on the RV. The only upside of the vent is that it might make them easier to oxygenate but only if the cause of the hypoxia was a big shunt physiology like a pnuemonia. Oxygen is a great tool for reducing PVR so if we can leverage that then that’s great. However, a lot of hypoxia in end stage PH is reduced mixed venous oxygenation due to low cardiac output and the vent does nothing good for this.

Once on the vent we want a goldilocks’s zone of lung unit recruitment. Too little PEEP we have atelectasis and shunt and hypoxia and vasoconstriction. Too much PEEP and we have overdistension which itself can raise PVR by squeezing the pulmonary vasculature. Finding that sweet spot for the PEEP is a whole post or 10 on its own.

While on the vent it’s a good opportunity to deliver some inhaled therapies. The original gangster here is of course nitric oxide which is one of our target molecules in PH. In a crisis and a failing RV, this might get you out of a tricky spot. But given its expense and not being widely available its worth considering other inhaled options, particularly intermittent nebs of iloprost or a continuously nebulised eporprostenol solution both of which i have seen implemented to good effect.

In terms of monitoring should we be reaching for a PAC? Well, take a step back to start with. We probably need the CVP more. The RV is the first downstream organ that suffers under the burden of worsening PH and if the RV is failing then the CVP will be rising. Like any monitoring tool, a PAC in itself is going to do nothing but provide you with scary looking numbers, particularly the PA pressures which, remember, you should largely ignore. But picking up a severely raised wedge for example might push you to be much more aggressive with your diuresis and left heart management. A continuous cardiac output monitor will allow you to titrate your vasoactives with a great deal more confidence and accuracy

The other monitor I would reach for would be echo. I am a self confessed echo phile so take that into consideration but one of my targets of treatment is going to be how the heart looks. Is the IVS becoming less flattened, is the RV less distended, is the TAPSE improving etc… Echo early, echo often in my book.

Atrial fibrillation is something of a right of passage in the ICU. Have you really been critically ill if you haven’t even had an episode of fast AF? When it comes to PH it’s often poorly tolerated and the approach to rhythm and rate control probably needs to be a bit more aggressive than usual. Our usual choice of vitamin A, amiodarone is a good start but you may need other agents like dig or even DCC to get control. A consistent message from the reviews is to avoid beta blockers. The negative inotropic effect on an RV that is already functioning at peak capacity is not going to be good.

Our first reaction when faced with hypotension is often to load with fluid, this makes sense when we think of the frank starling mechanism, we want to be sure our LV is appropriately pre loaded. But in PH the issue is a failure to deliver volume or flow from the right heart to the left. We can dump a litre into the venous side of the circulation but the PVR just stops it getting efficiently through to the LV. If your patient is hypotensive then the RV is already failing in its basic function of delivering volume and flow to the LV while keeping the CVP low. More fluid is almost never going to fix this.

Indeed diuresing the hypotensive patient may well be the way to go. If you can decongest the right side and reduce the bowing of the septum you’ll get both the RV and the LV working more efficiently

This is only a taster of things you might want to try in a critically ill patient with severe PH. It is important to emphasis that they are not evidence based overall. Most of it is interpretation of clinical physiology at the bedside and applying the available manipulations. Which is of course what makes it so much fun.2

My own rambling review of pulmonary hypertension on JFICMI website.

2022 ESC Guidance

Welcome back to the tasty morsels of critical care podcast.

This time we’re looking at pulmonary hypertension. Mainly cause I recently had to give a talk on it so it’s fresh in my rapidly diminishing brain cells and thought I should get it all written down before I forget it. We’re going to try it as a 2 parter. Part 1 will cover a broad overview of pulmonary hypertension and part 2 will focus on management strategies for a PH patient in the ICU.

Saying a patient has PH does not really tell you very much. All we mean is that pressures in pulmonary circulation are higher than they should be. Saying someone has PH and not quantifying it is a little like saying someone has cancer but not saying which organ or how advanced it is. We need to go a bit further than just say they have PH and quantify the cause or rather which group of PH they’re in. We also need some way of quantifying the severity of it.

The definition of PH since the 2022 ESC guidelines is a mean PAP of 20mmHg on a right heart catheter. Echo can be used to screen for “probability” of PH but the right heart cath is needed to make the diagnosis. Once you’ve defined that the pressure is high the real doctory work begins as you have to figure out the likely cause. The language the guidelines use is “group”. You should be able to put your patient into 1 of 5 groups.

To give an example you are handed over someone who has known PH. You dig a little deeper and see they have an mPAP of 27 on a recent right heart cath. Their echo shows a poorly functioning LV and severe MR. The PH here is going to be group 2, PH secondary to left heart disease. This is by far the commonest.

Or another example, you are told someone has PH. You dig a little deeper and see an echo report that says the left heart works well but the right side is dilated. You dig a little deeper and see the clinic letters describing severe end stage emphysema. This is likely to be group 3 PH, PH secondary to lung disease.

In both those examples the PH is a problem but it is a downstream effect of other disease. And unless you can fix the heart or lung disease then the patient is in trouble, indeed if the patient dies in the coming weeks to months it’s likely going to be the left heart disease or the lung disease that kills them.

Let’s spend a few minutes talking about group 1 PH, sometimes called PAH. This is rare but often very severe and progressive and comes with some unique medications so it’s worth discussing. These people should have normal lung parenchyma and normal left hearts. There are a variety of specific causes in group 1 but a lot of it is described as “idiopathic”. It is a progressive pulmonary vasculopathy where the tiny arterioles suffer intimal proliferation and eventual fibrosis due to a variety of vasoactive molecules. This transforms the pulmonary circulation from a very compliant, low resistant circuit into a narrow and stiff group of pipes. The right heart is evolved and very comfortable with assisting large volumes of blood through a low resistance circuit. In hroup 1 PH, the change in pulmonary vascular resistance is more than the right heart can cope with and the right heart over time starts to fail in its primary purpose of maintaining a low CVP while delivering preload to the LV.

Over the past decades a number of classes of drugs have been developed that target the vasoactive molecules that cause the vascular changes. These can be split into 3 classes

1) endothelin receptor angtagonists which do exactly what the name says: reducing endothelin. Drugs like macitentan fall in that category

2) PDE5 inhibitors. These inhibit the enzyme you expect from the name but the key outcome is that there is an increase in nitric oxide something that causes pulmonary vasodialtion. Sildenafil or tadalafil are two common drugs in this group

3) prostacyclins. These vasodilate and reduce proliferation in the vascular bed and typically IV epoprostenol is the drug of choice here.

These drugs have proven disease modifying benefit but only in group 1 PH. We have not been able to prove any benefit for those with PH from left heart disease or lung disease. The more severe their disease the more drugs they might be on. Some patients are even on IV epoprostenol in the community to keep their PVR compatible with life.

These 3 classes of drugs have had a significant impact on both length and quality of life in PH. But the prognosis in group 1 PH is still one of progressive irreversible disease in the longer run. There are lots of features that are well validated on an outpatient basis to determine prognosis however that is rarely the question we’re faced with. For example we know ICU admission is a poor prognostic sign in severe PH but this is generally the very point we get involved at. As usual i suspect decisions about prognosis in the ICU setting are typically decisions about limitation or withdrawal of life sustaining therapy and they all depend on reversibility. If someone with severe PH has a pneunonia then we can probably turn that around then that’s something to consider. However if the right heart and liver are failing due to worsening congestion from progressive PH then that’s a different question.

That’s enough for today and to give you an overview of PH, next time we’ll focus on some management strategies.

My own rambling review of pulmonary hypertension on JFICMI website.

2022 ESC Guidance

Welcome back to the tasty morsels of critical care podcast.

Last time i was butchering my way through a diagnostic approach to hyponatraemia, particularly the forms likely to end up in the critical care end of the hospital. This time we’ll take a punt at how you might approach management. In an ideal world of course you would have all of the diagnostic tests back and you’ve been able to make a very solid diagnosis of the cause of hyponatraemia and you would institute a bespoke treatment course for the underlying disease and the resultant hyponatraemia. But as we all know in critical care we often work with less than ideal information and have to begin treatment while the diagnostic process is ongoing. Hopefully what follows will provide enough broad brush strokes to get you through a night on call or even worse a viva.

We’ll start with truly emergent situations. Older person presents to the ED after being unwell for several weeks. They have a seizure on arrival and a Na comes back at 105. This is a fairly solid indication to give hypertonic saline. In this scenario they are seizing because of the low Na and rapid increase of the Na is needed to stop the seizure. The European Hyponatraemia Guidelines would suggest 150mls of 3% saline over 20 mins aiming for a rise in the Na of 5mmol/L. This bit is usually pretty straightforward. The sodium rises, the patient stops seizing everyone relaxes but then the Na continues to rise, well above the 5mmol we wanted and a panic ensues.

The guidelines suggest a max rise of 10mmol in the first 24 hrs and 8 mmol/day after that. It is hard to overemphasise how easy it is to blow past that target unless you are paying attention. So how do you control the rise in the Na? If it’s rising too quick it’s often because the patient is losing lots of water through the kidneys which concentrates the plasma raising the Na in the blood. You can replace that water loss by giving a decent bolus of free water in the form of something like 5% dextrose. An alternative method involves using the wonderfully named DDAVP clamp. In this scenario you’re using the DDAVP to tell the kidneys to excrete less water therefore limiting the rise of the Na. I have not seen particularly strong data on one method vs the other for limiting the rise and indeed I have seen clinicians use either or indeed both to good effect.

The European guidelines do use the phrase “severe symptoms” as an indication for a bolus of hypertonic. Unfortunately it’s a little less clear what constitutes severe symptoms. A seizure seems fairly easy to define but “coma” is a little bit more vague.  The guidelines are clear that you have to be able to put the symptoms down to the hyponatraemia and not some other cause. But as we all know patients often have multiple reasons to be obtunded including sepsis or intoxication or multiple other causes. As such the decision to give hypertonic can be a little subjective and fudgeable.

For many patients the best thing you can do is very little. A former consultant I worked for had somewhat facetious plans to start a hyponatraemia clinic that involved locking the patient in a room and denying them access to water and letting the body sort it out over several days. There is an element of truth to that as for many of the hyponatraemics simple fluid restriction and time will correct things.

Lastly, our hypertonic of choice is typically 3% saline with an osmolality somewhere in the range of 1000 or so. Typically we’re a bit reticent to give such concentrated solutions through a peripheral IV but there are a few papers suggesting that this is fine at least on a limited basis. I will say that once the hypertonic is in and you’re reaching for a 2nd or a 3rd you should probably be thinking about a CVC as the access for administration and indeed regular sampling is really helpful.

European Hyponatraemia Guidelines

Oh Chapter 95

Welcome back to the tasty morsels of critical care podcast.

Today we cover an incredibly common inpatient issue – hypnatraemia. We’ll often find 1 or 2 of these in our high dependency unit at any given time, mainly due to the requirement for frequent testing of Na levels that seems beyond the remit of normal ward level care. The approach I describe here is neither comprehensive or especially robust but it is how I approach it. Caveat emptor and all that.

The over bearing demyelinating elephant in the room in hyponatraemia is the risk of osmotic demyelinating syndrome (the pathology formerly known as central pontine myelinolysis). If we correct the Na too fast will our patients end up with a severe brain injury? This is rare but is a very real phenomenon.The brain is actually quite good at adapting to sodium levels that have lowered over a few days or weeks. Hence why the slow developing sodium of 120 often causes minimal or no symptoms. However once the patient is in this adapted state (as mentioned this probably is after a few days at a minimum) then a rapid return to baseline sodium can cause ODS. By contrast a rapid drop in sodium, eg over a few hours drinking litres of unnecessary water during a marathon, is poorly tolerated but the plus side is it can be corrected fairly rapidly without harm.

Most of the hyponatraemia we see admitted through the ED will be hypoosmotic hyponatraemia. The bucket here will include heart failure, cirrhosis, SIADH, tea and toast and beer potomania. I’m going to put these common ones to one side for a minute and look at some of the niche exam ones.

For example, i said hypoosmotic hyponnatraemia there, so presumably there could be an isotonic and a hypertonic verison. There is indeed. The isotonic hyponatraemias are usually from spurious results. For example, when you have high lipids (super high, like high enough to cause pancreatitis high) or high proteins (eg high paraproteins like myleoma) the measurement method can underestimate the sodium. You can work this out by always sending a serum osmolality. If this is normal but the Na is 125 and your calculated osmolality is low, then you have an isoosmotic hyponatraemia. You should then check the lipids and the protein. Hypertonic hyponatraemia is another strange beast. This time the tonicity is high from something else such as high glucose or mannitol drawing water from cells into plasma. Again a mix of clinical context and a serum osm will help you out here.

Let’s go back to the bread and butter (or should i say the “tea and toast”) hyponatraemia, the hypotonic or hypoosmotic hyponatraemia. Context as always will give you lots of clues, if the patient has consumed nothing but beer for weeks then the likely causes is beer potomania. If the patient has a new cancer then SIADH is high up your list.

I confess I lean heavily on the approach you can see on Deranged Physiology and have Alex Yartsev’s flow diagram saved on my phone and i look at it almost every time i’m trying to work this out. The first test (assuming you’ve confirmed this is hypotonic hyponatraemia) in this algorithm is urinary osm, the question you are asking here is whether the kidneys are doing what they’re meant to be doing in the face of a low sodium. A normal sane and functioning kidney will try and lose water to conentrate the plasma in order to bring the sodium back up to normal, in other words the kidney should be producing a dilute urine with a low osm. Next step is to check the concentration of sodium in this dilute urine. If the kidney is doing what it should be doing it should be holding onto to all the sodium it can and urine sodium should be low. The problem here is too much water, not enough solute. Think, beer potomania, tea and toast, and polydipsia. If the urine is dilute but the sodium is high then you know something has gone wonky in the kidney itself, typically AKI or resolving ATN.

On the other side of the algorithm we have a concentrated urine, in other words, a high urine osm. The kidney is holding onto water and concentrating the urine. This may be a very sane and sensible response by the kidney if you are frankly hypovolaemic from eg gastroenteritis. The kidney also gets tricked by a few conditions into thinking its hypovolaemic, things like CHF or cirrhosis where the kidney itself just mightn’t be being perfused very well.  In this scenario you should have a concentrated urine with a high osm and a low urinary Na as the kidney holds onto Na for all its worth in an effort to maintain effective circulating volume. On the other hand you might find a concentrated urine with a high osm but a high sodium also. This tells us that the kidney is handling water reabsorption OK but has lost the run of itself when it comes to regulating sodium. Something may be strong arming the kidney into losing more sodium than it should, like thiazides or an external actor like ADH, in this case it would be inappropriate ADH, hence the syndrome of inappropriate ADH. In addition a lack of steroid (and in particular the mineralocorticoid part) or a dodgy thyroid may cause the kidney to lose sodium when you shouldn’t. Or of course this scenario could be due to intrinsic renal disease.

So that’s 8 or 900 hundred words running through the deranged physiology algorithm and you can imagine that simply looking at the algorithm would probably be a better use of your time so go do that.

Next time we’ll have a look at how we might manage hyponatraemia

Deranged Physiology – Wonderfully titled ” A Lazy Man’s Classification”

Oh Chapter 95

Welcome back to the tasty morsels of critical care podcast.

Today we’ll cover some key exam content, all be it not something you’re likely to run into in the ICU too often. The thyroid is a deceptive little organ, tucked in the neck, quietly secreting hormones and interfering in negative feedback loops. It usually restricts its mischief to outpatient clinics by running hot or cold on a chronic basis, occasionally hypertrophying and interfering with its more important neighbour the airway. But every now and then in a pique it decides it’s fed up of this low level mischief and uses its deeply embedded relationship with the rest of the body to wreak havoc.

We’ll split this into 2 parts, one when the thyroid goes on strike and is under active and the other when it goes bananas and secretes far too much hormone

Some basic physiology. Thyroid hormones are essential for all organ systems. The active forms are T3 and T4. T3 is generally the more active one. They are synthesised by incorporating iodine into tyrosine residues in thyroglobulin in the thyroid gland. Hence how iodine deficiency can cause a deficit in thyroid hromone. Their release into the circulation is stimulated by TSH. TSH causes endocytosis of this thyroglobulin into the follicular cells where they undergo hydrolysis into T3 and T4 which is released into the circulation. Both are highly protein bound to thyroid binding globulin.

Our first relevant condition is the wonderfully named thyroid storm. Most commonly you might see this as part of untreated Grave’s disease. It can be precipitated by the usual physiological stressors such as surgery or sepsis etc…

Expect to see (at least in an exam scenario)

For awareness there is a clinical prediction tool that rejoices in the name Burch-Wartofsky Point Scale. This includes most of the features listed above. It’s clear that the features listed above are fairly non specific and like always it’s likely just sepsis. But if something in the spidey sense tingles then finding undetectable TSH and high T3 or T4 should really get you going. In reality this is an incredibly rare diagnosis, one which in its fulminant form i have yet to see. Or perhaps more accurately one that i have failed to diagnose as yet. This is of course hardly surprising as it is hopefully clear by now on this podcast that I am not especially good at what i do and continue to put my appointment to my current job down as some kind of administrative error that is yet to be detected.

Once you’ve decided you’ve made the diagnosis then you’ll need a few basic principles of treatment. Firstly do a bit of resuscitation. There may well be some co existing sepsis so give some antibiotics. If they’re hypoxic give some oxygen. They may need some fluid or indeed they may be in congestive heart failure. The key is to do an assessment, this likely includes having a sneaky peak at the heart and the lungs with ultrasound. A commonly recommended treatment is propanolol to help with the tachycardia. Many patients will be hyperdynamic and tachycardic and giving a beta blocker may well be a good idea but giving a negative inotrope to someone who’s heart is a bit clapped out is generally considered bad form. The key message is to assess comprehensively and then decide.

For specific therapies, your list should include some steroids, this reduces the release of thyroid hormone from the gland. There is occasionally some coexisting adrenal insufficiency so you’ll treat that as well.  You’ll need to use something like PTU (propylthiouracil) or carbimazole in order to block new production of thyroid hormone. Good luck finding PTU at 3am. Having performed one miracle in locating PTU you are now expected to perform a further miracle and find something that sounds more like a tonic you’d buy from a wild west apothecary. This is of course “Lugol’s Solution”. Only give this once thyroid production has been blocked (recommendations suggest an hour afterward). It contains typically a bunch of iodine and will block the release of any T3/4 left in the gland.

Your next patient in your exam viva comes from the opposite end of the spectrum. Myxoedema coma. As an aside, myxoedema coma is a terrible name, the patient may not have oedema or be in a coma. Again, Farkas on the IBCC uses the term “decompensated hypothyroidism” which i think is much more descriptive and accurate. This is hypothyroidism but not as you’ve seen it before. Typical features or as Josh Farkas calls them “cognitive triggers” to consider myxoedema coma include

Again lots of these are non specific so keep the differential broad before anchoring too early. As expected TFTs will be helpful and in general expect to find a high TSH and low T3/4

Management will involve your usual assessment and resuscitation but the specific therapy here is IV thyroid hormone. It does exist but is also hard to track down at 3am. Your ICU probably has it and you’ve typically only seen it as part of management of a potential organ donor in brain death. T4 is the one typically recommended (which will converted intracellularly to the more active T3). But in Ireland the most commonly available will be T3.

These people also need steroid, typically some hydrocortisone for the ubiquitous adrenal insufficiency. Indeed giving thyroid hormone without steroid may cause an adrenal crisis.

Oh Chapter 61

Tasty Morsels of EM 130

The IBCC

Deranged Physiology

Welcome back to the tasty morsels of critical care podcast.

Today we’ll talk about one of the niche and shall I say “advanced” in inverted commas therapies in intensive care practice. ECMO. And to be precise we’ll be talking about VV ECMO. Indeed saying that you are “putting someone on ECMO” is a woefully incomplete sentence as the support and physiological difference between venovenous ECMO and venoarterial ECMO is really rather profound.

The post will be an intentionally broad description of the therapy and perhaps less on the nuances of managing a patient on VV ECMO, as at fellowship exam level I suspect you’d only be expected to have an overview of what it it is, what it can (and can’t do) and when to ask for it. I acknowledge the glaring gaps in the post and the likely criminal omission of the oxygen carrying capacity calculation. It would be fair to call this an idiot’s guide. And given that these posts are generated from my own notes then we all know who the idiot in that title it refers to is.

We’ll start at its simplest level, which is how i try to describe to friends and non medical people about how ECMO works. Blood is removed from the veins in one pipe and put through an artificial lung type device where CO2 is removed and Oxygen added, then blood is returned to the veins via a second pipe. If you’re lungs don’t work so well then the device can replace a lot of their function in the short term. Lay person explanation ends.

The degree to which we can replace lung function, primarily the degree to which we can oxygenate, is determined by the amount of the venous return coming back to the heart we can divert through the machine. Let’s say the cardiac output is a healthy 5L/min. That means that 5L/min is being ejected from the left ventricle and 5L/min is returning to the right ventricle. If the lungs aren’t working well then we need to capture at least 60% or so of this venous return and stick it through the oxygenator in order to maintain tolerable saturation of haemoglobin with oxygen. So in our example we’ll have to be siphoning off at least 3L/min from the venous return, putting it through the oxygenator and returning it back to the right side of the heart. With me so far?

It is at this stage that we immediately run into one of the physics challenges of VV ECMO. Pulling off 3L/min of blood requires pipes of substantial diameter. Typically these are in the 23 to 27Fr range. (ie 8-9mm internal diameter). You want to place this drainage pipe somewhere where there is a high flow of blood in a large vessel capable of accommodating it. Typically this will be in the SVC or the IVC, typically reached by an insertion point in the IJ or femoral vein respectively. It becomes really quite tricky to drain more than 3L/min of blood (or 60% of the venous return) with a single pipe as you can really only drain either the SVC (venous return from the upper body) or the IVC (venous return from the lower body) and as should be obvious the venous return from the body is split between these. In addition to the limitations of the physical size of the pipes you have to remember that the vessels within which these pipes are placed are not rigid fixed stented things, they dilate and contract in response to intravascular volume and intravascular tone. If you try to suck blood out of them with too much negative pressure the vessels will collapse around the pipe blocking all the holes and stopping all drainage.

All this to say that oxygenation is determined by the proportion of venous return we can divert through the ECMO machine. And capturing that venous return should be the priority when it comes to deciding on drainage pipe size and placement. once the blood is out of the body and through the oxygenator it turns out that it’s quite east to get it back into the budy. Pushing blood back into the body is much easier and can be done with a much smaller pipe. Pulling is harder than pushing in this context.

The key factor in returning blood back to the body in VV ECMO is 1) it has to return to a vein, hence the second V in VV ECMO and 2) it needs to return to the venous circulation at a healthy distance from the drainage pipe. It would be bad form to return 3L/min of beautifully oxygenated blood directly into the inlet holes of the drainage pipe and through the circuit for a second entirely pointless run. We call this re circulation and we get around it by placing the tip of our return pipe somewhat remote to the access pipe. For example we could drain blood from the SVC and return it to the IVC or if both of our pipes were in the IVC we could ensure that the return of blood happens much closer to the heart in the IVC or even in the RA but importantly a healthy distance away from where the drainage pipe is in the IVC.

Why would one want to initiate such a therapy? There are a number of indications, or should i say circumstances where VV ECMO might have a role. A reasonable list for examination purposes might run as follows.

There is lots of debate and indeed variations in practice on when you might initiate VV ECMO. There are a number of published criteria for when ECMO is indicated but you have to remember that a single PaO2 of 6.5kPa on 100% does not actually tell you how sick a patient is or if you’ve truly exhausted your conventional management. Either way at the end of a viva question on management strategies for severe ARDS, once you’ve been through high PEEPs, permissive hypercapnoea, diuresis, proning, and a nuanced discussion on steroids and nitric you should probably mention VV ECMO.

It’s worth noting some reasons when doing VV ECMO is not a great idea. As with almost all intensive care organ supports there’s not much point in adding it if you don’t have a way to fix the underlying organ. For example, if you have ARDS from pneumonia we can probably fix that, however if you have end stage COPD we can’t fix that. Adding the device will not change things. There are some programs who will use VV ECMO as a bridge to transplant but this is beyond the scope of this post. But it is commonly used in the immediate post lung tx phase when the new lungs are a bit heavy, wet and not working too well.

If your circulation is falling apart and you’re on 2mcg/kg/min of noradrenaline and the LV is clapped out from septic cardiomyopathy then it’s hard to see how fixing the hypoxic part of the multiorgan failure is going to turn things around. That being said if you’re hypoxic with a struggling RV then adding VV ECMO might be enough to correct the circulatory issues simply by fixing the hypoxia and hypercarbia.

If you wanted to end your SAQ with a flourish and add some complications then a convenient top 5 might run as follows

I am of course being a little bit facetious here but i’m not that far off the mark. Anticoagulation is typically used to keep the VV ECMO circuit running but even on those without anticoagulation they still bleed. And the bleeding is often spontaneously into non compressible sites like the pleura, GI tract or retroperitoneum.

Oh chapter 41

The ELSO red book

Alfred ECMO Site

Welcome back to the tasty morsels of critical care podcast.

Way back in the way back in tasty morsel number 43 we discussed inotropes and vasopressors but there was a noticeable AHD analogue shaped hole in that post that i promised to discuss at a future stage. Well, that time has come and it’s time to run through vasopressin.

You probably first encourntered vasopressin when you heard about ADH in medical school. Anti diruetic hormone, named for what it stops Its discussion in medical school involved delving into the world of endocrinology and negative feedback loops. Something we will be studiously avoiding here. Vasopressin is an ADH analogue, very simillar in structure with very similar effects. As such vasopresin exhibits the same ADH effects but this maxes out at very low doses, much lower than what we use in sepsis. At the very high doses we use, much higher than the pituitary can secrete, it acts as a pure pressor without the inotropic effect we’re use to when using more familiar agents like noradrenaline or adrenaline.

How does it work? Well this is where the fun beings. We’re used to messing around with the adrenergic receptors but vasopressin opens up a whole new bunch of confusing letters that have a whole myriad of effects. Some of these receptors are even shared with other molecules like oxytocin. The main we’re interested in is the V1 receptor, this is found throughout vascular smooth muscle. Stimulating it causes calcium release from the sarcoplasmic reticulum leading to increased vascular tone. Note noradrenaline has the same mechanism (ca release) just through a different receptor. This vasoconstriction affects pretty much all the vasculature including things like the coronaries (not so good) but does seem to spare the pulmonary arteries meaning it may be good in those with pulmonary hypertension.

What other receptors is it worth knowing about? both for exams and the all important one-upmanship on the ward round. V2 receptors are mainly in the renal collecting ducts, this is where we get the ADH effect primarily be increasing the number and effect of something called aquaporin 2 channels. The V3 receptor causes increased ACTH, increasing cortisol secretion, and then there are the OTR and P2 receptors which my notes make no elaboration upon and i will make the dangerous assumption that they have no relevance to what we do in ICM.

Why pull out the vaso when we can get the same vasopressor effect from our beloved noradrenaline. In theory the vasopressin receptors should remain fully funcitonal in the depths of horrific metabolic acidosis that has led your patient into intensive care, the same acidosis in theory should be causing issues with the effectiveness of your catecholamines. It should cause less pulmonary arterial constriction than a catecholamine and should even have less tachyphylaxis. the above list of advantages seems to come straight from the manufacturers advert, so why doesn’t it come pre attached to every patient?

The issue gets a bit clouded due to the somewhat clouded evidence base. I’m going to run through a few of the bigger name trials that one may trot out in a viva type setting, and with all good controversial issues in ICM you could easily go the track of “on the one hand this and the other hand that” and come up with an answer with both buttocks firmly on the fence of the issue.

First up is the VASST trial, (Russel et al 2008 NEJM). Done in North America and Oz, they enrolled septic patients and randomised them to vasopressin vs a blinded infusion of 15mcg/min of norad. Once maxed out on the study drug, then open label additional norad could then be titrated to keep the MAP at target. Enrolled 800 fairly typical ICU patients, and found a 35% vs 39% mortality benefit favouring the vaso but of course this was below the somewhat arbitrary statistical significance. A somewhat underwhelming start

Second is the VANISH trial by Gordon et al, JAMA 2016. This was closer to home in the UK, with 18 ICUs. Septic patients randomised similar to the VASST trial, vaso vs blinded norad, this time at 12mcg/min. Primary outcome here was kidney based rather than mortality. 400 pts here, no clear benefit for vaso. Again, hardly compelling

Enter the meta analysis. Nagendran 2019 in CCM. This was of a decent standard being not just a mix of numbers from the trials but an individual patient meta analysis that takes individual patient data points rather than the trial aggregate. This included all the trials but unsurprisingly VASST and VANISH make up most of the numbers. No mortality benefit found but there was less need for CRRT and less arrhythmias. There was some more digital ischaemia but no clear sign of increased mesenteric ischaemia.

Hence the “on the one hand this, on the other hand that” and widely varying opinions on use of vaso. It would be perfectly reasonable to say this drug adds nothing to usual practice and i’ll stick with my catecholamines. And lots of other reasonable people look at the data and say, well this is a catecholamine sparing agent and is a balanced approach to receptor manipulation and just might spare a few filters. As you can imagine (though please don’t actually imagine this) my buttocks remain firmly on the fence getting splinters.

Deranged Physiology

LITFL

Russell, J. A. et al. Vasopressin versus norepinephrine infusion in patients with septic shock. New Engl J Medicine 358, 877–87 (2008).

Welcome back to the tasty morsels of critical care podcast.

Today we’re going to verge into challenging territory for an audio podcast in that we’re going to the discuss the very visual topic of dynamic LV outflow tract obstruction. This is something fairly dependent on echocardiography for diagnosis which as you can imagine translates poorly to audio format.  This also means you’ll be denied my interpretative dance as i simulate the mitral valve leaflets being pulled over towards the septum via the Venturi effect. But alas i digress.

In essence dynamic LVOTO occurs when the closure point and tips of mitral valve tips are pulled into the left ventricular outflow tract during systole forming an anatomic obstruction to LV outflow thus reducing SV, CO, perfusion etc… This is reflected in poor blood pressure to which we respond by giving more catecholamines which makes this whole thing worse in a horrible cycle of nastiness.

Perhaps it’s best to start by identifying contexts where we should be on the look out for this. We’ll start with sepsis. Sepsis is a state of low systemic vascular resistance leading to reduced preload and afterload in the heart. The LV receives less than usual volume to stretch it and the low afterload makes it incredibly easy for the LV to empty itself of this load. This results in a small cavity LV where the LVOT and the mitral valve find themselves in much closer proximity than they are normally used to. If it gets out of hand bits of the mitral valve find themselves in the LVOT itself causing all kinds of bother.

The incidence of dynamic LVOTO in those with septic shock  is remarkably high and is reported to be  20% in one study from ICU echo guru Michel Slama. Even if it’s not that common it’s yet another reason why the super shocked patient should get a timely echo.

So let’s say we’re worried about our septic patient: within that cohort who is at risk? Classically it would be the older person with LVH or a thickened septal bulge, sometimes called a sigmoid septum. Going with that is a stiff ventricle that fills poorly and has diastolic dysfunction.

As noted at the beginning it’s clear that echo is a key part of the diagnosis here and if you do one you may see some of the baseline features just mentioned but with the addition of SAM or systolic anterior motion of the mitral valve. Most dynamic LVOTO has SAM but not all SAM has LVOTO. SAM can be quite a common out patient echo finding and so in addition to SAM you might want to look for flow acceleration. Just as a river approaching a narrow point accelerates and becomes turbulent so does blood flow in the LVOT approaching an unwelcome and intrusive mitral apparatus. This flow acceleration can be easily measured with doppler and produces characteristic patterns that get echo nerds like me all hot under the collar and is largely beyond the scope of the podcast.

Before we get onto management I want to mention another at risk cohort. These are usually easy to spot as they return from theatre with a big sternotomy following an AV replacement or mitral valve repair. To take the example of the aortic valve. Aortic stenosis leads to severe LVH as the LV has to generate an enormous pressure to get the crusty calcified stenotic valve to open. The heart slowly adapts and learns to live with this very high afterload. Then one day someone opens their chest and pops in a nice shiny new valve that opens like a dream. The LV is not used to this and continues to eject blood like pompeii on a bad day. This hyperdynamic contraction in an LV not used to it has a tendency to drag the mitral apparatus into the LVOT forming an obstruction. Patients following MV repair are also at risk as the change in shape of the annulus and final position of the coaptation point at end repair can also lead to the MV getting pulled into the LVOT. If you’ve ever done the TOE board examinations you will curse yourself learning the 8 echocardiographic risk factors for SAM post MV repair…

So we’ve talked about what it is, and a few at risk populations (sepsis and cardiac surgery). The clinical appearance is typically rapidly worsening shock and rising pressors, hypotension and rising lactate. Let’s say you’ve even managed to diagnose it using echo. How should you manage the thing?

Increasing the preload can help, typically best done with volume expansion. Part of the mechanism here is an empty LV so filling it up can help. You can of course increase preload with noradrenaline but it comes with the unfortunate side effect of inotropy which is how we got into this mess in the first place. As such you find yourself having to do something quite uncomfortable and reduce or stop your catecholamines even when you think they might be keeping the patient alive. If that’s not bad enough, you probably want to swap them out for…. eughhh, and i struggle to say this… phenylephrine… Much derided and neglected it may well have a role here and it does hold the title of “pure alpha” and as such gives pressor effect without inotropy. Vasopressin is of course a reasonable option but as you know it is not the most titratable drug.

The next step might make you even more uncomfortable. Understandably these patients are often very tachycardic. This tachycardia leaves less time for diastole, less time for cardiac filling, again worsening the LVOTO. As such beta blockers become a very attractive option. Even though beta blocking a very sick patient seems like a slightly insane idea. But of course you’ll only be doing this after a high quality, well interpreted echo that shows a hyperdynamic LV so that should give you a little reassurance. Esmolol is probably the agent of choice here given it’s titratability. The dosing can be tricksy as it’s in the mcg/kg/min range and is not a daily use drug for most of us. I tend to skip the loading dose. In other words look it up.

Usually with filling and removal of inotropy and chronotropy you can ride it out and you should see some improvement in your haemodynamics. A failure to respond to treatment should make you question your diagnosis, that high lactate may well be due to dead gut rather than low output due to LVOTO.

Welcome back to the tasty morsels of critical care podcast.

Following hot on the heels of tasty morsel number 72 on cardio renal syndrome is its partner in nephron injury: hepatorenal syndrome. This gets covered in a sub section of Oh’s manual chapter 44 on liver issues but there are a variety of other sources mentioned at the end that are worth a read.

It can be a little tricky to pin down this diagnosis. A lot of that comes because it is a “syndrome”, ie a collection of clinical findings that someone has put into a big bucket and mixed around without paying too much attention to hard core diagnostic information like histology or a true pathological diagnosis.

To start with we need context. We should have an AKI in the setting of advanced chronic liver disease and portal hypertension ie cirrhosis. But of course there are multiple reasons for AKI in this context so we have to work through them a little before the label of hepatorenal gets attached. Our friends in the international club of ascites (yes that’s a thing, i didn’t make it up) suggest that you need an AKI with a failure to respond to simple things like withdrawal of nephrotoxic agents, treatment of infection and, importantly a decent trial of albumin. You also have to exclude intrinsic renal diseases that lose protein and blood but this is usually fairly straightforward to exclude. However you can quickly see that a lot of this is pretty nebulous and it can be hard to really draw a line under. As such it’s fair to say that your patient may have several causes for their AKI in cirrhosis and hepatorenal may only be part of the problem.

To take hepatorenal per se, what’s the purported pathogenesis? Well we think that increasing portal venous pressures and cirrhosis leads to splanchnic vascular vasodilation. The vessels in our gut lose tone and we develop this chronic high output, low SVR state. This state of reduced pressure leads to activation of the RAAS causing increased resistance in the renal arteries (in distinction to the very low resistance state of the splanchnic vasculature). As such, perfusing pressure to the glomerulus falls and GFR falls. This is reflected in oliguria and the usual renal response to a crisis of hanging onto Na for all its worth with a urine Na typically <10 if you do go looking for it. In the background you’ve got all this chronic ascites that is adding to the compartment pressure in the abdomen making things worse. That’s the basic bedtime story version of the pathophys that i received, I understand there’s a competing theory where the chronic bacterial translocation of a leaky liver leads to a chronic inflammatory process buggering the kidneys but i digress.

At this stage, it’s worth noting that just like our cardiorenal syndrome we can split hepatorenal into a couple of types. This is all about timing of onset. The in-patient with a rapidly rising creatinine in the hospital setting is more likely to have type I or acute HRS while the stable out patient cirrhotic with a gradually rising creatinine is going to have the type II or chronic HRS.

HRS itself can be precipitated by the usual chronic liver disease decompensations – ie , bleeding, infection, SBP and also large volume paracentesces without appropriate albumin replacement.

How should we treat. First off an important reminder that cirrhosis is not a reversible pathology and if you’re decompensating then the only real treatment to turn the whole thing around is a transplant. All the rest of it is a little bit like rearranging the deck chairs on the Titanic. I don’t mean to imply that the deck chairs should not be rearranged; rearranging the deck chairs may well lead to some meaningful short term outcomes and quality of life improvement but the hole in the hull from the iceberg isn’t going to get any better without the transplant.

As mentioned in the diagnostic bit, they probably should’ve got some albumin. I don’t deny that albumin has a limited role in the ICU, gone are the days where we could make bold assertions about colloid osmotic pressures and how albumin was better than crystalloid. We studied it and it turns out it doesn’t seem to be any better than salty water. But i still think that in the livers it still has an important role and HRS is probably one of those scenarios. The data for its use in HRS is not exactly stellar, and no where near as good as it is say for albumin in ascites drainage in SBP where it has a mortality benefit.

I confess i have done the sneaky move “push the MAP up to 75 mmHg” for a few days to see if that makes a difference but i can’t say I’m standing on solid ground doing that but it does get a mention in the relevant UpToDate article.  Of course given that we have access to terlipressin in Ireland and the UK, these types of haemodynamic manipulations could even be done on the ward before rushing to commit yourself to CRRT and an ICU admission.

TIPS has been described to be helpful in improving renal function in HRS but only on a very limited basis so probably worth a discussion with your local liver experts but not something you can use evidence to make a compelling argument for.

If the HRS fails to respond to such therapies then the option of CRRT is usually raised. This will indeed replace the failing kidney but like many things in intensive care, there’s not much point in starting it if there’s no chance of reversibility. Some HRS may be in the context of something like acute alcoholic hepatitis where the chance of recovery of some hepatic function is reasonable and hence a trial of the CRRT while the liver settles is reasonable. In addition if the patient is already on the transplant list then keeping them going with CRRT pending the transplant would also be reasonable. However the more common scenario is the burnt out cirrhotic who is having increasing frequency of decompensations. The prognosis here is dismal and adding more machines at the end of life is probably not going to help. The tricky bit for us is teasing out so much of what we’ve discussed above, eg is this one of the many other eminently reversible causes of AKI in the context of cirrhosis or is this all driven by the chronic progressive liver failure. This as you’ve probably discovered by now is a core part of ICM and best learnt in the unit and on the wards rather than on the podcast.

ICA Guidelines

Deranged Physiology

LITFL CCC