Cardiac Output

Eight o'clock ward round, day two after a coronary artery bypass, and the monitor shows atrial fibrillation at a hundred and fifty. Calum's answer is amiodarone — and that is the single most common wrong answer on the unit. Not because amiodarone is wrong, but because of where it sits in the order.
This is atrial fibrillation on the cardiothoracic intensive care unit, top to bottom. And it matters beyond the bedside: in a recent Fellowship paper it was the worst-performing question on the whole exam, with a pass rate of about 43 per cent. The examiners were specific about where candidates fell down — the management, and the anaesthetic for elective DC cardioversion. The bit everyone assumes is easy.
We start with how common it is — about a third of patients after coronary surgery, up to half after valve or combined procedures, peaking on days two and three — and why it isn't benign. Then a framework you can use on a ward round: substrate and trigger. The substrate is atrial; the triggers are catecholamines, pain, hypovolaemia, hypoxia, anaemia, electrolytes and the inflammatory response to bypass. Almost every trigger is something you can change.
Then the four strands of management — rhythm, rate, recurrence and thromboembolism — and the step that comes before all of them. Correct the physiology first: potassium, magnesium, pain, volume; exclude tamponade and pneumothorax with a scan and a film, not a glance at the drain; look for sepsis; and check for a misplaced epicardial pacing wire irritating the atrium.
Only then a drug. Amiodarone as the workhorse — why it acts on both rhythm and rate, and why the slow loading dose is a safety feature rather than fussiness in a shocked post-cardiotomy patient. Then the drugs we avoid, and the specific reason for each: digoxin, verapamil and diltiazem, flecainide and the lesson of the CAST trial, beta-blockers (and why esmolol if you must), and adenosine as a diagnostic rather than a treatment. Plus the Vaughan Williams classification in ninety seconds, and when an unstable patient needs a synchronised shock rather than a drug.
Anticoagulation next: our local position on new-onset post-operative AF and the reasoning behind it, anticoagulant mechanisms and reversal agents, the case for and against the direct oral anticoagulants, and neuraxial timings. Then a trap specific to cardiac surgery — andexanet alfa can look exactly like heparin resistance on bypass, which is why you tell the perfusionist before you reverse, not afterwards.
We finish with the section the examiners said was done worst — the anaesthetic for a DC cardioversion list, including what to avoid and the bailout you set up before the first patient — and the long-term options for rhythm, rate and stroke prevention.
Chapters
(00:00) Cold open — AF at 150, and the commonest wrong answer
(00:50) Why this was the worst-answered exam question
(01:20) Incidence, and why it isn't benign
(02:10) Substrate and trigger
(02:40) Causes of AF in general
(03:30) The four strands of management
(03:50) Correct the physiology first
(04:20) The misplaced pacing wire
(04:50) Amiodarone, and why the slow load matters
(06:30) The drugs we avoid — and why
(08:00) The Vaughan Williams classification
(08:50) The unstable patient
(09:20) Anticoagulation: our local position
(10:00) Mechanisms and reversal
(11:00) For and against the DOACs
(11:30) Neuraxial timings
(11:50) Reversal before bypass — andexanet and apparent heparin resistance
(13:10) Anaesthesia for DC cardioversion
(14:30) Long-term rhythm, rate and stroke prevention
(15:10) Wrap-up
Key takeaways
Post-operative AF affects about a third of patients after CABG and up to half after valve or combined surgery, peaking on days two to three — often self-terminating, but not benign
Think substrate (atrial trauma, inflammation, stretch, fibrosis, a large left atrium) and trigger (catecholamines, pain, hypovolaemia, hypoxia, anaemia, electrolytes, the inflammatory response)
Four strands of management: restore rhythm, control rate, prevent recurrence, prevent thromboembolism
Correct the physiology first: K⁺ 4.5–5.5, Mg²⁺ above 1.0, treat pain and hypovolaemia, exclude tamponade and pneumothorax, look for sepsis, and check the pacing wires
Amiodarone 300 mg over 20–60 minutes (never as a rapid bolus), centrally where possible, then 600–900 mg over 24 hours — rapid loading causes negative inotropy and vasodilation
Digoxin is not an acute agent; verapamil and diltiazem are negatively inotropic; flecainide is contraindicated in structural heart disease and ischaemia (the CAST lesson); esmolol if you must beta-block; adenosine is diagnostic, not therapeutic
Adverse features — shock, syncope, myocardial ischaemia, heart failure — mean synchronised DC cardioversion, not a drug; know the Resuscitation Council adult tachycardia algorithm
Know anticoagulants by mechanism, because reversal follows mechanism
Reversing a DOAC before bypass? Tell the surgeon and the perfusionist first — andexanet sequesters heparin–antithrombin and mimics heparin resistance
Cardioversion list: pre-oxygenate, titrate propofol, support the airway with spontaneous ventilation, avoid midazolam, volatiles and TCI, and set out the bailout (rocuronium, intubation, sugammadex) before you start
Doses, thresholds and anticoagulation practice are Wythenshawe-specific — check your own guidelines
References / further reading
Resuscitation Council UK. Adult tachycardia algorithm. Resuscitation Guidelines 2021
Van Gelder IC et al. 2024 ESC Guidelines for the management of atrial fibrillation. Eur Heart J 2024
National Institute for Health and Care Excellence. Atrial fibrillation: diagnosis and management (NG196). 2021
Echt DS et al. Mortality and morbidity in patients receiving encainide, flecainide, or placebo: the Cardiac Arrhythmia Suppression Trial. N Engl J Med 1991; 324: 781–8
Horlocker TT et al. Regional anesthesia in the patient receiving antithrombotic or thrombolytic therapy: ASRA evidence-based guidelines (4th edition). Reg Anesth Pain Med 2018; 43: 263–309
Royal College of Anaesthetists. Final FRCA written examination — chair's reports
Follow the podcast
Bluesky: @cardiacoutput.bsky.social
X: @CardiacOutputMC
If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a topic you'd like covered.
This podcast is for medical education for healthcare professionals. It is not clinical advice. Medications, doses and practice described reflect local Wythenshawe practice at the time of recording — always follow your own centre's guidelines and current local policy.

Last episode we put a probe on the chest. This one goes down the oesophagus — and it starts somewhere uncomfortable.
Inserting a transoesophageal probe is something we do to a patient that can cause them significant harm. Maybe once a year, sometimes more, a patient comes to harm because an anaesthetist put a probe down. We do a great many of these, so it is not unreasonable to think we sometimes cause damage and never find out. But sometimes the injury is severe: an oesophageal tear, upper gastrointestinal bleeding. Severe harm runs at roughly one in two thousand, from a UK audit by the Association of Cardiothoracic Anaesthetists, with female sex, steroid use and a low body mass index as risk factors — all observational, so hold it loosely.
And then the argument that follows from it. Contraindications are genuinely hard to find, and Mike's position has changed over time: transoesophageal echo should be mandatory for essentially all cardiac surgery unless there is a good reason not to. Not because every operation demands it, but because of what turns up — valvular problems and dissections in patients listed for bypass grafts with supposedly normal hearts, things that would never have been found without scanning them on the table.
From there, the practical half. Three ways to get an ejection fraction, in ascending order of rigour: the eyeball, which in experienced hands correlates well and is what you'll actually use in a crashing patient; fractional area change from the transgastric short axis; and Simpson's biplane, the most rigorous and the most dependent on an endocardial border you can genuinely see. Garbage in, garbage out.
Then the right ventricle, and an argument worth having: there are many accepted ways to assess it, and the very existence of so many tells you that none is good enough alone. Watch TAPSE being measured in theatre and ask honestly whether that is a precise, scientific number — particularly off-axis on a transoesophageal probe. Most of the time what you are doing is judging performance by eye and following the trend: how it looked before, how it looks now, what you did in between, and whether it worked. Plus the levers when the right ventricle is struggling — rate and rhythm, preload, afterload, contractility and coronary perfusion, and what comes after them.
Systolic anterior motion gets done properly this time, and the key idea is that echo lets you predict it rather than just diagnose it: a long anterior mitral leaflet, a small hypertrophied ventricle with a narrow outflow tract, and a short coaptation-to-septum distance. Then the management, which runs almost exactly opposite to instinct.
Then the probe as a theatre instrument. Siting a balloon pump tip just distal to the subclavian — and what goes wrong if it sits too high or too low. The Protek Duo, and why its tip must sit beyond the pulmonary valve. And the rule worth carrying out of the whole topic: any patient on mechanical support with haemodynamic instability does not have to do very much to earn a scan.
The section Mike calls most underrated is the one where you help the surgeon and the perfusionist. The venous drainage cannula that has slipped into a hepatic vein — presenting as low flows and poor drainage that everyone assumes is volume — and the non-standard view that finds it. Confirming the retrograde cardioplegia cannula in the coronary sinus, which protects the heart for the entire operation. And watching the wires: the femoral venous wire in peripheral bypass, and the balloon pump wire in the descending aorta, with a live commentary nobody else in the room can give.
We finish with transplantation — what is and isn't useful to scan, why a heart that has just been ischaemic cannot be judged like a normal one, and the three questions all of this monitoring exists to answer before you take that patient out to the unit.
Chapters
(00:00) Cold open — the probe can hurt people
(01:00) One in two thousand, and the risk factors
(01:40) Contraindications, and how few there are
(02:20) Who should get one — and why the answer changed
(03:20) Accreditation for cardiac anaesthetists
(03:50) Ejection fraction: eyeball, FAC and Simpson's biplane
(05:00) The right ventricle, and why so many methods is a warning
(06:20) The levers when the right ventricle is failing
(07:00) Predicting systolic anterior motion before it happens
(08:00) Treating it when it happens
(08:40) Siting a balloon pump
(09:20) The Protek Duo
(10:10) Instability on support earns a scan
(11:00) Which pipes you can actually see
(11:50) Helping the surgeon: the cannula in the hepatic vein
(12:50) Confirming retrograde cardioplegia in the coronary sinus
(13:20) Watching the wires
(13:50) Transplantation
(15:00) Wrap-up
Key takeaways
Putting a probe down can seriously harm a patient: severe harm is around 1 in 2000, with female sex, steroids and low BMI as observational risk factors
Contraindications are few — oesophageal pathology or swallowing difficulty, and a pharyngeal pouch is difficult but not absolute
TOE should be mandatory for essentially all cardiac surgery unless there's a good reason not to, because we find valve lesions and dissections in hearts we believed were normal
Ejection fraction three ways: visual estimation, fractional area change from the transgastric short axis, and Simpson's biplane — the most rigorous and the most dependent on image quality
The number of accepted methods for assessing the right ventricle is itself evidence that none is good enough alone — use the trend and your eyes
When the right ventricle struggles, work the levers: rate and rhythm, preload, afterload, contractility, coronary perfusion — then mechanical support
Predict systolic anterior motion before bypass comes off: long anterior mitral leaflet, small hypertrophied ventricle, narrow outflow tract, short coaptation-to-septum distance
Treat it by filling, decreasing the rate, stopping the inotropes and adding a vasoconstrictor
Balloon pump tip just distal to the left subclavian: too high risks the arm and brain vessels, too low risks the mesenteric and renal arteries and won't augment properly
The Protek Duo tip belongs beyond the pulmonary valve, in the pulmonary artery
Any patient on mechanical support with haemodynamic instability earns a transoesophageal echo — look at flows, septal position, ventricular size and the position of the pipes
A venous drainage cannula that has slipped into a hepatic vein presents as low flows and poor drainage that everyone assumes is volume — and one non-standard view finds it
Confirming the retrograde cardioplegia cannula in the coronary sinus protects the heart for the whole operation
A freshly transplanted, recently ischaemic heart has a fixed stroke volume, is highly rate-dependent and is on inotropes — you are interpreting a picture that is abnormal by design
Everything you are monitoring answers one of three questions: is this patient stable, in a low cardiac output state, or sliding into one?
References / further reading
Association of Cardiothoracic Anaesthesia and Critical Care. UK audit of transoesophageal echocardiography probe-related injury
Lambert AS, Allen SJ, Sidhu S, eds. Practical Perioperative Transoesophageal Echocardiography, 4th edn. Oxford: Oxford University Press, 2025
Hahn RT et al. Guidelines for performing a comprehensive transesophageal echocardiographic examination: recommendations from the American Society of Echocardiography and the Society of Cardiovascular Anesthesiologists. J Am Soc Echocardiogr 2013
European Association of Cardiovascular Imaging / EACTAIC. Transoesophageal echocardiography certification
British Society of Echocardiography. Transoesophageal echocardiography accreditation
Follow the podcast
Bluesky: @cardiacoutput.bsky.social
X: @CardiacOutputMC
If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a topic you'd like covered.
This podcast is for medical education for healthcare professionals. It is not clinical advice. Practice described reflects local Wythenshawe practice at the time of recording — always follow your own centre's guidelines and current local policy.

Three in the morning. A patient on the unit is hypotensive, the noradrenaline is going up, and somebody has already given a litre. What is the fastest way to actually know what's wrong?
Put a probe on them.
This is the first of two foundation episodes on echocardiography — the thing that turns a guess into a diagnosis, at the bedside, in about ninety seconds. Today it's transthoracic echo on the intensive care unit: the views, and the patterns you are looking for in a shocked patient. Next time we take the transoesophageal probe into theatre.
We start with accreditation, because Mike's advice is to start early. The routes come in two shapes and it helps to see that first: a mentored portfolio, or an examination. FICE is the portfolio route — an approved basic course, a logbook of fifty studies with at least ten directly supervised, all fifty reviewed with your mentor, and a triggered assessment, with the first and last scans no more than twelve months apart. EDEC is the examination route — a hundred transthoracic and thirty-five transoesophageal cases, a mentor and a supervisor, and a formal exam you can start once you've done thirty and ten. Plus the machine you actually need: a cardiac phased-array probe, 2D and M-mode, colour and spectral Doppler, ECG gating — and proper image archiving, for three reasons most people haven't thought about.
Then the five views that will get you a very long way. Parasternal long axis for overall size and function and the mitral and aortic valves. Parasternal short axis — the doughnut — which at papillary muscle level is the single best view for regional wall motion and for the shape of the septum. Apical four-chamber for comparing the ventricles, for TAPSE, and for Doppler through the mitral and tricuspid valves. Subcostal, the one people neglect and the one that works when nothing else does, because it doesn't care about ventilation, chest drains or dressings — and the best view for a pericardial effusion. And the inferior vena cava from there.
Then six patterns. Hypovolaemia, with its hyperdynamic ventricle, end-systolic effacement and kissing walls. Low afterload, which can look almost identical — and why that means echo is never read in isolation. Right ventricular failure, usually obvious the moment the probe goes on: a dilated right ventricle squashing a small, underfilled left one. Tamponade, with diastolic collapse of the right atrium and ventricle. Dynamic left ventricular outflow tract obstruction and systolic anterior motion — which looks like a mitral valve problem and is really a haemodynamic one, where the management runs almost exactly opposite to your instincts. And type A dissection: a normal-looking heart with a flap in the root.
The case we'd want every new starter to know is the aortic valve replacement done for aortic stenosis. That hypertrophied ventricle now ejects vigorously through a brand-new valve, empties, and becomes profoundly hypovolaemic. The vasopressor requirement climbs, somebody quite reasonably asks for an echo to exclude tamponade — and the scan shows hypovolaemia instead. That is echo earning its keep: not confirming what you suspected, but telling you that you were looking at the wrong thing.
We finish on fluid responsiveness, with some honest context about why it matters less than it used to, the real pitfalls of the vena cava, and why lung ultrasound and venous congestion assessment tell you more than any single number.
Chapters
(00:00) Cold open — 3am, hypotensive, a litre already in
(00:40) Why accreditation is worth starting now
(01:30) Two shapes of pathway: FICE and EDEC
(03:00) The machine you actually need — and why archiving matters
(04:00) The five views: parasternal long axis
(04:40) Parasternal short axis — the doughnut
(05:30) Apical four-chamber
(06:00) Subcostal, and the inferior vena cava
(06:50) Hypovolaemia: kissing walls and a collapsed cava
(07:40) Why low afterload looks the same
(08:20) Right ventricular failure
(09:00) Tamponade
(09:40) Dynamic LVOT obstruction and systolic anterior motion
(10:30) Type A dissection
(11:00) The classic case: the scan you ordered for tamponade
(12:00) Fluid responsiveness, honestly
(13:00) Wrap-up
Key takeaways
Echo turns a guess into a diagnosis at the bedside in about ninety seconds — put the probe on early
Start accreditation now: FICE is a mentored portfolio route, EDEC is examination-based, and both take time. Requirements change, so check the current version
Your machine needs a cardiac phased-array probe, 2D and M-mode, colour and spectral Doppler, ECG gating — and image archiving, without which you have no logbook, no comparison and no record
Five views will take you a long way: parasternal long axis, parasternal short axis, apical four-chamber, subcostal, and the inferior vena cava
The parasternal short axis at papillary muscle level is the best single view for regional wall motion and for septal shape
The subcostal view is the rescue view — it works when nothing else does, and it's the best view for a pericardial effusion
Hypovolaemia is a hyperdynamic ventricle with end-systolic effacement and a small, collapsing cava
Low afterload can look almost identical, so read the scan alongside the clinical context and the vasopressor requirement
Right ventricular failure is a dilated right ventricle squashing a small underfilled left one, with a displaced septum — usually obvious by eye
Tamponade shows diastolic collapse of the right atrium and right ventricle, with normal left ventricular size and function
Systolic anterior motion looks like a mitral problem but is haemodynamic — fill them and get the rate/inotropy down
After an aortic valve replacement for stenosis, expect the scan you ordered for tamponade to show you hypovolaemia
Treat fluid responsiveness with humility: the vena cava has real pitfalls, and lung ultrasound and venous congestion assessment tell you more than any single number
References / further reading
Intensive Care Society. Focused Intensive Care Echocardiography (FICE) accreditation
European Society of Intensive Care Medicine. European Diploma in Advanced Critical Care Echocardiography (EDEC)
British Society of Echocardiography. Personal and departmental accreditation
Lancellotti P et al. The use of echocardiography in acute cardiovascular care: European Association of Cardiovascular Imaging and Acute Cardiovascular Care Association recommendations
Mitchell C et al. Guidelines for performing a comprehensive transthoracic echocardiographic examination in adults: recommendations from the American Society of Echocardiography. J Am Soc Echocardiogr 2019
Vieillard-Baron A et al. A decade of progress in critical care echocardiography: a narrative review. Intensive Care Med 2019
Follow the podcast
Bluesky: @cardiacoutput.bsky.social
X: @CardiacOutputMC
If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a topic you'd like covered.
This podcast is for medical education for healthcare professionals. It is not clinical advice. Accreditation requirements change — always check the current version with the relevant body, and follow your own centre's guidelines.

Recorded minutes after the live launch at Annual Congress 2026, and published the same day the paper appears in Anaesthesia.
GLIMPSE — the GLP-1 receptor agonist Management in the Peri-operative Setting project — is a national, prospective, multicentre cohort study delivered through the UK resident research networks across all four nations, and it is the largest thing anyone has done on this question. Mike and Calum take it apart the way they did ITACS: the question, the methods, the results, the critique. Then the half that wasn't in the room — what it means for cardiothoracic anaesthesia and intensive care.
Two numbers to start. One in 36 patients presenting for anaesthesia in the UK is taking a GLP-1 receptor agonist. And in those patients, the incidence of pulmonary aspiration or regurgitation was 1 in 71, against 1 in 802 in everyone else — an odds ratio of about 11.
Then the work of deciding how much of that to believe.
The framing that governs everything: the primary outcome is a prevalence, not a comparison. GLIMPSE was designed to answer "how many?" The aspiration finding is a secondary outcome, and every conclusion about the odds ratio has to sit inside that. We cover the design — each site choosing its own fortnight, screening every patient's drug history — the exclusions, the deliberate discarding of twelve sites that screened fewer than three quarters of their eligible patients, and the 97% screening rate that makes this a remarkably complete national snapshot.
The prevalence findings are more interesting than the headline. Between institutions, use ranged from 0.4% to 7.8% — a twenty-fold difference — so the national average is close to useless for planning your own list. Tirzepatide accounts for 70% and semaglutide 26%, and this is now predominantly a weight-loss population rather than a diabetes one. And the number that should change your practice this week: 41% obtained the drug from somewhere that is neither a GP nor a hospital specialist. It will not be on the GP summary, the discharge letter or the drug chart. It is only there if somebody asks the patient directly, by name.
Management was highly variable. Thirty per cent were asked to stop, most for 8–14 days — consistent with "stop it for a week", which the paper quietly demolishes: with half-lives of five and seven days, five half-lives means 25 and 35 days. A week does not eliminate the drug, and national consensus guidance says to continue it anyway. Meanwhile gastric ultrasound — the one bedside test that answers precisely the question being asked — was used in 1.3% of patients.
The safety finding gets the full appraisal, including the qualifier that changes how you read it: of the 19 events in the exposed group, all 19 involved regurgitation and only two reached the lung, which suggests the intubation, head-up positioning and pre-oxygenation were doing their job. And the finding we'd most like to change practice: more than half the events happened at emergence, not induction.
The critique covers association without any ability to adjust for confounding, observer bias inflating the ratio from both ends (with the NAP7 cross-check that partly defends it), small event numbers driving a large effect, and generalisability. The strengths get their due too.
Then the cardiothoracic half, which the paper does not address. Why our prevalence is probably higher than the national figure. Why the months between assessment and surgery make the drug-history gap worse for us. Why emergence happening on the intensive care unit, hours later, makes this a handover problem as much as an airway problem.What it means that we put a probe into the stomach of nearly every patient we anaesthetise — and that we already own the skill that was used in 1% of cases nationally. The double-lumen tube versus tube-exchange dilemma, taken from a real case in the paper. The conflict between a rapid sequence induction and haemodynamic stability in critical aortic stenosis. Why stopping the drug costs our patients more, through hyperglycaemia and sternal wound infection. And the exclusion that matters most to us: GLIMPSE excluded patients whose airway management began elsewhere — ICU transfers explicitly — so our sickest patients are entirely outside the dataset.
Chapters
(00:00) Cold open — outside Hall 1A, minutes after the launch
(00:50) The two headline numbers
(01:30) Why the question needed asking
(02:40) Methods: design, sites and the fortnight
(03:30) The exclusions — including ICU transfers
(04:10) The primary outcome is a prevalence, not a comparison
(05:00) Powering, and discarding twelve sites on purpose
(05:50) One in 36 — and the twenty-fold variation between hospitals
(07:00) Which drugs, and why people were taking them
(07:40) Where they got them: the 41% that isn't on any record
(08:40) What clinicians did: stopping for a week
(09:40) Why a week is pharmacologically meaningless
(10:30) Airway management — and gastric ultrasound at 1%
(11:40) The safety outcome: 1 in 71 against 1 in 802
(12:40) Nineteen events, two aspirations — the mitigation worked
(13:30) Emergence, not induction
(14:20) The critique, in four parts
(16:20) The strengths it deserves
(17:00) Cardiothoracic: our prevalence, and the pre-assessment gap
(18:00) Emergence on the unit — a handover problem
(18:50) The probe we already own
(19:30) Double-lumen tube, or exchange?
(20:00) Rapid sequence versus haemodynamics in critical AS
(20:30) Whether to stop the drug at all
(21:00) The exclusion that leaves out our sickest patients
(21:30) Wrap-up
Key takeaways
1 in 36 patients presenting for anaesthesia in the UK is taking a GLP-1 receptor agonist — but between hospitals the rate ranged from 0.4% to 7.8%, so find out your own number
The primary outcome was prevalence; the aspiration comparison is a secondary outcome and should be read as such
41% obtained the drug from neither a GP nor a hospital specialist — ask patients by drug name, at assessment and again on the day
Tirzepatide 70%, semaglutide 26%; 97% on weekly injections; 59% taking it for weight loss
Aspiration and/or regurgitation occurred in 1 in 71 of exposed patients against 1 in 802 of the unexposed, odds ratio 11.4 (95%CI 6.5–20)
All 19 events in the exposed group involved regurgitation; only two were pulmonary aspiration, giving an aspiration rate of about 1 in 674 — the mitigation appears to have worked
More than half the events occurred at emergence, not induction — the extubation plan matters at least as much as the intubation plan
Stopping for a week is pharmacologically meaningless against half-lives of five and seven days; national consensus guidance recommends continuing
Gastric ultrasound was used in 1.3% of patients — a skills gap, and an open goal for a specialty that already images the stomach
This is an association from a study that could not adjust for confounding, in which observers knew exposure status — hold it firmly but not tightly
For cardiac patients, emergence usually happens on the ICU hours later, which makes GLP-1 use a handover item
Stopping the drug in cardiac surgical patients risks hyperglycaemia and its associated sternal wound infection, without reliably emptying the stomach
GLIMPSE excluded patients whose airway management began outside the procedure location, so the critical care population is not represented
References / further reading
Potter TE, Cronin JN, Kua J, et al., on behalf of the GLIMPSE Resident Research Networks and Collaborators. Peri-operative glucagon-like peptide-1 receptor agonist use and outcomes: a national prospective multicentre cohort study. Anaesthesia 2026. doi:10.1111/anae.70380
Annual Congress 2026, Association of Anaesthetists. Session T8a, Peri-operative medicine — "GLIMPSE: … live launch and results", Dr Tom Potter, Thursday 17 September 2026, 16:35–16:55
Association of Anaesthetists and colleagues. Multidisciplinary consensus statement on the peri-operative management of patients receiving GLP-1 receptor agonists
Royal College of Anaesthetists. 4th National Audit Project (NAP4)
Royal College of Anaesthetists. 7th National Audit Project (NAP7): activity survey and peri-operative cardiac arrest
Follow the podcast
Bluesky: @cardiacoutput.bsky.social
X: @CardiacOutputMC
If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a paper you'd like taken apart.
This podcast is for medical education for healthcare professionals. It is not clinical advice. Discussion of published research reflects our own reading and interpretation — always read the primary source and follow your own centre's guidelines.

Before we touch a single image, we sort out the vocabulary — because almost everyone muddles these devices, and once you have muddled them you cannot reason about them. People say "VAD" and mean five different machines, sitting in different places, for different lengths of time.
So: three questions, and every device answers all three. Which ventricle? How long is it meant to stay in? Surgical or percutaneous?
This is part two of two, following the structure of the mechanical support chapter from the fourth edition of Practical Perioperative Transoesophageal Echocardiography. Part one covered balloon pumps and ECMO.
Every assist device has the same four components — a pump providing continuous flow, an inflow cannula, an outflow cannula, and an external controller connected by a driveline — and all of them sit in parallel with the native circulation. An LVAD drains from the left ventricular apex and returns to the ascending aorta end to side; an RVAD drains from the right atrium and returns to the pulmonary artery. Temporary means days to weeks, durable means weeks to years, and those are genuinely different machines rather than the same one left in longer. The predominant durable device worldwide is now the HeartMate 3.
Then the word that causes the most trouble: BiVAD. Two quite different situations get called the same thing. A true biventricular assist device means two devices implanted for long-term biventricular support, and that is occasional. Whereas a patient receiving a durable LVAD not infrequently needs a temporary RVAD at the same operation, because the right heart cannot cope — two devices, but one durable and one coming out. BiVAD describes the anatomy and tells you nothing about the timescale, which is precisely why people get confused.
With that grid in place, the rest follows the chapter. Before implantation: ventricular assessment, intracardiac thrombus, the valves, the aorta, and intracardiac shunts. Shunts matter because of a pressure change you are about to create — once support starts, left atrial pressure falls below right atrial pressure, so a patent foramen that has been silent for a lifetime can shunt right to left, giving arterial hypoxaemia or systemic embolisation of right-sided gas or thrombus. And because a foramen can be genuinely hard to detect before support starts, you re-examine the atrial septum afterwards.
Aortic regurgitation gets the attention it deserves, because it is the lesion that quietly destroys LVAD efficacy: blood leaves the outflow cannula, flows back across the incompetent valve, and returns straight into the inflow cannula — a circuit inside the chest that never reaches the patient. How you correct it depends on the intention of the device, why a bioprosthesis is preferred if replacement is needed, why colour Doppler underestimates it in end-stage failure and you should therefore assess on bypass, which measures work and which do not — and a free clue that costs nothing, in the left ventricular vent flows during implantation. Plus the exception on aortic stenosis that matters only for partial-support devices, and the mitral and tricuspid lesions corrected at the same operation.
After implantation: de-airing, and then the single most useful structure on the screen. The interventricular septum should be flat and neutral. Marked rightward displacement means the ventricle is inadequately decompressed — underpumping. Marked leftward means it has collapsed — the suction event, or ventricular suckdown. That is how you run a ramp study, and it is the same logic you will use later for the Impella.
Right ventricular function afterwards gets four reasons why it does not always improve despite reduced afterload, the echo signs of acute right ventricular failure, and then — in its proper place — temporary right ventricular support: either a surgically grafted CentriMag, or the Protek Duo, a percutaneous dual-lumen cannula from the right internal jugular with its inflow in the right atrium and its outflow in the proximal main pulmonary artery.
We then cover aortic valve opening and the HeartMate 3's programmed rhythmic flow changes, cannula assessment including the biplane tip for confirming inflow orientation, early and late causes of obstruction, why you reinterrogate both cannulas at chest closure, how to read the echo alongside the console, and the surveillance schedule.
The episode closes with temporary ventricular assist devices as their own topic — the Impella. Indications, the models and their licensed durations, the axial Archimedes screw, the left- and right-sided configurations, the contraindications to rule out first, the insertion sequence including the measurement trap of the pigtail tip, and the ongoing assessment, which comes back to exactly the same septum.
Chapters
(00:00) Cold open — sorting out the vocabulary first
(00:50) The four components every device shares
(01:20) Which ventricle, and for how long
(02:30) BiVAD — two situations, one word
(03:20) Surgical or percutaneous
(04:00) The rule that runs underneath, and where this comes from
(04:40) Before implantation: the five headings
(05:10) Intracardiac shunts, and the pressure change you create
(06:30) Aortic regurgitation, and recirculation inside the chest
(07:40) Why you grade it on bypass
(08:40) The vent flow clue
(09:10) Aortic stenosis, mitral stenosis and tricuspid regurgitation
(10:00) Intracardiac thrombus
(10:30) Right ventricular function, predictors and the aorta
(11:40) De-airing
(12:20) The septum: flat, rightward, leftward
(13:40) Ramp studies
(14:30) Why the right ventricle doesn't always improve
(15:40) Signs of acute right ventricular failure
(16:20) Temporary RVAD: the CentriMag and the Protek Duo
(17:30) Aortic valve opening, and the HeartMate 3 artificial pulse
(18:10) Assessing the cannulas, and the biplane tip
(19:20) Obstruction, early and late
(20:00) Reading the console alongside the echo
(20:40) Complications, tamponade and surveillance
(21:30) Temporary devices: the Impella
(22:10) Contraindications and insertion
(22:50) Ongoing assessment and what goes wrong
(23:20) Wrap-up
Key takeaways
Classify every device by three questions: which ventricle, how long, surgical or percutaneous
All assist devices share four components — pump, inflow cannula, outflow cannula, and an external controller on a driveline — and sit in parallel with the native circulation
LVAD: LV apex to ascending aorta, end to side. RVAD: right atrium to pulmonary artery
Temporary means days to weeks; durable means weeks to years. The HeartMate 3 is the predominant durable device worldwide
A true BiVAD is two devices for long-term biventricular support and is occasional — distinct from the common situation of a temporary RVAD at the time of durable LVAD implantation
Once LVAD support starts, left atrial pressure falls below right atrial pressure, so a silent PFO or ASD can shunt right to left — re-examine the atrial septum after support begins
Aortic regurgitation causes LVAD recirculation: outflow cannula, back across the valve, straight into the inflow cannula
Correction depends on intention — repair for bridge to recovery, suture closure for transplant or destination therapy, and a bioprosthesis if replacement is needed
Colour Doppler underestimates AR in end-stage failure because diastolic pressures equalise; assess on cardiopulmonary bypass and again at different pump speeds
Use vena contracta width and jet-to-outflow-tract ratio; pressure half-time and holodiastolic flow reversal are unreliable here
High LV vent flows on bypass may point to clinically significant AR
Aortic stenosis usually doesn't matter — except for partial-support devices requiring ongoing ejection
The interventricular septum should be flat and neutral: rightward means underpumping, leftward means suckdown
Ramp from minimum speed, aiming for a flat septum, a filled but not overdistended ventricle, and the aortic valve opening at least once every 3–5 beats
RV function doesn't always improve after LVAD: bypass, elevated PVR, a suddenly normalised preload, and altered LV geometry all work against it
Temporary right ventricular support is either a surgically grafted CentriMag or a percutaneous Protek Duo — right internal jugular, inflow in the right atrium, outflow in the proximal main pulmonary artery, up to about 4.5 L/min
Image the inflow cannula in two orthogonal long-axis views with biplane imaging, and reinterrogate both cannulas at chest closure
Low cardiac output with low device flows and a high central venous pressure is tamponade until echo says otherwise
The Impella is a temporary percutaneous device supporting either ventricle, with an inlet in the ventricle and outlet beyond the valve; optimal inlet position about 3.5 cm from the aortic valve, and the pigtail tip is not included in that measurement
Any deterioration in a patient on mechanical support gets a TOE, looking specifically for cannula malposition or occlusion
References / further reading
Charlesworth M, Allen SJ. Echocardiography for mechanical support (Chapter 20). In: Lambert AS, Allen SJ, Sidhu S, eds. Practical Perioperative Transoesophageal Echocardiography, 4th edn. Oxford: Oxford University Press, 2025. ISBN 9780198873686
Stainback RF, Estep JD, Agler DA, et al. Echocardiography in the management of patients with left ventricular assist devices: recommendations from the American Society of Echocardiography. J Am Soc Echocardiogr 2015; 28: 853–909
Mehra MR, Uriel N, Naka Y, et al. A fully magnetically levitated left ventricular assist device — final report. N Engl J Med 2019; 380: 1618–27
Crowley J, Cronin B, Essandoh M, D'Alessandro D, Shelton K, Dalia AA. Transesophageal echocardiography for Impella placement and management. J Cardiothorac Vasc Anesth 2019; 33: 2663–8
Potapov EV, Stepanenko A, Dandel M, et al. Tricuspid incompetence and geometry of the right ventricle as predictors of right ventricular function after implantation of a left ventricular assist device. J Heart Lung Transplant2008; 27: 1275–81
Chumnanvej S, Wood MJ, MacGillivray TE, Melo MF. Perioperative echocardiographic examination for ventricular assist device implantation. Anesth Analg 2007; 105: 583–601
Follow the podcast
Bluesky: @cardiacoutput.bsky.social
X: @CardiacOutputMC
If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a topic you'd like covered.
This podcast is for medical education for healthcare professionals. It is not clinical advice. Practice described reflects local Wythenshawe practice at the time of recording — always follow your own centre's guidelines and current local policy.

We've spent a lot of this series on machines — balloon pumps, ECMO circuits, ventricular assist devices. This episode is about the tool that makes all three of them safe.
The argument is that in mechanical circulatory support, transoesophageal echo is not an investigation you order. It is part of the device. It confirms the indication, guides the cannulas in, optimises the support once it is running, and finds the complication before the patient tells you about it. And one rule runs through the whole topic: any deterioration in a patient on mechanical support gets a TOE, with specific attention to cannula malposition or occlusion.
This is part one of two, covering balloon pumps and ECMO. Part two takes on assist devices and the Impella.
We start with the balloon pump, where TOE has real advantages over the chest film and fluoroscopy — real-time structures, portable, no ionising radiation. Scan before you insert, for two reasons: to confirm the indication, and to exclude the absolute contraindications, which are severe aortic regurgitation, severe atheroma in the descending thoracic aorta, aortic dissection, and severe distal occlusive disease. We explain why regurgitation is so absolute — the balloon inflates in diastole to augment diastolic pressure, so a leaking valve means you drive that augmented pressure straight back into the ventricle, worsening the regurgitation without improving coronary perfusion pressure. The device fails at its own job. Then the relative contraindications, including dynamic left ventricular outflow tract obstruction, where reducing afterload makes systolic anterior motion worse.
Positioning gets done properly: the tip about two centimetres distal to the origin of the left subclavian, with the proximal balloon above the diaphragm — too high and you occlude the subclavian, too low and you lose augmentation and may intermittently occlude the mesenteric or renal arteries. How to find the subclavian origin in the upper oesophageal arch short axis, how to tell the echo-dense tip and lucent balloon body from the wire, and what to use as a landmark when you genuinely cannot see the origin. Plus a technique that needs no technology at all: identify the balloon tip in the descending aorta short axis, put your fingertips on the probe at the level of the teeth, withdraw until the subclavian origin appears — and the distance your fingers have travelled is the distance from tip to subclavian. You have measured it with your hand.
The ECMO section opens with the decision that matters most, which is mode — and the most useful piece of thinking in the episode. When a severe respiratory failure patient becomes haemodynamically unstable, it is usually one of four things: right ventricular dysfunction from aggressive ventilation and acutely raised pulmonary vascular resistance, ventricular dysfunction from sepsis or underlying cardiac disease, vasodilation from sepsis, or hypovolaemia. Most of those do not need arterial support. Ventilator-induced right ventricular dysfunction usually resolves once ECMO lets you drop to rest settings, and vasodilation is usually managed with VV plus vasopressors. So instability alone does not mean VA — it means work out the mechanism first, and echo is how you do that.
Then what to look for before you commit: significant aortic or mitral regurgitation, which can cause acute left ventricular distension in a VA patient who is barely ejecting; severe aortic disease as a contraindication to VA; severe atheroma as a relative one, because of the sandblasting effect of the return jet; and the precise position on a PFO, which may complicate VV but is not a contraindication.
Cannula positioning is laid out by configuration, because the rules are clean once you separate them — confirming the guidewire before you dilate, where the return tip belongs in VV, where the drainage tip goes for jugular–femoral versus femoral–femoral versus a double-lumen cannula, and why VA is simpler. Then recirculation: how to diagnose it with 2D and colour Doppler, the two findings that confirm it, and how to fix it under echo guidance.
We finish with the complications that are hard to see coming. The specific screen appearance of left ventricular distension on VA and the two families of solution. Circuit and intracardiac thrombus. And tamponade after cardiac surgery, which is genuinely difficult because the circuit largely bypasses flow through the heart, so it is initially well tolerated and you never get the classic picture — until the right atrium collapses and your circuit flow falls. Raised central venous pressures with low circuit flows on central VA means tamponade until proven otherwise. Plus the TOE predictors associated with successful weaning from VA.
Chapters
(00:00) Cold open — the tool that makes the machines safe
(00:50) Any deterioration on support gets a TOE
(01:20) The balloon pump: why scan before you insert
(02:00) Absolute contraindications, and why AR is absolute
(03:10) Relative contraindications, including dynamic LVOT obstruction
(04:00) The pre-insertion study
(04:50) Positioning the tip — and what happens if it's wrong
(05:40) Finding the subclavian origin
(06:30) Measuring the distance with your fingers
(07:20) Post-insertion complications, and is it working?
(08:00) ECMO: choosing the mode
(08:40) The four causes of instability in respiratory failure
(09:50) Why instability alone doesn't mean VA
(10:30) Lesions that will complicate things — and the PFO
(11:30) Using the probe during cannulation
(12:10) Cannula positions by configuration
(13:10) Recirculation: diagnosis and fix
(14:10) Left ventricular distension on VA
(15:10) Thrombus
(15:40) Tamponade — why it hides on central VA
(16:30) Weaning, and the TOE predictors
(17:20) Wrap-up
Key takeaways
In mechanical circulatory support, TOE is part of the device rather than an investigation you order
Any deterioration in a patient on mechanical support gets a TOE, looking specifically for cannula malposition or occlusion
Absolute contraindications to an IABP on echo: severe aortic regurgitation, severe descending aortic atheroma, dissection, and severe distal occlusive disease
Aortic regurgitation is absolute because augmenting diastolic pressure across a leaking valve worsens the regurgitation without improving coronary perfusion pressure
Reducing afterload worsens dynamic LVOT obstruction and systolic anterior motion, so that is a relative contraindication
Position the balloon tip about 2 cm distal to the left subclavian origin, with the proximal balloon above the diaphragm
You can measure tip-to-subclavian distance by walking the probe back with your fingers at the level of the teeth
ECMO mode follows the mechanism: instability alone does not mean VA
Ventilator-induced RV dysfunction and septic vasodilation are usually managed on VV, with rest settings and vasopressors
Assess both the aortic and mitral valves before VA, because regurgitation predisposes to acute LV distension in a barely ejecting ventricle
Severe aortic dissection or aneurysm contraindicates VA; severe atheroma is relative, because of the sandblasting effect of the return jet
A PFO may complicate VV ECMO but is not a contraindication to ECMO
Confirm the guidewire with the probe before dilating and passing a large-bore cannula
In VV, the return tip sits in the right atrium with the jet aimed at the tricuspid valve, and the tips must be kept apart to avoid recirculation
Diagnose recirculation by finding both tips on 2D, then looking for a high-velocity jet down the IVC or a jet that includes the drainage tip
LV distension on VA looks like a distended ventricle, minimal contractility, a closed aortic valve, a distended left atrium and significant mitral regurgitation, with a non-pulsatile arterial trace
Severe distension causes pulmonary oedema, delays recovery and predisposes to intracavity thrombus
Tamponade is poorly detected on central VA because the circuit bypasses the heart — raised venous pressures with low circuit flows means tamponade until proven otherwise
Have a low threshold for rescanning after starting ECMO: complications are common and hard to detect early
References / further reading
Charlesworth M, Allen SJ. Echocardiography for mechanical support (Chapter 20). In: Lambert AS, Allen SJ, Sidhu S, eds. Practical Perioperative Transoesophageal Echocardiography, 4th edn. Oxford: Oxford University Press, 2025. ISBN 9780198873686
Douflé G, Roscoe A, Billia F, Fan E. Echocardiography for adult patients supported with extracorporeal membrane oxygenation. Crit Care 2015
Platts DG et al. The role of echocardiography in the management of patients supported by extracorporeal membrane oxygenation. J Am Soc Echocardiogr 2012
Hahn RT et al. Guidelines for performing a comprehensive transesophageal echocardiographic examination. J Am Soc Echocardiogr 2013
Truby LK et al. Left ventricular distension and venoarterial extracorporeal membrane oxygenation. ASAIO J 2017
Follow the podcast
Bluesky: @cardiacoutput.bsky.social
X: @CardiacOutputMC
If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a topic you'd like covered.
This podcast is for medical education for healthcare professionals. It is not clinical advice. Practice described reflects local Wythenshawe practice at the time of recording — always follow your own centre's guidelines and current local policy.

"Draw me a cardiopulmonary bypass circuit." Venous cannula, reservoir, pump, oxygenator, back to the aorta, cross-clamp, cardioplegia, vents, suckers. Good — that's a pass. The mark comes when you can say what each component does to the patient rather than what it does to the blood, and that is the difference between a trainee who has memorised a diagram and a senior who understands a machine.
This is cardiopulmonary bypass top to bottom, pitched at FRCA level and a little beyond, and built around the questions that have actually appeared in past papers. Almost everything else we discuss on a cardiac unit assumes this episode.
Please note: the thresholds, ACT practice and doses are Wythenshawe-specific local practice. Take the principles, and check your own guidelines.
We start with the circuit, answering the examiner's real question component by component. The reservoir is a volume buffer — somewhere to hold blood so flow is maintained when the surgeon lifts the heart and venous return suddenly falls — and it is where air gets trapped rather than delivered. The heat exchanger is the patient's entire thermoregulation for the duration, not an accessory. The arterial filter is about the brain, because embolic load is one of the two mechanisms behind neurocognitive injury after bypass, and hypoperfusion is the other.
Then cannulation, including the one absolute rule in the whole chapter: never site a right radial arterial line if the axillary artery is going to be cannulated, because clamping it takes out your trace for a prolonged period and you will be flying blind at the worst possible moment. Plus the circuit variants worth knowing — hybrid circuits switchable between full bypass and VA-ECMO for lung transplantation, haemoadsorption for the septic patient, and ultrafiltration, which must be handed over explicitly to intensive care rather than buried in the chart. And a question most people are never taught properly: how is the patient anaesthetised on bypass? An isoflurane vaporiser on the circuit, capnography switched to the oxygenator — and depth monitoring throughout, because the transitions on and off pump are exactly where awareness happens.
Haemodilution gets a proper answer rather than a hand-wave. The haematocrit falls the moment bypass starts because you have connected the patient to a crystalloid prime, and it is tolerated because haematocrit is two competing things at once: oxygen-carrying capacity, which you want high, and viscosity, which you want low — especially when cold. Hence a target somewhere around 21–24%, and retrograde autologous priming to reduce the crystalloid load in selected patients.
Anticoagulation is given both ways. Heparin 300–400 units/kg and an ACT over 480 seconds is the version the exam wants; locally the threshold is over 400, because our machines double-count. Then the scenario: the ACT is 210 after a full dose. Don't give more heparin. Confirm it reached the circulation — go back to the lumen you used and aspirate it, because a surprising amount of this is a delivery problem rather than a pharmacology one. Then think heparin resistance, which almost always means acquired antithrombin deficiency, and give antithrombin III 500 units.
Cardioplegia covers the mechanism precisely — a high extracellular potassium depolarises the membrane, inactivating the fast inward sodium channels so there is no action potential upstroke — the five things cardioplegia actually achieves, the six advantages of blood over crystalloid (the exam wants three), warm versus cold, and the three indications for retrograde delivery with the 40 mmHg coronary sinus limit. Then the material that takes you past Fellowship: del Nido, and why Custodiol arrests the heart by hyperpolarisation rather than depolarisation. Opposite direction, same destination.
Hypothermia is delivered in the shape the question asks for — four advantages and three disadvantages — including the paradox worth saying out loud, that you are cooling to reduce oxygen demand while simultaneously making oxygen harder to offload. Plus flow and pressure targets, and alpha-stat versus pH-stat.
We finish with an eight-step checklist for separation from bypass, the vasoplegia that commonly follows, and protamine: the electrostatic reversal, why you must account for the heparin in the prime as well as the dose you gave, the four groups of adverse reaction with the Horrow type III as the one to fear, and why too much protamine is itself an anticoagulant.
Chapters
(00:00) Cold open — "draw me a bypass circuit"
(01:00) What each component does to the patient
(03:10) Cannulation, and the right radial rule
(04:20) Circuit variants: hybrid ECMO, haemoadsorption, ultrafiltration
(05:40) How is the patient anaesthetised on bypass?
(06:50) Priming, and why the haematocrit falls
(08:20) Heparin, the ACT, and what the ACT is actually for
(09:50) The ACT is 210 — what do you do, in order?
(11:20) Cardioplegia: how potassium arrests the heart
(12:30) Blood versus crystalloid, warm versus cold
(13:20) Antegrade and retrograde delivery
(14:10) del Nido, Custodiol and hyperpolarising arrest
(15:20) Laplace, and why we vent
(15:50) Hypothermia: four advantages, three disadvantages
(17:00) Flow, pressure, alpha-stat and pH-stat
(17:40) Coming off bypass — the eight-step checklist
(19:10) Vasoplegia
(19:40) Protamine: pharmacology, reactions and technique
(20:40) What arrives on the unit afterwards
(21:40) Wrap-up
Key takeaways
Name the parts, then say what each does to the patient: the reservoir is a volume buffer and an air trap, the heat exchanger is the patient's thermoregulation, the arterial filter protects the brain
Neurocognitive injury after bypass comes from embolic load and hypoperfusion
Never site a right radial arterial line if the axillary artery is to be cannulated
Hybrid circuits can switch between full bypass and VA-ECMO — used for lung transplantation, because it avoids full-dose anticoagulation and lets the heart keep ejecting
If ultrafiltration has been used, hand that over explicitly — over-filtration makes postoperative fluid and electrolyte management difficult
Monitor depth of anaesthesia throughout: going on and coming off pump disrupts both volatile and intravenous delivery, and that is where awareness happens
The haematocrit falls because of the crystalloid prime, and it is tolerated because haematocrit is oxygen carriage and viscosity at the same time — target around 21–24%
Heparin 300–400 units/kg with an ACT over 480 seconds for the exam; over 400 locally because the machines double-count
The ACT's genuinely useful role is confirming it is safe to turn the suckers on and go onto bypass
If the ACT won't rise, don't give more heparin — confirm it reached the patient, then give antithrombin III 500 units for acquired antithrombin deficiency
Potassium cardioplegia depolarises the myocyte membrane and inactivates the fast inward sodium channels, abolishing the action potential upstroke
Blood cardioplegia adds oxygen carriage, hydrogen ion buffering, free-radical scavenging, improved microvascular flow, reduced myocardial oedema and nutrient delivery
Retrograde delivery is needed for significant aortic regurgitation, root surgery and severely diseased coronaries — keep coronary sinus pressure below 40 mmHg
Custodiol arrests the heart by hyperpolarisation through sodium depletion, the opposite mechanism to potassium depolarisation
Laplace's law is the basis of every unloading strategy: venting, the IABP, and LV decompression on VA-ECMO
Hypothermia reduces cerebral and myocardial oxygen consumption but shifts the oxyhaemoglobin curve left, so offloading is impaired at the same time as demand falls
Separation from bypass in eight steps: warm, rhythm and AV synchrony, electrolytes and acid–base, lungs, de-air under TOE, rate to suit the lesion, titrate support, separate incrementally
Protamine 300 mg reverses roughly 30,000 units of heparin — and you must account for the heparin in the pump prime too
The Horrow type III reaction is pulmonary hypertension with right heart strain, and excess protamine is itself an anticoagulant
References / further reading
Charlesworth M. CTCCU Handbook, 2nd edition, sections 2b and 2d
Wahba A et al. EACTS/EACTAIC/EBCP Guidelines on cardiopulmonary bypass in adult cardiac surgery. Eur J Cardiothorac Surg
Murphy GS, Hessel EA, Groom RC. Optimal perfusion during cardiopulmonary bypass: an evidence-based approach. Anesth Analg 2009
Ltaief Z et al. Pathophysiology and clinical implications of vasoplegic syndrome after cardiopulmonary bypass. J Clin Med 2022
Finley A, Greenberg C. Heparin sensitivity and resistance: management during cardiopulmonary bypass. Anesth Analg 2013
Boer C et al. 2017 EACTS/EACTA Guidelines on patient blood management for adult cardiac surgery. J Cardiothorac Vasc Anesth 2018
Follow the podcast
Bluesky: @cardiacoutput.bsky.social
X: @CardiacOutputMC
If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a topic you'd like covered.
This podcast is for medical education for healthcare professionals. It is not clinical advice. All thresholds, doses and practice described reflect local Wythenshawe practice at the time of recording — always follow your own centre's guidelines and current local policy.

Two in the morning, day one after a long CABG. The patient has a balloon pump in, the augmented pressure has dropped, and the urine output has fallen away over the last three hours. The easiest thing in the world at that hour is to re-zero the transducer, decide the trace looks a bit better, and go back to what you were doing. By the end of this episode you'll know exactly why that's the wrong answer.
This is the intra-aortic balloon pump, top to bottom, pitched at Final FRCA and a bit past it — it has appeared in past papers three separate ways, as principles, as indications and contraindications, and as complications. But the version worth having is the one where you can look at a trace on the unit and know what's wrong with it.
Please note: the weaning approach and anticoagulation discussed are Wythenshawe-specific local practice. Take the principles, and check your own guidelines.
We start with helium, and the two reasons to give rather than one — it's low density, so it shuttles down a long narrow catheter fast enough to work inside a fraction of a cardiac cycle, and it's highly soluble in blood, so rupture is far more forgiving than air would be. Then counterpulsation, and the sentence the whole device hangs on: the balloon pump increases myocardial oxygen supply and reduces demand at the same time, which almost nothing else does. Give adrenaline to an ischaemic ventricle and coronary perfusion may improve, but rate, contractility and wall stress have all gone up, so you've bought a little flow at a large metabolic price. Diastolic augmentation raises coronary perfusion pressure, presystolic deflation drops aortic end-diastolic pressure and therefore afterload and wall stress — supply up, demand down.
Then the trace, as a learnable set piece. The rule first: put the pump on 1:2, so every other beat is unassisted and you have a control sitting next to your test. Then three comparisons in order — augmented diastolic higher than unassisted systolic, assisted end-diastolic lower than unassisted end-diastolic, and assisted systolic lower than unassisted systolic. That last one catches people every time: a lower assisted systolic pressure is not the pump failing, it's direct evidence you have unloaded the ventricle.
The four timing errors get sorted by harm rather than by name, which is the distinction that shows understanding rather than recall. Early inflation and late deflation both load the ventricle during ejection and are the dangerous pair — late deflation worst of all, because the ventricle is ejecting against an inflated balloon. Late inflation and early deflation merely waste benefit, although early deflation can drive retrograde coronary and carotid flow and cause angina. Then triggers, the asynchronous mode and why you'd ever want it, and why arrhythmia is the balloon pump's great enemy.
Indications and contraindications follow, including the most satisfying piece of physiology in the episode: why a balloon pump helps in acute severe mitral regurgitation. The ventricle has two exits, and how much blood goes each way depends on the relative resistance of the two routes — so dropping aortic end-diastolic pressure makes the forward path easier, the regurgitant fraction falls, and forward output rises. You're not fixing the valve, you're changing the arithmetic while somebody organises theatre. And on the other side, why aortic regurgitation is an absolute hard stop: everything the balloon does in diastole raises aortic root pressure, so in an incompetent valve you are augmenting the leak straight back into a failing ventricle.
Then placement and the landmarks that matter, what TOE adds, the daily chest film, complications split into insertion, use and removal, and back to the 2am patient — falling urine output means think down, a lost left radial pulse means think up, and helium or blood in the tubing means rupture and it comes out now.
We finish on the trap. IABP-SHOCK II was negative, and Altshock-2 in 2025 was stopped for futility in heart failure–related shock. So why is there one running in bed four? The answer isn't to ignore the evidence — it's to notice what those trials actually studied, which was routine, unselected use in two specific shock populations. That is a different question from the patient who cannot come off bypass, the one with acute severe mitral regurgitation waiting for theatre, or the one who needs to survive four hours until the cath lab.
Chapters
(00:00) Cold open — the augmented pressure has dropped
(01:00) Why this episode, and what level we're pitching at
(01:40) What it actually is, and why helium — two reasons
(02:45) Counterpulsation: supply up and demand down at once
(03:45) Inflation, and why the left ventricle is perfused in diastole
(04:15) Deflation, afterload and wall stress
(05:15) The trace — and why you put it on 1:2
(06:05) Reading a pair of beats
(07:35) Lower is better: the comparison everyone misreads
(08:05) The four timing errors, sorted by harm
(10:35) Triggers, and the asynchronous mode
(11:15) Arrhythmia and tachycardia
(12:05) Indications
(12:35) Acute mitral regurgitation, and post-infarct VSD
(13:25) Contraindications — and why aortic regurgitation is absolute
(15:15) Insertion, TOE, and the daily film
(16:55) Complications: insertion, use, removal
(18:35) Back to the 2am patient — migration up and down
(19:55) Balloon rupture
(20:15) Anticoagulation and weaning
(21:15) IABP-SHOCK II, Altshock-2, and the trap
(24:15) Wrap-up
Key takeaways
Helium for two reasons: low density so it shuttles fast, and high solubility so rupture is survivable
The balloon occupies 80–90% of the aortic cross-section, inflating in diastole and deflating before systole
It raises myocardial oxygen supply and lowers demand simultaneously — almost nothing else does
Diastolic augmentation raises coronary perfusion pressure; presystolic deflation drops aortic end-diastolic pressure, and therefore afterload, wall stress and demand
Read the trace on 1:2 so you have an unassisted control beat next to your assisted one
Augmented diastolic should exceed unassisted systolic — the tallest thing on the screen should be the balloon, not the heart
A lower assisted systolic pressure is good: it's evidence of unloading, not of pump failure
Early inflation and late deflation load the ventricle during ejection and are the dangerous errors; late inflation and early deflation waste benefit, though early deflation can cause retrograde coronary flow and angina
Arrhythmia is the balloon pump's great enemy; above a rate of about 120 you may get better support at 1:2 than 1:1
In acute severe mitral regurgitation, reducing aortic end-diastolic pressure shifts flow from the regurgitant route to the forward one
Aortic regurgitation is an absolute contraindication because you would be augmenting the leak into an already failing ventricle — along with dissection, aneurysm, a prosthetic aorta, and futility
Tip at the level of the carina and at least 2 cm below the aortic knob, just distal to the left subclavian, balloon above the diaphragm — checked every single day
Falling urine output or a distended abdomen means think downward migration; a lost or damped left radial trace means think upward
Helium or blood in the tubing means rupture: stop and remove urgently, before clot inside the balloon entraps it
Wean by ratio or augmentation percentage, judged on haemodynamics, urine output and lactate rather than on the machine
IABP-SHOCK II and Altshock-2 were negative for routine use in specific shock populations — which is not the question you are usually asking on a cardiac unit
References / further reading
Thiele H et al. Intraaortic balloon support for myocardial infarction with cardiogenic shock (IABP-SHOCK II). N Engl J Med 2012; 367: 1287–96
Thiele H et al. Intra-aortic balloon pump in cardiogenic shock complicating acute myocardial infarction: long-term 6-year outcome of the randomised IABP-SHOCK II trial. Circulation 2019
Early intra-aortic balloon support for heart failure-related cardiogenic shock (Altshock-2): a randomised clinical trial. J Am Coll Cardiol 2025
Møller JE et al. Microaxial flow pump or standard care in infarct-related cardiogenic shock (DanGer Shock). N Engl J Med 2024
Thiele H et al. Extracorporeal life support in infarct-related cardiogenic shock (ECLS-SHOCK). N Engl J Med2023
Byrne RA et al. 2023 ESC Guidelines for the management of acute coronary syndromes. Eur Heart J 2023
Krishna M, Zacharowski K. Principles of intra-aortic balloon pump counterpulsation. Contin Educ Anaesth Crit Care Pain 2009
Follow the podcast
Bluesky: @cardiacoutput.bsky.social
X: @CardiacOutputMC
If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a topic you'd like covered.
This podcast is for medical education for healthcare professionals. It is not clinical advice. Weaning and anticoagulation practice described reflects local Wythenshawe practice at the time of recording — always follow your own centre's guidelines and current local policy.

Something different this episode. No 3am emergency — instead, a pre-assessment clinic on a Tuesday afternoon, a woman booked for an aortic valve replacement in six weeks, and a haemoglobin of 118. Do you give her intravenous iron?
ITACS, published in The BMJ in August 2026, is the best answer we have ever had to that question. It is also a trial with a great deal to teach about how to read a paper properly — so this is a journal club, and we go through the methods slowly, because that is where the meaning lives.
The question first. A third of patients coming for cardiac surgery are anaemic, twenty to fifty per cent are transfused, and both are powerfully associated with complications, longer stays and death — in a specialty that consumes around ten per cent of the entire NHS blood supply. Correcting the anaemia in clinic attacks both problems at once. But anaemia may be a marker as well as a mechanism: if patients do badly because of the kidney disease or inflammation that made them anaemic, then fixing the number fixes the screen and changes nothing about the patient. A strong prognostic marker is not automatically a treatment target — and that idea runs through the whole episode.
Then the methods. Thirty-three hospitals across ten countries, 955 anaemic adults for elective cardiac surgery, a single 1,000 mg dose of intravenous iron or placebo one to twenty-six weeks beforehand — and, deliberately, no requirement to prove iron deficiency. We look at how you blind a brown drug (a black syringe and an opaque line), and at the detail that separates a good trial from a trial that merely says "double blind" in its abstract: they checked whether the masking had worked by asking patients to guess their allocation.
The primary outcome is days alive and at home at 90 days — what it captures, and the two things it hides. And then the finding that isn't in the abstract at all: the primary outcome was originally days at home at thirty days, and was amended to ninety in December 2020. We give the defence and the concern, and then we look at what the original outcome showed. It was null. Had the investigators kept it, this would be a negative trial, and that belongs in any honest summary of the paper.
The results deserve care. The iron worked biochemically — ferritin rose from around 110 to over 500 — but haemoglobin rose by under 4 g/L, and three quarters of treated patients were still anaemic on the day of surgery, which explains the size of everything that follows. The primary result is one day, with a confidence interval touching zero, in a trial powered for a day and a half. Transfusion is the solid finding: 68% down to 61%, about fifteen patients treated to prevent one transfusion, and roughly 44 units of blood saved per hundred patients. But length of stay was identical, complications were identical — so where did the extra day come from? The paper answers it, and the answer is a smaller claim than the headline sounds.
The best thing in the trial isn't in the abstract either. Rather than only reporting the median, the investigators reported the treatment effect across the whole distribution — and it turns out that patients who recovered well gained a fraction of a day, while at the 25th centile the difference was 6.1 days. All the benefit sits with the patients who did badly. It is post hoc, it was requested at peer review, and one of its confidence intervals is enormous — but it is biologically coherent, and it may be the most important idea to come out of this trial.
We also correct an argument we would have made before reading the full paper. The obvious criticism — that enrolling patients without proven iron deficiency dilutes the effect — was pre-specified, tested, and not supported. Calum then supplies the more sophisticated objection, which is that an interaction test in a trial this size cannot exclude a subgroup difference. Plus the safety signal that matters to us specifically: a sixteen-fold increase in hypophosphataemia.
We finish with the four-point critique, the strengths the trial genuinely deserves, and what to do in clinic on Monday.
Chapters
(00:00) Cold open — a haemoglobin of 118 and six weeks to go
(01:00) Why the question matters, and the ten per cent of the blood supply
(02:10) Marker or mechanism? The idea that runs through everything
(03:00) What we knew before ITACS
(03:40) The methods: 33 hospitals, 955 patients, no iron deficiency required
(04:40) How do you blind a brown drug — and how do you prove it worked?
(05:40) Days alive and at home: what it captures and what it hides
(07:10) The primary outcome was changed mid-trial
(08:30) What the original outcome showed
(09:20) Sample size, two interim analyses, and the 95.4% interval
(10:30) Did the iron actually do anything? Under 4 g/L
(11:40) The primary result — one day, and an interval touching zero
(12:50) Transfusion: the solid finding
(13:50) Same length of stay, same complications — so where did the day come from?
(15:00) The result the headline misses: who actually benefited
(16:30) Iron deficiency — an argument corrected, and a better objection
(18:00) Safety, and the phosphate
(18:40) The critique in four parts
(19:30) The strengths, and Monday morning
(20:30) Wrap-up
Key takeaways
Anaemia is a marker as well as a mechanism — a strong prognostic marker is not automatically a treatment target
ITACS: 33 hospitals, 10 countries, 955 anaemic adults, a single 1,000 mg dose of IV iron or placebo, 1–26 weeks before elective cardiac surgery
The primary outcome was changed from days at home at 30 days to 90 days in December 2020 — and the original outcome was null
Always ask what a trial's original primary outcome was, and what it showed
The iron filled the stores but raised haemoglobin by under 4 g/L, and three quarters of treated patients were still anaemic on the day of surgery
Primary result: one extra day at home in 90, with a confidence interval touching zero, in a trial powered for 1.5 days
Transfusion is the robust finding: 68% to 61%, NNT about 15, and roughly 44 units of blood saved per 100 patients treated
Length of stay and complications were identical — the extra day accrued after discharge as shorter readmissions, not fewer of them
A post hoc quantile analysis suggests all the benefit sits with the poor recoverers: 6.1 days at the 25th centile against a fraction of a day at the 75th
Benefit was the same with or without demonstrable iron deficiency — but that interaction test is underpowered, so read it as no reason to restrict treatment rather than proof that deficiency is irrelevant
Intravenous iron caused a sixteen-fold increase in hypophosphataemia, which is not a footnote in a patient about to be weaned
The 95.4% confidence interval is the fingerprint of pre-planned interim analyses and alpha spending
Find the anaemia early: the biggest barrier in this trial was time, not biology
Read the methods before the abstract, and convert every relative number into an absolute one
References / further reading
Myles PS, Klein AA, Smith JA, et al. Intravenous iron to treat anaemia before cardiac surgery (ITACS): international, double blind, placebo controlled randomised trial. BMJ 2026; 394: e100407
Myles PS, Richards T, Klein A, et al. Rationale and design of the intravenous iron for treatment of anemia before cardiac surgery trial. Am Heart J 2021; 239: 64–72
Boer C et al. 2017 EACTS/EACTA Guidelines on patient blood management for adult cardiac surgery. J Cardiothorac Vasc Anesth 2018
Agarwal S, Choi SW, Fletcher SN, Klein AA, Gill R. The incidence and effect of resternotomy following cardiac surgery on morbidity and mortality: a 1-year national audit on behalf of the Association of Cardiothoracic Anaesthesia and Critical Care. Anaesthesia 2021; 76: 19–26
Myles PS et al. Validation of days at home as an outcome measure after surgery. BMJ Open 2017
Follow the podcast
Bluesky: @cardiacoutput.bsky.social
X: @CardiacOutputMC
If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a paper you'd like taken apart.
This podcast is for medical education for healthcare professionals. It is not clinical advice. Discussion of published research reflects our own reading and interpretation — always read the primary source and follow your own centre's guidelines.

You will have noticed that consultants tend to give the protamine themselves, and if you're an ST3 who hasn't done much cardiac, nobody asks you to do it. Most people assume that's habit. It isn't.
This is the theatre half of the bleeding week — the drugs, the traps in the tests, and a salvage case at the far end of what's possible. Last episode was the unit.
Please note: the doses and drug choices are Wythenshawe-specific local practice. Take the principles, and check your own guidelines.
The scale first: cardiac surgery consumes around ten per cent of the entire NHS blood supply, and ninety per cent of those products go into ten per cent of patients. Which makes anticipating who that will be most of the job. Bypass is hostile to clotting in five ways at once — a large heparin dose, haemodilution, cooling, platelet activation and consumption, and contact with the circuit — so know the cases where you will predictably need products (deep hypothermic circulatory arrest, redos, long pump runs, transplants and VAD explants, endocarditis, and anyone arriving anticoagulated) and act on it while they're still on bypass. Order them, get them thawed, get them physically in the room. The worst position is realising you need four units of something that's still in a freezer twenty minutes away.
Then the antifibrinolytics. Tranexamic acid at three to four grams is the modern default; aprotinin — Trasylol — was withdrawn in 2008, reintroduced in 2012 with very narrow licensing, and is now used off-license for high bleeding risk, with the honest admission that there is very little evidence directly comparing the two. And a trap the crib sheet asks about directly: aprotinin artificially prolongs the R time on your TEG, so the trace suggests a coagulopathy that isn't there and you can go chasing it with products the patient doesn't need.
We do the acquired von Willebrand story properly, because it's lovely physiology: turbulent flow across a tight aortic valve shears platelets and cleaves the large von Willebrand multimers, so the valve lesion itself causes a bleeding disorder — and explains the patient who oozes at the end of an aortic valve replacement for no other obvious reason. DDAVP, twenty micrograms in a hundred millilitres, given slowly, because giving it fast to a marginal patient at the end of a case will drop their pressure.
The centrepiece is protamine. What it actually does (electrostatic neutralisation, not a receptor effect), the three types of reaction, and why type three — profound pulmonary vasoconstriction, pulmonary hypertension and right ventricular failure — is the one that empties the ventricle and stops the heart. Then the technique that follows from the principle that when you decide to stop, no more protamine must reach the patient: not through a running flush with a column of drug behind it, but through the central line with the tap turned between patient and syringe. And the counterintuitive part — protamine in excess is itself an anticoagulant, so the generous dose given "to be sure" makes the patient oozier, not drier.
Heparin rebound gets its own section, because it's the four-in-the-morning diagnosis that saves a patient a lot of unnecessary products: protein-bound and extravascular heparin redistributing into a circulation whose protamine has already been cleared. Unexplained oozing in a patient who left theatre dry needs a small further dose of protamine, not factors. We also cover heparin resistance as an antithrombin problem, and recombinant factor seven — including a real case where it arrived from another hospital by ambulance and stopped the bleeding immediately, with proper caveats about cost, availability, thrombosis and haematology authorisation.
We finish at the salvage end: a patient with acute HIT and thrombotic complications who needs an emergency BiVAD, with no time for plasma exchange or immunoglobulin. Heparin is out, argatroban is a unit drug rather than a theatre one, and almost nobody is comfortable with bivalirudin. The answer starts with getting the right surgeon — because some will go onto cardiopulmonary bypass first, and a bypass circuit demands full anticoagulation. Instead: VA-ECMO first as a heparin-free run, then the BiVAD from there, decided at a full MDT that is honest about what this is. And for the patient with HIT who needs surgery but isn't an emergency, the first and best intervention is simply time.
Chapters
(00:00) Cold open — the most dangerous drug we give
(00:50) The scale: ten per cent of the blood supply
(01:40) Why bypass is hostile to clotting
(02:30) Who predictably needs products — and acting while still on bypass
(03:40) Tranexamic acid, and aprotinin
(05:20) Aprotinin prolongs the R time — don't chase it
(06:10) Acquired von Willebrand disorder in aortic stenosis
(07:40) DDAVP, and why you give it slowly
(08:50) Protamine — what it does and why it's dangerous
(10:00) The three types of reaction
(11:00) Giving it so that stopping means stopping
(12:20) Too much protamine is an anticoagulant
(13:00) Heparin rebound — the 4am diagnosis
(14:40) Heparin resistance and antithrombin
(15:40) Recombinant factor seven, and the ambulance
(17:20) Acute HIT needing an emergency BiVAD
(19:40) After the salvage: no platelets, and buying time
(20:30) Elective HIT: delay is the treatment
(21:40) Bivalirudin, if you had to
(22:40) Wrap-up
Key takeaways
Cardiac surgery uses around 10% of the NHS blood supply, and 90% of those products go to 10% of patients — anticipate, and get products thawed and in the room while the patient is still on bypass
Bypass impairs clotting five ways: heparin, haemodilution, cooling, platelet activation and consumption, and circuit contact
Tranexamic acid 3–4 g is the default; aprotinin is off-license for redos and high bleeding risk, with little comparative evidence
Aprotinin prolongs the R time on the TEG — don't chase a coagulopathy that isn't there
Severe aortic stenosis causes acquired von Willebrand disorder through shear; DDAVP 20 micrograms in 100 mL, given slowly, and cryoprecipitate is a rational choice
Protamine neutralises heparin by charge, and can cause pulmonary hypertension, RV failure and cardiac arrest
Type 1 is rate-related hypotension, type 2 anaphylactoid, type 3 the catastrophic pulmonary vasoconstriction — type 3 is what kills
Give protamine so that stopping means stopping: no flush behind it, tap turned between patient and syringe
Excess protamine is itself an anticoagulant — titrate it, don't pour it in
Unexplained oozing hours later on the unit is heparin rebound: give more protamine, not more products
Heparin resistance is an antithrombin problem — antithrombin III around 500 units, not more heparin
Recombinant factor seven works but is expensive, hard to locate, prothrombotic and needs haematology sign-off
For emergency surgery in acute HIT: the right surgeon, VA-ECMO before the BiVAD, a heparin-free run, no platelets, and an MDT honest about it being salvage
Where there is time, time is the treatment — delay towards remote HIT, plasma exchange, IVIG, and a single intra-operative heparin exposure only
References / further reading
Boer C et al. 2017 EACTS/EACTA Guidelines on patient blood management for adult cardiac surgery. J Cardiothorac Vasc Anesth 2018
Task Force et al. 2024 EACTS/EACTAIC Guidelines on patient blood management in adult cardiac surgery
Myles PS et al. Tranexamic acid in patients undergoing coronary-artery surgery (ATACAS). N Engl J Med 2017
Fergusson DA et al. A comparison of aprotinin and lysine analogues in high-risk cardiac surgery (BART). N Engl J Med 2008
Nybo M, Madsen JS. Serious anaphylactic reactions due to protamine sulfate: a systematic literature review. Basic Clin Pharmacol Toxicol 2008
Vincentelli A et al. Acquired von Willebrand syndrome in aortic stenosis. N Engl J Med 2003
Cuker A et al. American Society of Hematology 2018 guidelines: heparin-induced thrombocytopenia. Blood Adv2018
Koster A et al. Anticoagulation during cardiopulmonary bypass in patients with heparin-induced thrombocytopenia. Ann Thorac Surg
Follow the podcast
Bluesky: @cardiacoutput.bsky.social
X: @CardiacOutputMC
If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a topic you'd like covered.
This podcast is for medical education for healthcare professionals. It is not clinical advice. All drugs and doses discussed reflect local Wythenshawe practice at the time of recording — always follow your own centre's guidelines.