Part II — VV ECMO · Chapter 7
Evidence search date: 6 September 2026. Two source documents held in the project library were read in full for this chapter: the 2024 Perfusion review of mechanical ventilation during ECMO, and the 2022 Intensive Care Medicine summary of anticoagulation management. Both are cited directly.
Clinical Question
The circuit is running. What is set, what is stopped, what is started, and what is checked — in the first minutes, the first hours, and the first day?
Why This Matters
There is a specific reason this chapter is not an administrative appendix to Chapter 6.
Evidence — the first three days are the ones that count
In a pooled individual patient data analysis of earlier studies of mechanical ventilation during ECMO for ARDS, totalling 545 patients, driving pressure during the first three days of ECMO was the sole independent risk factor for mortality (Serpa Neto A, et al. Intensive Care Med 2016, as reported in Szuldrzynski K, et al. Perfusion 2024).
Separately, a retrospective analysis across three high-volume ECMO centres found higher PEEP during the first days of ECMO associated with reduced mortality (Schmidt M, et al. Crit Care Med 2015, as reported in the same review).
Certainty: moderate. Observational and pooled-observational data, vulnerable to confounding by severity — a sicker lung tolerates less. But the signal is consistent, and it is specific to the early period.
The implication is uncomfortable and useful: the ventilator settings chosen in the first hour of ECMO may matter more than most of what follows. This chapter is organised around that.
The First Ten Minutes
The physiological reasoning for each of these is in Chapter 6 and is not repeated here.
Action | Setting | Why |
Establish flow | Increase to the target, or to maximum sustainable flow — chatter and erratic flow define the ceiling | Chapter 6. Effective flow relative to cardiac output sets arterial saturation |
FdO₂ | 1.0 | Standard for emergent VV initiation; weaned later |
Sweep gas | Deliberately low — commonly about half the blood flow | To avoid abrupt CO₂ correction. See the Danger below |
Temperature | Set the heat exchanger to normothermia unless hypothermia is intended | A circuit outside the body is a large heat-exchange surface; patients cool fast, and shivering raises VO₂ (Chapter 2) |
Anticoagulation | Continue the infusion started around cannulation, to an early target | See Starting Anticoagulation below |
Blood gases | Arterial, pre-oxygenator and post-oxygenator | Establishes the three-point baseline every later interpretation depends on (Chapters 2 and 6) |
Echo and radiograph | Confirm cannula position; screen for complications | Chapter 5. The distal cannula is radiolucent, so the radiograph is not the answer on its own |
Danger — the two things that go wrong in the first ten minutes
Abrupt CO₂ correction. The membrane can normalise a PaCO₂ of 100 mmHg within minutes; cerebral vasoconstriction and a rapid CSF pH shift follow. Correct over roughly 4–8 hours (Chapter 6). This is what the low starting sweep is for. Resist the urge to fix the first blood gas.
Hypotension. VV ECMO is haemodynamically neutral by construction — the same volume is drained and returned to the venous compartment. So hypotension at initiation is not the circuit doing something to the circulation. It has other causes, and they need naming rather than treating reflexively with flow.
Why the pressure falls when the circuit starts
Cause | Mechanism | Response |
Dilution by the prime | An adult circuit prime is roughly 500–700 mL (Chapter 3), delivered as a bolus of non-oxygen-carrying fluid | Expect a fall in haemoglobin and in platelets. Anticipate it rather than discovering it |
Sedation and neuromuscular blockade | Given at or around cannulation; removes sympathetic drive | Vasopressor, titrated — not fluid by reflex |
Blood–surface contact | Inflammatory activation on first exposure to the circuit | Usually transient; supportive |
Unmasked hypovolaemia | Drainage reveals a volume state that positive-pressure ventilation was concealing | Judicious volume — with the warning below |
Tamponade, pneumothorax, bleeding | Procedural complications of cannulation | Echo immediately. Chapter 5 |
Pitfall — the fluid the patient spends the next week losing
The reflex at initiation is to give volume: the pressure is low, the circuit chatters, and fluid fixes both in the moment. Chapter 6 gave the counter-argument — higher-than-necessary flow requires greater volume expansion, worsening extravascular lung water and complicating later diuresis and sedation reduction (Shekar K, et al. J Thorac Dis 2020).
The de-resuscitation problem of day five is created on day zero. Use vasopressor for vasoplegia, volume for hypovolaemia, and be able to say which one you are treating.
The Ventilator Change
This is the single most consequential act of the first hour, because it is the moment the reason for cannulating is actually delivered. ECMO was started to take over gas exchange; the benefit only materialises if the ventilator is then turned down.
Evidence — what initiation actually changes
In an international multicentre prospective study of 350 patients, initiation of ECMO produced these changes in ventilation:
- Tidal volume 6.4 ± 2.0 → 3.7 ± 2.0 mL/kg predicted body weight
- Driving pressure 20 ± 7 → 14 ± 4 cmH₂O
- Respiratory rate 26 ± 8 → 14 ± 6 breaths/min
- Mechanical power 26.1 ± 12.7 → 6.6 ± 4.8 J/min
58% of patients were ventilated with driving pressure ≤ 15 cmH₂O and 65% with tidal volume ≤ 4 mL/kg. Notably, no association between day-two respiratory parameters and mortality was found in this cohort. (Schmidt M, et al. Am J Respir Crit Care Med 2019, as reported in Szuldrzynski K, et al. Perfusion 2024)
Certainty: moderate for the description of practice; low for causal inference. This describes what centres do and the magnitude of the change, not that any particular setting causes survival.
Recommended settings
Parameter | ELSO guidance | Contemporary practice (2023 survey of 48 centres) |
Plateau pressure | Limited to 30 cmH₂O; recommended < 25 cmH₂O | — |
PEEP | Maintained ≥ 10 cmH₂O | Personalised; historically often fixed for the whole run |
Respiratory rate | 4–15 breaths/min | < 10/min as the stated lung-rest goal |
Tidal volume | — | < 5 mL/kg in 54% of centres |
Driving pressure | — | Generally < 15 cmH₂O, with 58% declaring ≤ 10 cmH₂O |
FiO₂ | As little as possible | — |
ELSO guidance and survey figures as reported in Szuldrzynski K, et al. Perfusion 2024, citing the 2021 ELSO VV guideline and van Minnen O, et al. ASAIO J 2023.
Clinical Pearl — the governing principle
The ELSO guidance contains one sentence that resolves most first-day arguments: if gas exchange deteriorates, increase extracorporeal support rather than ventilator settings.
That is the whole logic of the therapy expressed as an instruction. The circuit exists so the lung does not have to work. Escalating the ventilator to fix a gas on ECMO spends the thing you cannulated the patient to protect.
Pitfall — respiratory rate is the forgotten half of mechanical power
Driving pressure gets the attention; rate quietly delivers energy too. In an analysis across several lung-protective ventilation populations, the relative contribution of respiratory rate to mechanical power was calculated as about 25% of that of driving pressure — so increasing rate by 4 breaths/min is roughly equivalent to increasing driving pressure by 1 cmH₂O (Costa ELV, et al. Am J Respir Crit Care Med 2021, as reported in Szuldrzynski 2024).
Applied to EOLIA: the ECMO arm ran a driving pressure of about 13 cmH₂O at a rate of about 22/min, which in mechanical-power terms is roughly equivalent to a driving pressure of 16 cmH₂O at a rate of 10/min. More protective than the control arm — but the review's own question stands: should it have been lower still?
Animal data suggest the relationship between ventilation intensity and lung injury is linear, with no safe threshold, which argues for as little mechanical power as the patient tolerates rather than for hitting a number.
Controversy — proning during ECMO, resolved as far as the evidence allows
Chapter 4 flagged that ELSO treats proning before ECMO as mandatory while the 2026 ANZ GRADE guideline gives a conditional recommendation against routine proning during ECMO. The underlying evidence, all as reported in Szuldrzynski 2024:
- Pooled individual patient data, nearly 900 patients from 5 studies: no survival benefit — but proning was applied late and in low doses, a median of two sessions beginning a median of five days after ECMO initiation (Giani M, et al. Crit Care 2022).
- Meta-analysis of 13 studies, 1,836 patients: better 28-day survival in proned patients — 74% (95% CI 71–77) versus 58% (95% CI 55–62); RR 1.31 (95% CI 1.21–1.41) — consistent across COVID and non-COVID populations (Papazian L, et al. Intensive Care Med 2022).
- PRONECMO randomised trial, predominantly COVID (94%) with very severe ARDS (mean compliance 15.0 mL/cmH₂O): no benefit of prone positioning on weaning from ECMO within 60 days, nor on ECMO- or ventilator-free days, length of stay, pressure injury, serious adverse events or 90-day mortality. No severe adverse events attributable to proning (Schmidt M, et al. JAMA 2023).
Current consensus: observational data favourable, the randomised trial neutral, and 74% of surveyed centres prone during VV ECMO anyway. Notably, most of these studies did not titrate PEEP by body position, even though compliance, transpulmonary pressure and PEEP requirement all change with proning.
Practical approach: proning during ECMO is safe in these data but is not established as beneficial. It is not an early-hours decision — Chapter 53 owns it.
Sedation and Neuromuscular Blockade
Chapter 2 made the case that VO₂ is a treatable variable, and Chapter 6 made it one of the four determinants of arterial saturation. The first hours are where that lever is used hardest — and where it is easiest to overuse.
Goal in the first hours | Rationale |
Suppress respiratory drive enough to allow the ventilator to be turned down | Vigorous spontaneous effort defeats lung rest and adds self-inflicted injury |
Reduce VO₂ | Lowers the demand side of the mixing equation — Chapters 2 and 6 |
Prevent shivering during temperature correction | Shivering is a large and avoidable VO₂ load |
Prevent movement that threatens cannula position | Chapter 5 |
Then start reducing it | Deep sedation is not a therapeutic goal; awake and mobilising ECMO is the direction of travel — Chapters 59, 83 |
Evidence — neuromuscular blockade
Early neuromuscular blockade in severe ARDS improved survival in the ACURASYS trial, but a later trial using lighter sedation in the control arm undermined that finding; a subsequent meta-analysis again indicated benefit. Guidelines diverge: the American Thoracic Society gives a conditional recommendation in favour, while ESICM withholds any recommendation.
In ECMO practice specifically, a 2023 survey found NMBAs administered by 93% of respondents but withdrawn as soon as possible, against 41% use in a 2019 prospective multicentre study. (All as reported in Szuldrzynski K, et al. Perfusion 2024)
Certainty: low to moderate, and contested. Practice is far ahead of the evidence, and the direction of that practice is short courses rather than routine continuation.
Starting Anticoagulation
Part VII owns this subject. What belongs here is the early target and one number that frames the whole problem.
Evidence — the scale of the problem, and the honesty about targets
An analysis of 7,579 VV ECMO patients in the ELSO database from 2010 to 2017 found that 40.2% experienced one or more bleeding or thrombotic event, with circuit thrombosis the most common, accounting for 54.9% of events. In 11,984 VA ECMO patients from the same database there were 8,457 events, of which 62.1% were bleeding. (Levy JH, Staudinger T, Steiner ME. Intensive Care Med 2022)
Suggested monitoring targets from the same source: ACT 180–200 seconds; aPTT 40–50 seconds initially, then titrated to 60–80 seconds depending on bleeding and thrombosis risk; anti-Xa 0.3–0.7 IU/mL. Unfractionated heparin remains the mainstay because it is titratable, reversible, and short-acting (about one hour).
The authors then say the thing that matters most: "There is no uniform ECMO anticoagulation practice as there is no evidence-based consensus guiding anticoagulation agents, monitoring, therapeutic targets, optimal levels, and management of complications and outcomes."
Certainty: expert practice. These are suggested ranges, not validated targets. No anti-Xa concentration has been shown to improve outcomes.
Pitfall — the tests disagree with each other, and that is normal
ACT correlates poorly with anti-Xa and aPTT; aPTT and anti-Xa are frequently discordant for heparin monitoring. ACT is prolonged by hypothermia, platelet number and function, and coagulation factor levels; anti-Xa is influenced by antithrombin deficiency, hyperlipidaemia, and haemolysis with raised plasma haemoglobin or bilirubin.
The practical consequence for the first day: decide which test your unit is titrating to, and record it. Chasing whichever number looks worst produces oscillating heparin doses and no better haemostasis. Antithrombin falls during ECMO, but no data support improved outcomes from supplementation (Levy 2022). Chapters 40 to 43 develop all of this.
The First-Hours Checklist
When | Check | What would change management |
0–15 min | Flow achieved and stable; no chatter; circuit intact and de-aired; cannula depths recorded at the skin; blood pressure and rhythm; three-point blood gas | Chatter → drainage problem, not a fluid order. Sudden hypotension → echo for tamponade or bleeding |
15–60 min | Ventilator turned down to lung-rest settings; sedation adequate; vasopressor titrated; heat exchanger set; anticoagulation infusion running; cannula sites inspected | If the ventilator has not come down, the reason for cannulating has not yet been delivered |
1 h | Repeat gas — confirm PaCO₂ is falling gradually; SaO₂ against the four determinants (Chapter 6); haemoglobin and platelets after prime dilution; lactate | PaCO₂ falling too fast → reduce sweep. SaO₂ inadequate at maximum sustainable flow → Chapter 6 titration tree |
4 h | Echo: cannula position, RV size and function, no pericardial collection. Coagulation to the unit's chosen test. Cannula site bleeding. Temperature | New RV dilatation → reassess whether RV failure was primary (Chapters 2 and 3). Site bleeding → review anticoagulation before escalating it |
24 h | Driving pressure and mechanical power documented; sedation and NMBA reviewed for reduction; fluid balance and a de-resuscitation plan; the destination restated (Chapters 1 and 4) | Driving pressure still high → this is the variable the early evidence associates with mortality. NMBA still running without a reason → stop it |
Clinical Pearl — write the destination on day zero
Chapter 1 made the bridge taxonomy a selection instrument and Chapter 4 made it a gate. The first-day note is where it becomes a fact: recovery, decision, or transplant — with, for a bridge to decision, an explicit review date. A run whose destination is never written down tends never to acquire one.
What Goes Wrong Early
Problem | Mechanism | Recognition | Action |
Abrupt CO₂ correction | Sweep set too high at initiation | Large fall in PaCO₂ on the first gas | Reduce sweep; accept a slow correction over 4–8 h |
Hypotension | Prime dilution, sedation, vasoplegia, unmasked hypovolaemia — or tamponade | Timing relative to initiation; echo | Name the cause. Vasopressor for vasoplegia; echo before assuming volume |
Dilutional anaemia and thrombocytopenia | 500–700 mL prime | Predictable fall on the 1-hour bloods | Anticipate; transfuse on the delivery reasoning of Chapters 1 and 2, not reflexively |
Chatter / erratic flow | Vena caval collapse around the drainage cannula | Movement of the venous limb; flow instability | Reduce pump speed first; consider position and volume state — Chapters 3, 5, 6 |
Cannulation-site bleeding | Anticoagulation on a fresh site | Ooze or expanding haematoma | Compression; review the anticoagulation target before increasing it — Chapter 36 |
Cannula migration | Movement, transfer, inadequate securing | Falling flow, rising recirculation, changed depth at the skin | Echo; re-secure; Chapter 34 |
Air in the circuit | Entrainment at a connection or access site | Visible air; alarm | Circuit emergency — Chapter 35 |
Hypothermia | Large extracorporeal heat-exchange surface | Falling core temperature; shivering | Set the heat exchanger; suppress shivering — it is a VO₂ load (Chapter 2) |
Persistent hypoxaemia | The four determinants of Chapter 6 | SaO₂ below target at maximum sustainable flow | Work the Chapter 6 tree, then Chapter 9 |
Hypotension in the First Hour
The final box is the point of the diagram. Because VV ECMO drains and returns the same volume to the same compartment, the circuit is not a haemodynamic explanation. Anything that looks like one is something else.
Clinical Pearls
- Turn the ventilator down. If it has not come down within the first hour, the therapy has been started but not delivered.
- If gas exchange deteriorates, increase extracorporeal support rather than ventilator settings. The ELSO sentence that settles most first-day arguments.
- Start the sweep low and correct the PaCO₂ over hours.
- Respiratory rate is a quarter of driving pressure, in mechanical-power terms. Four breaths per minute ≈ 1 cmH₂O.
- VV ECMO does not cause hypotension. Name the actual cause before treating it.
- Anticipate the prime. Haemoglobin and platelets will fall; that is arithmetic, not a complication.
- Pick one anticoagulation test and titrate to it. The tests disagree, and that is expected.
- Write the destination in the day-zero note.
Pitfalls
- Leaving the ventilator where it was because the gases look acceptable.
- Fixing the first blood gas by raising the sweep.
- Giving fluid for every episode of chatter, then spending a week de-resuscitating.
- Escalating the ventilator to treat a falling saturation on ECMO.
- Continuing neuromuscular blockade past the point where it has a purpose.
- Chasing whichever coagulation test looks worst, producing oscillating heparin doses.
- Treating the driving pressure as a day-five problem when the evidence points at the first three days.
Controversies
Controversy — how low should the ventilator go?
Clinical question: Is there a floor below which reducing ventilation on ECMO stops helping, or becomes harmful?
Evidence supporting going as low as possible: Driving pressure in the first three days was the sole independent mortality risk factor in pooled patient-level data. Animal data suggest the relationship between ventilation intensity and lung injury is linear with no safe threshold, which implies minimising mechanical power rather than reaching a target. Contemporary centres declare lung-rest goals of tidal volume under 5 mL/kg, driving pressure under 15 and often ≤ 10 cmH₂O, and rate under 10/min.
Evidence supporting caution: Higher PEEP in the early days was associated with reduced mortality, so "less ventilation" cannot mean less PEEP. A misaligned individualised PEEP strategy has been shown to increase mortality in ARDS, so personalisation is not automatically safe. And in the largest prospective cohort, day-two respiratory parameters showed no association with mortality — the association is not consistent across datasets. Complete lung rest also risks derecruitment and loss of the native lung contribution that Chapter 6 showed reduces required circuit flow.
Current consensus: minimise tidal volume, driving pressure and rate; maintain PEEP; personalise rather than apply a fixed recipe.
Practical approach: treat driving pressure and rate as the levers to minimise and PEEP as the one to defend.
Knowledge gap: no randomised trial of ventilation strategy during VV ECMO. Trials in the VA population are in progress.
Controversy — what anticoagulation target in the first 24 hours?
Clinical question: How intensely should a patient be anticoagulated immediately after cannulation?
Evidence supporting a lower early target: The cannulation sites are fresh, and site bleeding is the commonest early bleeding problem. Suggested practice starts at an aPTT of 40–50 seconds before titrating upward. Modern circuits tolerate reduced anticoagulation for periods (Chapter 4).
Evidence supporting a conventional target: Circuit thrombosis was the single most common hemocompatibility event in 7,579 VV ECMO patients, accounting for 54.9% of events. Under-anticoagulation risks the oxygenator.
Current consensus: there is none. The source review states plainly that no evidence-based consensus exists on agents, monitoring, targets or optimal levels.
Practical approach: follow the local protocol, record which test is being titrated to, and treat the early target as deliberately conservative while the sites are fresh.
Knowledge gap: essentially the whole field — a modality in use since the 1980s without established targets. Part VII.
Evidence Summary
Statement | Certainty | Basis |
Driving pressure during the first three days of ECMO is independently associated with mortality | Moderate | Pooled individual patient data, 545 patients (Serpa Neto 2016) |
Higher PEEP during the early days of ECMO is associated with reduced mortality | Low to moderate | Retrospective multicentre analysis (Schmidt 2015) |
ECMO initiation substantially reduces tidal volume, driving pressure, rate and mechanical power in practice | Moderate — descriptive | International prospective cohort, 350 patients (Schmidt 2019) |
Specific ventilator settings during ECMO improve survival | Not established | No randomised trial; day-two parameters showed no association with mortality in the largest cohort |
If gas exchange deteriorates, increase extracorporeal support rather than ventilator settings | Consensus | ELSO guidance as reported in Szuldrzynski 2024 |
Respiratory rate contributes to mechanical power at about 25% of the weight of driving pressure | Moderate — derived | Costa 2021 analysis across lung-protective ventilation populations |
Prone positioning during ECMO improves survival | Conflicting | Meta-analysis favourable (RR 1.31); pooled IPD neutral; PRONECMO randomised trial neutral |
Bleeding or thrombotic events affect roughly 40% of VV ECMO patients | Moderate | ELSO registry analysis, 7,579 patients 2010–2017 |
Suggested early anticoagulation targets (ACT 180–200 s; aPTT 40–50 s then 60–80 s; anti-Xa 0.3–0.7 IU/mL) | Expert practice | Levy 2022 — explicitly not evidence-based consensus |
Antithrombin supplementation improves outcomes on ECMO | Not established | No supporting data (Levy 2022) |
VV ECMO itself causes hypotension | Contradicted | Equal drainage and return to the venous compartment — Chapters 2 and 6 |
Key Takeaways
- The first three days carry the association with mortality that later days do not — this chapter exists because of that.
- Turning the ventilator down is the act that delivers the therapy. Doing it late wastes the cannulation.
- ELSO's governing sentence: if gas exchange deteriorates, increase extracorporeal support, not ventilator settings.
- Initiation in practice roughly halves tidal volume and cuts mechanical power by about three-quarters.
- Rate matters. Four breaths per minute is worth about 1 cmH₂O of driving pressure.
- PEEP is the setting to defend while everything else comes down.
- Start the sweep low; correct PaCO₂ over 4–8 hours.
- VV ECMO is haemodynamically neutral, so hypotension always has another explanation — find it.
- The prime will dilute haemoglobin and platelets. Expect it.
- Anticoagulation targets are suggestions, not evidence. Pick a test, record it, start conservatively while the sites are fresh.
- Neuromuscular blockade: widely used at initiation, withdrawn as soon as possible.
- Write the destination on day zero.
Key References
- Szuldrzynski K, Kowalewski M, Swol J. Mechanical ventilation during extracorporeal membrane oxygenation support — new trends and continuing challenges. Perfusion. 2024;39(1S):107S–114S. DOI: 10.1177/02676591241232270 — read in full from the project library; the source for the ELSO ventilator settings, the Schmidt 2019 initiation figures, the Serpa Neto driving-pressure finding, the Costa mechanical-power relationship, the proning evidence and the NMBA survey data
- Levy JH, Staudinger T, Steiner ME. How to manage anticoagulation during extracorporeal membrane oxygenation. Intensive Care Med. 2022. DOI: 10.1007/s00134-022-06723-z — read in full from the project library; the source for the ELSO registry event rates and the suggested monitoring targets
- Tonna JE, Abrams D, Brodie D, et al. Management of adult patients supported with venovenous ECMO: guideline from ELSO. ASAIO J. 2021;67(6):601–610. DOI: 10.1097/MAT.0000000000001432
- Shekar K, Buscher H, Brodie D. Protocol-driven daily optimisation of venovenous ECMO blood flows: an alternate paradigm? J Thorac Dis. 2020;12(11):6854–6860. DOI: 10.21037/jtd-20-1515
- Primary studies cited as reported in reference 1: Schmidt M, et al. Am J Respir Crit Care Med 2019 (initiation cohort, n=350); Serpa Neto A, et al. Intensive Care Med 2016 (pooled IPD, n=545); Schmidt M, et al. Crit Care Med 2015 (PEEP, three centres); Costa ELV, et al. Am J Respir Crit Care Med 2021 (mechanical power); Giani M, et al. Crit Care 2022 and Papazian L, et al. Intensive Care Med 2022 (proning); Schmidt M, et al. JAMA 2023 (PRONECMO); van Minnen O, et al. ASAIO J 2023 (survey)
- Cross-references: Chapter 1 (the destination), Chapter 2 (VO₂ as a treatable variable; DO₂:VO₂), Chapter 3 (prime volume; drainage limits), Chapter 5 (cannula depth and securing), Chapter 6 (all initiation physiology and the titration tree), Chapter 8 (daily management), Chapter 9 (persistent hypoxaemia), Chapter 10 (weaning), Chapters 36 and 40–43 (bleeding and anticoagulation), Chapter 53 (ventilation during ECMO in depth, including proning), Chapter 57 (sedation and delirium)
Citation verification status. Both principal sources were read in full from the project library during the 6 September 2026 cycle, and every figure, threshold and quoted phrase attributed to them is taken directly from those texts: Szuldrzynski 2024 (Perfusion 39(1S):107S–114S, DOI 10.1177/02676591241232270) and Levy 2022 (Intensive Care Med, DOI 10.1007/s00134-022-06723-z). The ELSO 2021 VV guideline and Shekar 2020 citations were verified in earlier cycles for Chapters 4 and 6.
Second-hand attribution, declared. The Schmidt 2019, Serpa Neto 2016, Schmidt 2015, Costa 2021, Giani 2022, Papazian 2022, PRONECMO and van Minnen 2023 figures are quoted as reported in Szuldrzynski 2024, not from the primary papers, which were not retrieved in this cycle. They are labelled that way throughout. Retrieve the primaries before quoting these numbers in a formal document — particularly the PRONECMO and Papazian results, which point in opposite directions.
Transcription note. The source review's sentence describing PRONECMO reads "no benefit of ECMO"; from the surrounding text and the trial's design this plainly means no benefit of prone positioning, and it is rendered that way here.
Not quoted: ventilator settings and oxygenation targets specific to VA ECMO, which appear in the same review but belong to Chapters 14 and 53.
Educational use only. This chapter does not replace institutional ECMO protocols, local policy, specialist consultation, current guidelines or patient-specific clinical judgement.