Chapter 22 — Post-ECPR Management

What this chapter covers — and what it does not

Chapter 21 stopped at the moment antegrade flow was established and compressions were discontinued. This chapter starts there and ends roughly twenty-four hours later. It is deliberately narrow, because almost every organ system it touches has a chapter of its own later in the book. What it owns is the set of decisions that are specific to a patient who was in cardiac arrest until a minute ago — and, above all, the handful that cannot be deferred.

The boundary with Chapter 23 is worth stating precisely, because it is the one readers will cross by accident. This chapter owns the acute neuroprotective decisions — the things done to the brain in the first hours. Chapter 23 owns the attempt to find out what happened to it. Continuous electroencephalography appears in both chapters and means different things in each: here it is a monitor for seizures that should be treated, there it is an instrument of prognosis. Neither chapter makes the other's argument.

22.1 The patient Chapter 21 handed over, and why this is not Chapter 14

Chapter 14 described the first hours on venoarterial ECMO. Much of it applies here unchanged, and this chapter does not repeat it. What it does not cover is the thing that makes this patient different, which is not the circuit but the forty minutes that preceded it.

A patient cannulated for cardiogenic shock arrives on the circuit having been perfused — badly, but continuously — throughout. A patient cannulated during cardiac arrest arrives having had no circulation at all for some minutes and a fraction of a circulation for some tens of minutes (Chapter 19 §19.2). Every capillary bed in the body has been ischaemic; every one of them is about to be reperfused simultaneously, at a rate and with a composition of blood that a machine decides. The consequences are not subtle.

That last row governs everything else in this chapter, and it needs a qualifier that is easy to get wrong. Most of the brain injury in an ECPR patient was inflicted before the cannula went in, and no amount of skill afterwards will undo it. But most is not all. The first hours are the one window in which a clinician can make an already-injured brain worse, quickly, using entirely conventional-looking actions. That asymmetry — limited ability to help, considerable ability to harm — is the correct frame for everything that follows.

22.2 The first ten minutes — three variables, one set of knobs

This is the organising idea of the chapter, and it is worth stating before any individual target.

At the moment antegrade flow begins, three quantities change. Arterial carbon dioxide falls, because a membrane lung with a sweep gas flowing across it is a far more efficient carbon dioxide exchanger than a pair of compressed lungs. Arterial oxygen rises, often steeply, because the gas blender is usually still at 100%. Arterial pressure rises, because a pump has been added to a circulation that already has two or three vasopressors running into it at doses titrated to a patient who had no cardiac output.

Each of those three is a potent cerebral vasoactive variable. Taken together they are close to a complete list of the things that determine cerebral blood flow in a patient who cannot regulate it themselves.

Now add the second variable. Cerebral autoregulation holds mean arterial pressure and cerebral blood flow loosely independent across a plateau. After global ischaemia that plateau is narrowed, shifted, or absent — which is precisely what the pressure reactivity index measures, and precisely what the experiments quoted later in this chapter find in the post-arrest brain. A brain that has lost autoregulation transmits arterial pressure directly to the capillary bed. So hypocapnia constricts the cerebral vessels while hypertension pushes pressure through whatever is left of them, at the same moment, in the same patient.

That combination is not a theoretical worry. It is the clearest empirical signal in the whole of this literature.

22.3 Carbon dioxide — the steepest gradient in medicine

Of the three variables, carbon dioxide deserves to be first because it is the one most likely to be moved furthest, fastest, and by accident.

Why the ECPR patient is the extreme case

Every ECMO patient experiences a fall in arterial carbon dioxide when the membrane lung is connected. The ECPR patient experiences the largest one available, for three reasons that compound.

1. The starting point is high. Chest compressions deliver perhaps a quarter to a third of normal cardiac output and considerably less alveolar ventilation than that. Carbon dioxide accumulates for the whole low-flow interval.
2. The exchange capacity is enormous. A contemporary adult membrane lung will clear the whole metabolic carbon dioxide production of an adult at a sweep gas flow of a litre or two per minute. At the flows circuits are habitually started at, it will clear a great deal more.
3. Nothing in the room is measuring it. End-tidal capnography, which was the team's carbon dioxide monitor five minutes earlier, now reads whatever is passing through the native pulmonary circulation, which may be almost nothing. The instrument the team has been staring at for forty minutes becomes uninformative at exactly the moment the variable it measures becomes dangerous.

That third point is worth pausing on. In the ECPR room, the end-tidal trace is the physiological signal with the longest continuous history and the strongest emotional weight — Chapter 19 §19.2 made it one of the gates. The moment flow starts, it stops meaning what it meant. Teams that do not say this out loud will keep reading it.

The evidence that the drop matters

This is the part of the chapter with the most data and the least agreement. It is worth laying out in full, because the disagreement resolves in an instructive way.

Four registry analyses across three age groups say the same thing. Two smaller studies with better measurement say it less clearly or not at all. The usual reading of that pattern is that the registry signal is confounded — the patients whose carbon dioxide falls furthest are the patients who were sickest, most acidotic and most hypercapnic to begin with, which is to say the patients who were always going to do worse.

That reading is entirely reasonable, and it is why the next study matters more than all seven above.

What to actually do

22.4 Oxygen — Chapter 19's argument, operationalised

Chapter 19 §19.5 established that reperfusion oxygen is an intervention, and quoted the registry data. This chapter does not re-argue it. It states the numbers once for reference, then moves to the bedside.

A second oxygen finding deserves separate mention, because it is the only study in the ECPR literature that examined both gases together and found the same shape in both.

The randomised evidence, such as it is, comes entirely from patients with a circulation. An updated systematic review of oxygen and carbon dioxide targets after cardiac arrest covering twelve trials found no difference in survival or favourable functional outcome between restrictive and liberal oxygen targets in either the prehospital or the intensive care setting, and no difference between mild hypercapnia and normocapnia. Certainty was rated low to moderate. That is the honest backdrop: in patients whose own hearts are beating, the level does not appear to matter much within the ranges tested. The ECPR-specific concern is not the level but the excursion, and no trial has tested an excursion.

22.5 Arterial pressure — and why the optimum moves when the heart comes back

The third knob is the one with the most published numbers and the least agreement about what they mean.

What the sources recommend

Two textbooks, three numbers, all of them narrow, none of them evidenced. The reason they are narrow is that they are bounded on both sides by different fears: too low and the brain and kidneys are underperfused; too high and the left ventricle cannot open the aortic valve against the circuit (Chapter 13 §13.4, Chapter 15).

What the human data show

The ECPR observational data describe an inverted-U with its peak somewhere in the seventies. The randomised data, from patients with a circulation, say the target does not matter within the range anyone has tested. The two are usually reconciled by dismissing the observational finding as confounding — a higher pressure marks a patient whose heart is working, not a patient whose brain is being protected.

Then there is an experiment which suggests the reconciliation is more interesting than that.

What to actually do

Given an unevidenced target, a narrow textbook consensus, an inverted-U observational signal and one animal experiment, the honest position is a range and a set of reasons to move within it.

22.6 Flow — and the thing flow is competing with

Flow is the one variable in this chapter that the general VA chapters already handle well, so this section is short and confined to what differs.

The ECPR-specific tension is the familiar one from Chapter 13, arriving earlier and more sharply. Raising the flow raises aortic pressure, which raises left ventricular afterload, which reduces whatever native contractility is trying to return. In a shock patient that trade-off is played out over hours. In an ECPR patient it is played out in the first thirty minutes, in a ventricle that has been fibrillating or standing still, at a moment when nobody yet knows whether that ventricle is going to recover at all.

The textbook's formulation of the compromise is the right one and is worth quoting as a principle rather than a number: support should be adjusted to achieve adequate organ perfusion and yet reduce aortic pressure enough to allow for any potential native left ventricular ejection.

22.7 The first echocardiogram and the first electrocardiogram

Both textbooks place these in the first minutes, and both give them jobs that are more specific than "assess function."

22.8 Coronary angiography — for everyone, or for some?

This is the largest decision in the chapter, and it is the one where the guidance and the data point in different directions.

What the sources say

The Red Book is unambiguous. "Emergent coronary angiography should be undertaken for all ECPR patients without an obvious noncardiac cause, independent of age and presenting rhythm." Table 32-4 lists coronary angiography first among the early post-arrest actions, alongside computed tomography pulmonary angiography, comprehensive echocardiography and repeated electrocardiograms. The stated rationale is that studies using protocolised catheterisation after ECPR initiation demonstrated increased survival.

Why the ECPR patient is not the post-arrest patient

The single most robust finding in this literature is that the coronary anatomy of a patient whose arrest was refractory is worse than the coronary anatomy of a patient whose arrest was not.

That matched comparison contains the key observation, and it is easy to read past. The acute culprit lesion is the same in both groups. What differs is the chronic disease burden underneath it. The same occlusion, dropped onto a heart with left main disease and a chronic total occlusion, produces an arrest that will not respond to defibrillation. Dropped onto a heart with single-vessel disease, it produces an arrest that will.

The coronary anatomy explains why the arrest was refractory. That is a different claim from saying it explains what happens next.

What happens when the lesion is treated

What to actually do

22.9 Temperature — you cannot not choose

Every other intervention in this chapter can be withheld. Temperature cannot, and this is the physiological fact that the entire temperature literature in ECPR has failed to take seriously.

This matters because it reframes the entire evidence base. The ECPR temperature literature is not comparing an intervention against no intervention. It is comparing two temperatures, both of which were produced by a machine, one of which was chosen deliberately and the other by default. A null result in such a comparison means something quite different from a null result in a conventional targeted-temperature-management trial.

The evidence, laid out

The finding that should be uncomfortable

22.10 The heparin bolus — one decision this chapter owns

Anticoagulation on ECMO belongs to Part VII, and this chapter does not attempt it. But there is exactly one anticoagulation decision that is made inside this chapter's window, by this chapter's team, and it is ECPR-specific: whether to give a heparin bolus at the moment the circuit is connected.

The Red Book's connection sequence says it plainly: "A heparin bolus can be given as soon as the system is connected and adjusted accordingly per institutional protocols. This may be omitted in hypothermic or near-drowning victims because of the risk of hemorrhage." The ECPR textbook agrees that systemic anticoagulation should be provided for all patients in the absence of bleeding, while noting that ECPR patients in particular are at high risk of bleeding, may develop disseminated intravascular coagulation after prolonged arrest times, and are often about to receive antiplatelet agents in the catheterisation laboratory.

Set against that, the patient in front of you has just had their femoral vessels punctured several times under compressions, has had a chest wall compressed for forty minutes, and may be going to a coronary intervention within the hour. The bleeding literature confirms the fear: bleeding was the most common complication in the 1,644-patient SAVE-J II cohort quoted in Chapter 21, with cannulation-site bleeding in 16.4% and other haemorrhage in 8.5%.

22.11 The lungs — post-arrest pulmonary oedema and the two ventilators

Chapter 53 owns mechanical ventilation during ECMO. This section covers only what is different when the patient has just been resuscitated, and it is different in two ways.

First, the lung is now one of two gas exchangers in parallel, and it is the weaker one. Pulmonary blood flow is whatever the native right ventricle is pushing, which may be very little. The consequences are practical:

• Minute ventilation must come down. The ECPR textbook: reduced minute ventilation is used to achieve a low-normal end-tidal carbon dioxide, because normal minute ventilation against a reduced pulmonary blood flow produces significant respiratory alkalosis within the pulmonary circulation.
• The end-tidal trace is now a native-circulation monitor (§22.3), not a ventilation monitor.
• Carbon dioxide can be controlled from either device. Sweep gas changes of 1–2 L/min with a blood gas after each, or ventilator minute ventilation changes, both work. The sweep is the more powerful lever and therefore the more dangerous one (§22.3).

Second, post-arrest pulmonary oedema in an ECPR patient is a mechanical problem until proven otherwise. The Red Book's Table 32-4 makes the diagnostic order explicit, and the order is the message:

22.12 The four-hour systems review

Both textbooks converge on the same structure: an immediate phase measured in minutes, and an early phase measured in hours. The Red Book splits it into Table 32-3 and Table 32-4; the ECPR textbook frames the first four hours as a discrete task with the dual goals of stabilising the patient and identifying early complications, followed by a comprehensive systems review once the emergency is over.

The review matters because of how these patients arrive. In the ECPR textbook's own description: they arrive at an emergency department with little warning and little history, undergo cannulation, and are then moved for imaging and intervention. By the time they reach an intensive care bed, nobody in the room has yet had a moment to think about them as a whole patient — and several time-limited decisions have quietly been running in the background.

22.13 What the evidence base for this chapter actually is

This section exists because a reader who has come through Chapters 19, 20 and 21 will arrive here expecting the evidence to get better, and it gets dramatically worse. It is worth saying why, and saying it precisely.

Into that vacuum, in 2026, came the first structured attempt at guidance.

22.14 The first ten minutes, as one figure

Everything above resolves into a short sequence. It is drawn here not because it is complicated but because the failures are failures of omission, and a list that is visible is harder to omit.

The figure has no branch for "give the heparin bolus" because §22.10 concluded that it is a unit-level decision rather than a bedside one, and no branch for prognostication because that is Chapter 23's.

22.15 The errors that recur

22.16 Key points

1. The patient handed over by Chapter 21 differs from every other venoarterial patient in one respect that governs everything: they had no circulation at all a few minutes ago. Most of their brain injury is already inflicted and cannot be undone. The first hours are, however, the one window in which it can be made considerably worse using entirely conventional-looking actions.
2. Three potent cerebral vasoactive variables — carbon dioxide, oxygen and arterial pressure — all move in the direction of maximum insult within the first two minutes, by default, and all three are set by knobs. Assign all three to one named person at the moment compressions stop. This is the operational form of Chapter 19 §19.5.
3. Carbon dioxide is the steepest gradient available. Between 20 and 60 mmHg, 1 mmHg changes cerebral blood flow by 2–3%, with onset in two minutes and equilibrium in twelve. The ECPR patient starts above that range and the membrane lung can cross it in minutes.
4. Three ELSO Registry populations associate a large early carbon dioxide fall with neurological complications, and the two best-measured human studies do not. A porcine ECPR experiment, which removes the confounding objection, found that rapid correction impaired cerebral autoregulation and produced histological neuronal injury. Set the initial sweep gas low, leave it for ten minutes, and titrate 1–2 L/min at a time against right-radial gases towards roughly 40 mmHg.
5. Severe hyperoxia above 300 mmHg carries the clearest associated harm in the only ECPR-specific registry analysis — adjusted odds ratios of 1.59 for acute brain injury and 1.58 for death in 3,125 patients. Target a saturation of 92–97% by blending, and come off 100% as soon as flow is confirmed. But do not chase a restrictive target downward: the randomised post-arrest evidence gives no support for it, and the one ECPR dataset examining the shape found harm at both ends.
6. The arterial pressure target is unevidenced and the textbooks agree on a narrow band anyway: roughly 65–80 mmHg, measured from the right upper limb. The observational ECPR data describe an inverted U peaking in the seventies; four randomised trials in patients with a circulation found no difference between targets.
7. A porcine experiment suggests the optimal pressure inverts at the moment the native heart returns — a high target improved cerebral autoregulation during extracorporeal resuscitation and worsened it after return of circulation. Six animals. Treat the return of native ejection as a prompt to reassess the pressure target, not merely as good news.
8. The coronary anatomy of a refractory arrest is worse than that of a non-refractory one — left main involvement 17% against 5.7% — but the acute culprit lesion is the same. The chronic disease underneath is what made the arrest refractory.
9. Angiography after ECPR is a diagnostic act, not a reperfusion-time act, because the pump has already restored coronary perfusion — which is why three-quarters of patients with obstructive disease organised their rhythm before intervention, and why a careful registry analysis found no survival association with percutaneous intervention across three causal-inference methods. The case for near-universal angiography stands; the case for doing it before stabilising does not.
10. On ECMO you cannot decline to choose a temperature, because the circuit is a heat exchanger and its default is cooling. "No targeted temperature management" in this literature frequently means 35.6°C. Choose a target in the 33–36°C band, document it, hold it; prefer 36°C if bleeding is difficult to control.
11. The pooled randomised evidence for hypothermia after ECPR shows no benefit; the pooled observational evidence does, and only at 32–34°C. The largest registry analysis finds hypothermia associated with survival but not with favourable neurological outcome — a dissociation that implies more survivors in poor neurological states and that nobody has addressed. A randomised trial in this population is under way.
12. The heparin bolus at connection is optional rather than mandatory, and a unit should decide in its protocol rather than at the syringe. Withholding it was associated with less early major bleeding and no excess thrombosis in a 59-patient study whose adjusted estimate was not significant.
13. After ECPR, unstable circuit flow means look for blood and tamponade before looking at the circuit; new pulmonary oedema means exclude left ventricular distension before treating the lung; new hyperkalaemia means look at the limb.
14. Do not cardiovert immediately. A myocardium that has been ischaemic for forty minutes will not hold a rhythm; the same myocardium after a few minutes of perfusion frequently will, and the wait costs nothing because the circulation is now mechanical.
15. This is the weakest evidence base in Part IV, and the reason is structural. Three randomised trials tested ECPR and every one of them randomised the decision to cannulate; none randomised anything afterwards. We have randomised evidence that ECPR works and none at all about how to run it. Of 133 post-ECPR studies, 90% are retrospective and 17% have more than 500 patients. The only structured guidance is a 2026 international Delphi of 52 experts.
16. Write the protocol anyway. The programme that moved from 9.1% to 69.2% good-outcome survival did so with a bundle that included standardised post-resuscitation care. Which component worked is unknown. A merely reasonable protocol beats merely reasonable improvisation, because it is the same every time and it can be audited.

Cross-references

Inherited by this chapter

• Chapter 13 — the three circulations, the moving mixing point, pulsatility as a readout of native ejection, and why the right radial is the sampling site that matters.
• Chapter 14 — the general initiation sequence on venoarterial ECMO. This chapter covers only what differs.
• Chapter 15 — left ventricular distension, the pause-the-balloon manoeuvre, the nine-strategy mechanical menu, and §15.9 on unloading in ECPR specifically, which this chapter cites and does not re-argue.
• Chapter 16 — differential hypoxaemia, and why a low right-radial oxygen tension is a configuration question before it is a blender question.
• Chapter 17 — limb ischaemia, distal perfusion routes and the four-hour window.
• Chapter 19 — §19.1 the absent neurological baseline; §19.2 the coronary-perfusion-pressure physiology and the end-tidal gate; §19.5 reperfusion as an intervention, which this chapter operationalises; §19.7 the three trials ordered by system maturity; §19.9 the programme as the unit of intervention and the Harefield bundle.
• Chapter 20 — §20.8 the refusal log, whose logic §22.13 extends from selection to management.
• Chapter 21 — the cannulation itself; the right radial line; the distal perfusion cannula; §21.4 the moment defibrillation stops, whose positive half is §22.7.

Handed forward by this chapter

• Chapter 23 — all neuroprognostication. Continuous electroencephalography appears in both chapters with different jobs: here a seizure monitor, there an instrument of prognosis. Chapter 23 also inherits the survival-without-neurological-benefit dissociation from §22.9, and the practice-survey finding that electroencephalography is routinely applied in only 13.9% of ECPR institutions.
• Chapter 24 — the decision to stop. §22.8's "a brain that is very unlikely to recover is a legitimate reason to defer angiography" is a Chapter 24 conversation held early, and it must be a named conversation rather than a silent omission.
• Chapter 27 — blood gas and oxygenation monitoring as a discipline.
• Chapter 37 — haemolysis, and the dark pink urine rule.
• Chapter 28 — cerebral monitoring as a set of instruments: near-infrared spectroscopy, transcranial Doppler and electroencephalography. This chapter orders them, Chapter 28 explains them, and Chapter 23 uses them to predict.
• Chapter 46, Chapter 55, Chapter 57 — transfusion thresholds, renal replacement (required in close to two-thirds of ECPR cases), and sedation with the ECMO sequestration problem.
• Chapter 53 — mechanical ventilation during ECMO. This chapter owns only the ECPR-specific reduction in minute ventilation and the PEEP-as-afterload argument.
• Chapter 11 — the propensity-matched finding that emergency coronary artery bypass outperformed percutaneous intervention in ECPR patients with triple-vessel disease, recorded in §22.8 and deliberately not acted on here. No chapter in this book currently owns revascularisation strategy on ECMO; Chapter 11 is the nearest home.
• Chapter 65 — accidental hypothermia, which is a different disease, and to which the rewarming-rate contradiction in §22.9 may belong.
• Chapter 59 — mobilisation and rehabilitation, which inherits the long-term multidimensional outcome data in ECPR survivors: left ventricular ejection fraction recovers substantially by 22 months while neuropsychological and physical function does not. Chapter 76, palliative care, inherits the same study's bearing on what survivorship actually looks like.
• Chapter 77 and Chapter 91 — the disposition question, the 30-runs-a-year threshold, and the pre-existing referral pathway.
• Part VII (Chapters 39–45) — everything about anticoagulation after the first hour.

References

Every external source below is known to this chapter through a structured abstract only. Nothing was retrieved in full text, and all rows seeded from this chapter are marked Verified = No.

Post-ECPR care — reviews, consensus, and the evidence gap

1. Dennis M, et al. Recommendations in extracorporeal cardiopulmonary resuscitation post-resuscitation care via an international, modified Delphi approach. Journal of the American Heart Association. 2026. DOI 10.1161/jaha.125.047548. NCT06552312. 52 of 53 experts, 3 rounds, 126 statements, 42 position statements. The only structured guidance that exists for this chapter's material.
2. Dennis M, et al. Extracorporeal cardiopulmonary resuscitation post-resuscitation care: a modified Delphi study with the CARES expert panel (abstract 123). ASAIO Journal. 2025. DOI 10.1097/01.mat.0001168852.56957.38. Reports 125 statements and 34 position statements across 7 domains for the same study — the discrepancy is flagged in §22.13.
3. Scquizzato T, et al. Post-resuscitation care after adult extracorporeal cardiopulmonary resuscitation: a scoping review. Resuscitation. 2025. DOI 10.1016/j.resuscitation.2025.110880. 133 studies; 90% retrospective; 17% with more than 500 patients. The quantitative statement of the gap.
4. Kang JK, et al. Post-cardiac arrest care in extracorporeal cardiopulmonary resuscitation. Critical Care Medicine. 2023. DOI 10.1097/ccm.0000000000006102.
5. Purington E, et al. Immediate post-ECPR management strategies in the prehospital and critical care transport medicine environments. Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine. 2025. DOI 10.1186/s13049-025-01448-6. The 4 to 6 hour window, for transport and non-ECMO-capable centres.
6. Supady A, et al. Extracorporeal cardiopulmonary resuscitation for refractory cardiac arrest. The Lancet Respiratory Medicine. 2025. DOI 10.1016/s2213-2600(25)00122-5.
7. Inoue A, et al. Extracorporeal cardiopulmonary resuscitation for out-of-hospital cardiac arrest in adult patients. Journal of the American Heart Association. 2020. DOI 10.1161/jaha.119.015291.
8. Han KS, et al. Experience of extracorporeal cardiopulmonary resuscitation in a refractory cardiac arrest patient at the emergency department. Clinical Cardiology. 2019. DOI 10.1002/clc.23169. The source of the 57.0% of deaths within 24 hours, and of 12 of 14 survivors having a good neurological outcome.

Carbon dioxide and oxygen

1. Cavayas YA, et al. The early change in PaCO₂ after extracorporeal membrane oxygenation initiation is associated with neurological complications. American Journal of Respiratory and Critical Care Medicine. 2020. DOI 10.1164/rccm.202001-0023oc. ELSO Registry, 11,972 patients.
2. Shah N, et al. Early changes in arterial partial pressure of carbon dioxide and blood pressure after starting extracorporeal membrane oxygenation in children: Extracorporeal Life Support Organization database study of neurologic complications. Pediatric Critical Care Medicine. 2023. DOI 10.1097/pcc.0000000000003216. 7,270 runs. The carbon-dioxide-by-pressure interaction.
3. Joram N, et al. Association between early change in arterial carbon dioxide tension and outcomes in neonates treated by extracorporeal membrane oxygenation. ASAIO Journal. 2022. DOI 10.1097/mat.0000000000001838.
4. Shou BL, et al. Arterial carbon dioxide and acute brain injury in venoarterial extracorporeal membrane oxygenation. ASAIO Journal. 2022. DOI 10.1097/mat.0000000000001699.
5. Shou BL, et al. Arterial oxygen and carbon dioxide tension and acute brain injury in extracorporeal cardiopulmonary resuscitation patients: analysis of the Extracorporeal Life Support Organization Registry. The Journal of Heart and Lung Transplantation. 2022. DOI 10.1101/2022.03.10.22272203. 3,125 ECPR patients — the hyperoxia analysis used in Chapter 19 §19.5 and restated in §22.4. [VERIFICATION REQUIRED] — the retrieved identifier is a preprint-server DOI attached to a journal citation, the same defect recorded for one of Chapter 19's sources.
6. Yu Y, et al. Association of early changes in arterial carbon dioxide with acute brain injury in adult patients with extracorporeal membrane oxygenation: a ten-year retrospective study in a German tertiary care hospital. Journal of Critical Care. 2024. DOI 10.1016/j.jcrc.2024.154880. The negative study, 618 patients.
7. Thiara S, et al. Characterizing the relationship between arterial carbon dioxide trajectory and serial brain biomarkers with central nervous system injury during veno-venous extracorporeal membrane oxygenation. Neurocritical Care. 2024. DOI 10.1007/s12028-023-01923-x. The biomarker study whose biomarkers did not follow the carbon dioxide.
8. Chen X, et al. Impact of early PaCO₂ and pH fluctuations on neurological outcomes in ARDS patients receiving VV ECMO: a retrospective cohort study from the CSECLS registry. Annals of Intensive Care. 2025. DOI 10.1186/s13613-025-01570-9.
9. Cao MF, et al. Exploring the association between early PaCO₂ correction speed and cerebrovascular autoregulation in a porcine model of extracorporeal resuscitation. Translational Stroke Research. 2025. DOI 10.1007/s12975-025-01376-8. The causal experiment behind Controversy 1, and the source of the sweep-gas-as-a-fraction-of-blood-flow protocol and of the CA-3 discrepancy flagged in §22.3.
10. Marchetto L, et al. It all circles back to cerebral autoregulation: understanding the risk of hypocapnia and arterial hypertension when initiating pediatric extracorporeal membrane oxygenation. Pediatric Critical Care Medicine. 2023. DOI 10.1097/pcc.0000000000003290. An editorial, and the source of the 2 to 3% per mmHg coefficient, the 2-minute onset and 12-minute equilibrium, and the 36-hour adaptation.
11. Hong S, et al. Optimal arterial blood gas tensions for the prognosis of favorable neurological outcomes in survivors after extracorporeal cardiopulmonary resuscitation. Journal of Clinical Medicine. 2022. DOI 10.3390/jcm11144211. The U-shaped analysis of both gases.
12. Holmberg MJ, et al. Oxygen and carbon dioxide targets after cardiac arrest: an updated systematic review. Resuscitation. 2025. DOI 10.1016/j.resuscitation.2025.110620. 12 randomised trials, all in patients with return of circulation.

Arterial pressure and flow

1. Lee Y, et al. Optimal mean arterial pressure for favorable neurological outcomes in survivors after extracorporeal cardiopulmonary resuscitation. Journal of Clinical Medicine. 2021. DOI 10.3390/jcm11020290.
2. Sun F, et al. Average mean arterial pressure in the first 6 hours of extracorporeal cardiopulmonary resuscitation in the prediction of the prognosis of neurological outcome. Perfusion. 2021. DOI 10.1177/02676591211027118.
3. Levy Y, et al. Targeted high mean arterial pressure aggravates cerebral hemodynamics after extracorporeal resuscitation in swine. Critical Care. 2021. DOI 10.1186/s13054-021-03783-3. The inversion at return of spontaneous circulation. Six animals per group.
4. Ryu JA, et al. Neurological outcomes in patients who undergo extracorporeal cardiopulmonary resuscitation. The Annals of Thoracic Surgery. 2019. DOI 10.1016/j.athoracsur.2019.03.033. The nECPR score; the pulse pressure under 25 mmHg and mean pressure under 70 mmHg thresholds.
5. Saemann L, et al. A systematic review with meta-analysis investigating the impact of targeted perfusion parameters during extracorporeal cardiopulmonary resuscitation in out-of-hospital and in-hospital cardiac arrest. The Journal of Extra-Corporeal Technology. 2022. DOI 10.1182/ject-191-202.
6. Niemelä V, et al. Higher versus lower blood pressure targets after cardiac arrest: systematic review with individual patient data meta-analysis. Resuscitation. 2023. DOI 10.1016/j.resuscitation.2023.109862. Four randomised trials, 1,087 patients, none on ECMO.
7. Kjaergaard J, et al. Blood-pressure targets in comatose survivors of cardiac arrest. The New England Journal of Medicine. 2022. DOI 10.1056/nejmoa2208687. NCT03141099. BOX.

Coronary angiography and revascularisation

1. Scquizzato T, et al. Coronary angiography findings in resuscitated and refractory out-of-hospital cardiac arrest: a systematic review and meta-analysis. Resuscitation. 2023. DOI 10.1016/j.resuscitation.2023.109869. 128 studies, 62,845 patients. The left main and left anterior descending comparison.
2. Franco D, et al. Coronary disease in refractory cardiac arrest undergoing resuscitation with extracorporeal membrane oxygenation. European Heart Journal: Acute Cardiovascular Care. 2023. DOI 10.1093/ehjacc/zuad012. The matched comparison in which the chronic burden differed and the acute culprit lesion did not.
3. Yannopoulos D, et al. Coronary artery disease in patients with out-of-hospital refractory ventricular fibrillation cardiac arrest. Journal of the American College of Cardiology. 2017. DOI 10.1016/j.jacc.2017.06.059.
4. Alhuneafat L, et al. The effect of percutaneous coronary intervention after extracorporeal cardiopulmonary resuscitation on survival for out-of-hospital cardiac arrest: a causal inference analysis. Resuscitation Plus. 2025. DOI 10.1016/j.resplu.2025.101103. 576 ELSO Registry patients; null across three methods.
5. Crespo-Diaz R, et al. Effects of perfusion, coronary artery disease burden, and revascularization in establishing organized cardiac rhythm during extracorporeal cardiopulmonary resuscitation for shockable refractory out-of-hospital cardiac arrest. Journal of the American Heart Association. 2024. DOI 10.1161/jaha.123.033907. The finding that three-quarters organised before intervention — and the source of the population-definition discrepancy flagged in §22.8.
6. Bogerd M, et al. Coronary angiography in cardiac arrest patients undergoing extracorporeal cardiopulmonary resuscitation. Netherlands Heart Journal. 2026. DOI 10.1007/s12471-026-02049-3.
7. Schussler JM, et al. Selective cardiac angiography following extracorporeal cardiopulmonary resuscitation for out-of-hospital cardiac arrest (abstract 252). ASAIO Journal. 2024. DOI 10.1097/01.mat.0001069916.55559.20.
8. Righetti S, et al. Coronary angiography findings in patients with refractory out-of-hospital cardiac arrest treated with extracorporeal cardiopulmonary resuscitation: prediction of lesion culprit of acute myocardial infarction. Resuscitation. 2025. DOI 10.1016/j.resuscitation.2025.110717.
9. Fu HY, et al. Emergent coronary revascularization with percutaneous coronary intervention and coronary artery bypass grafting in patients receiving extracorporeal cardiopulmonary resuscitation. European Journal of Cardio-Thoracic Surgery. 2024. DOI 10.1093/ejcts/ezae290. No chapter currently owns revascularisation strategy; nearest home Chapter 11.
10. Lemkes JS, et al. Coronary angiography after cardiac arrest without ST-segment elevation. The New England Journal of Medicine. 2019. DOI 10.1056/nejmoa1816897. COACT — patients with return of circulation, not ECPR.

Temperature

1. Huang M, et al. Does targeted temperature management improve neurological outcome in extracorporeal cardiopulmonary resuscitation? Journal of Intensive Care Medicine. 2021. DOI 10.1177/08850666211018982. 35 studies, 2,643 patients.
2. Sakurai T, et al. Extracorporeal cardiopulmonary resuscitation with temperature management could improve the neurological outcomes of out-of-hospital cardiac arrest. Journal of Intensive Care. 2022. DOI 10.1186/s40560-022-00622-7. The 249-minute median time to target temperature.
3. Watanabe M, et al. The impact of different targeted temperatures on out-of-hospital cardiac arrest outcomes in patients receiving extracorporeal membrane oxygenation: a nationwide cohort study. Critical Care. 2022. DOI 10.1186/s13054-022-04256-x.
4. Kim JH, et al. Target temperature management effect on the clinical outcome of patients with out-of-hospital cardiac arrest treated with extracorporeal cardiopulmonary resuscitation: a nationwide observational study. Journal of Personalized Medicine. 2024. DOI 10.3390/jpm14020185.
5. Kim YS, et al. Target temperature management may not improve clinical outcomes of extracorporeal cardiopulmonary resuscitation. Journal of Intensive Care Medicine. 2018. DOI 10.1177/0885066618801269. The 33.4 against 35.6°C comparison, and the authors' own observation that extracorporeal support produces the temperature.
6. Lee CC, et al. Effect of targeted temperature management on neurological and survival outcomes in patients undergoing extracorporeal cardiopulmonary resuscitation. PLOS One. 2026. DOI 10.1371/journal.pone.0342473.
7. Wang J, et al. What is the optimal temperature control strategy in patients receiving ECPR after cardiac arrest? A network meta-analysis. The American Journal of Emergency Medicine. 2024. DOI 10.1016/j.ajem.2024.11.001. Positive on direct comparison, null on the network.
8. Miao B, et al. Survival and neurological outcome in patients treated with extracorporeal membrane oxygenation and therapeutic hypothermia: an updated systematic review and meta-analysis. Frontiers in Medicine. 2026. DOI 10.3389/fmed.2026.1882223. PROSPERO CRD42023435353. The randomised-versus-observational separation that Controversy 3 is built on.
9. Inoue A, et al. Temperature control after extracorporeal cardiopulmonary resuscitation: a retrospective multicenter study in Japan (abstract Sun301). Circulation. 2025. DOI 10.1161/circ.152.suppl_3.sun301. The survival-without-neurological-benefit dissociation, and the announcement of a randomised trial in this population.
10. Kanda J, et al. Target temperature management and post-extracorporeal cardiopulmonary resuscitation outcome: a post hoc analysis of the SAVE-J II study. Preprint. 2023. DOI 10.1101/2023.06.15.23291462. The earlier version of the same analysis, with different denominators and point estimates — flagged in §22.9.
11. Miyamoto S, et al. Association between the rewarming duration and neurological outcomes after extracorporeal cardiopulmonary resuscitation followed by targeted temperature management for out-of-hospital cardiac arrests: a secondary analysis of the SAVE-J II study. Therapeutic Hypothermia and Temperature Management. 2024. DOI 10.1089/ther.2024.0036.
12. Hifumi T, et al. Details of targeted temperature management methods for patients who had out-of-hospital cardiac arrest receiving extracorporeal cardiopulmonary resuscitation: a questionnaire survey. Therapeutic Hypothermia and Temperature Management. 2022. DOI 10.1089/ther.2022.0004. All 36 SAVE-J II institutions. The source of the 13.9% routine electroencephalography figure.
13. Zhang X, et al. Impact of differential temperature trajectories following targeted temperature management on outcomes in out-of-hospital cardiac arrest patients receiving ECPR. Resuscitation. 2025. DOI 10.1016/j.resuscitation.2025.110932. The temperature-variability finding, read here as reverse causation.
14. Lascarrou JB, et al. Targeted temperature management for cardiac arrest with nonshockable rhythm. The New England Journal of Medicine. 2019. DOI 10.1056/nejmoa1906661. NCT01994772. HYPERION.
15. Fernando SM, et al. Targeted temperature management following out-of-hospital cardiac arrest: a systematic review and network meta-analysis of temperature targets. Intensive Care Medicine. 2021. DOI 10.1007/s00134-021-06505-z. The arrhythmia signal, rated high certainty.
16. Granfeldt A, et al. Targeted temperature management in adult cardiac arrest: systematic review and meta-analysis. Resuscitation. 2021. DOI 10.1016/j.resuscitation.2021.08.040.
17. Böttiger BW, et al. The effectiveness of targeted temperature management following cardiac arrest may depend on bystander cardiopulmonary resuscitation rates. European Journal of Anaesthesiology. 2022. DOI 10.1097/eja.0000000000001663. The insult-severity argument. A letter, not an original analysis.

Anticoagulation at cannulation

1. Redondo EA, et al. Heparin exposure adjustment to reduce thrombo-hemorrhagic complications after venoarterial extracorporeal membrane oxygenator cannulation: the HEART-ECMO observational cohort study. Journal of Cardiothoracic and Vascular Anesthesia. 2026. DOI 10.1053/j.jvca.2026.01.025. 59 patients. Direction used, adjusted point estimate not.

Programmes and outcomes

1. Akhtar W, et al. Improving quality and outcomes of extracorporeal cardiopulmonary resuscitation in refractory cardiac arrest: the Phoenix ECPR project. BMJ Open Quality. 2025. DOI 10.1136/bmjoq-2024-002934. The bundle that included standardised post-resuscitation care.
2. Thevathasan T, et al. Multi-dimensional outcomes following extracorporeal cardiopulmonary resuscitation. Resuscitation Plus. 2025. DOI 10.1016/j.resplu.2025.100888. Held for Chapter 59, mobilisation and rehabilitation, and Chapter 76, palliative care.

Textbooks

1. ELSO Red Book, 6th edition, Chapter 32 — the connection sequence including the heparin bolus statement and the cardioversion statement; Table 32-3, CPR-specific immediate post-arrest management (minutes); Table 32-4, CPR-specific early post-arrest management (within hours); the coronary angiography recommendation; the 4-hour distal perfusion window. [VERIFICATION REQUIRED] — page numbers not confirmed.
2. Shinar Z, Badulak J. ECPR and Resuscitative ECMO, Chapter 9 — the first four hours, in full: monitoring minimums, arterial line placement, circuit pressures and delta P, flow and central venous oxygen saturation targets, vasopressor and inotrope targets, gas exchange including the PEEP and minute-ventilation arguments, initial echocardiography and electrocardiography, temperature management and rewarming, anticoagulation and transfusion, the post-arrest systems review, and disposition. [VERIFICATION REQUIRED].
3. ISCCM Manual of RRT and ECMO in ICU — bleeding mechanisms on ECMO, thrombocytopenia and acquired von Willebrand syndrome; Chapter 35 on sedation, including the fentanyl sequestration data. Held largely for Part VII and Chapter 57. [VERIFICATION REQUIRED].
4. Taha AR, Caridi-Scheible M, Leiendecker E, et al. ECMO: A Practical Guide to Management — consulted for post-cannulation management; not separately quoted in this chapter. [VERIFICATION REQUIRED].