Cardiopulmonary Interactions During Shock and Resuscitation

During shock, the circulation has less reserve to tolerate the loading changes produced by breathing and ventilation. A respiratory intervention can improve arterial oxygenation while reducing cardiac output, or unload a failing left ventricle while preserving useful flow. Resuscitation changes those responses again. The task is to identify the limiting mechanism, test it when feasible, and reassess the combined effect on gas exchange, circulation, and tissue perfusion.

42.1 Shock Assessment Must Connect Pressure, Flow, and Cardiac Geometry

Arterial pressure supports organ perfusion, but pressure alone does not describe how much blood reaches the tissues. A circulation with low vascular resistance can generate inadequate pressure despite preserved or increased output. A failing ventricle can maintain an apparently acceptable pressure only while flow falls. Respiratory pressure changes can alter either state.

A useful simplified systemic relationship is:

Pressures are in mm Hg, output in L/min, and resistance in mm Hg·min/L. This describes a mean pressure–flow relationship, not the full pulsatile circulation. It cannot identify regional perfusion or establish whether a particular pressure is sufficient for an individual organ.

For illustration, output of 4 L/min through resistance of 15 mm Hg·min/L produces a mean gradient of 4 × 15 = 60 mm Hg. Output of 6 L/min through resistance of 10 produces the same 6 × 10 = 60 mm Hg gradient. Identical gradients can accompany different flows and different reasons for shock.

Cardiac geometry adds the missing anatomical context. A small left ventricle may reflect reduced circulating supply, but it may also lie downstream from a dilated right ventricle that cannot eject. A raised central venous pressure can reflect external pressure, right ventricular failure, pericardial constraint, or combinations. Timing identifies whether these findings appeared with induction, an inflation, fluid administration, or disease progression.

The 2025 ESICM shock-monitoring guideline supports echocardiography as the initial imaging approach to identify shock type. It also favors dynamic over static preload variables when applicable and calls for cardiac output or stroke volume monitoring when initial therapy fails. These recommendations support a mechanistic assessment; they do not make one measurement a treatment target for every shock phenotype. 1

42.2 Reduced Venous Supply and Vasodilation Create Different Forms of Underfilling

In hypovolemic shock, loss of blood or fluid can reduce the stressed vascular volume that generates pressure for venous return. The heart receives less effective filling even if its contractile function is initially preserved. Positive intrathoracic pressure can then impose a substantial additional obstacle because the upstream pressure driving return is already limited.

Distributive shock can produce underfilling through a different route. Increased venous capacitance allows more blood to reside without generating the same upstream pressure. Arterial vasodilation lowers pressure for a given flow, while myocardial dysfunction and microvascular abnormalities may coexist. A patient can consequently have low arterial pressure with high output, low output, or a changing mixture.

These mechanisms explain why loss of spontaneous inspiratory assistance can matter. Before induction, negative pleural pressure may help draw venous blood toward the chest. After induction and controlled inflation, right atrial pressure may rise while venous tone and upstream pressure change. The effect depends on both ends of the return pathway, not on the central venous pressure alone.

Fessler and colleagues studied venous pressure–flow curves in a closed-chest canine bypass preparation. Positive end-expiratory pressure altered upstream pressure, venous conductance, and the critical downstream pressure limiting flow. This experiment shows why reduced return cannot always be represented as an isolated rise in right atrial pressure with all other variables held constant. 2

An underfilled appearance therefore requires an explanation. Ongoing hemorrhage, venous dilation, excessive respiratory loading, impaired right ventricular ejection, and increased abdominal pressure can produce overlapping findings. Restoring missing volume, altering vascular tone, or reducing an excessive mechanical impediment acts on different parts of that chain.

The physiological interpretation should follow measured flow and perfusion as the suspected mechanism changes. A rise in arterial pressure without improved flow may be useful in some settings, but it does not establish correction of the entire oxygen transport problem. Likewise, increasing filling does not guarantee that the ventricle can convert it into forward output.

42.3 Left Ventricular Failure Can Benefit from Respiratory Unloading

A distressed patient with left ventricular failure may generate large negative inspiratory pressures. These efforts increase the pressure difference the ventricular wall must sustain during ejection. They also increase respiratory muscle oxygen demand and can favor pulmonary fluid filtration. The work of breathing then becomes part of the circulatory problem.

Positive airway support can reduce inspiratory effort and make pressure around the heart less negative. This can reduce the transmural component of left ventricular ejection load. If the ventricle has limited contractile reserve and already operates at high filling pressures, that reduction may improve the balance between filling and ejection.

Naughton and colleagues compared physiological responses to continuous positive airway pressure in patients with congestive heart failure and healthy participants. In the heart failure group, support reduced inspiratory effort and estimated left ventricular transmural systolic pressure without reducing cardiac index. The different response in healthy participants illustrates that the benefit depends on the starting circulatory state. This was a physiological study, not a trial of shock survival. 3

The potential benefit has limits. Excessive reduction in venous return can lower left ventricular filling after pulmonary transit. Increased pulmonary vascular load can impair the right ventricle upstream. Hypotension can compromise coronary perfusion, and relief of dyspnea can coexist with declining forward flow. A failing left ventricle does not make all positive pressure hemodynamically favorable.

Ejection fraction also cannot settle the issue by itself. It changes with loading conditions and describes a fraction of the end-diastolic volume. Preserved ejection fraction can coexist with abnormal relaxation and high filling pressure. A small stroke volume can occur despite a seemingly vigorous fraction when the chamber is poorly filled.

The relevant response is therefore combined: reduced effort, improved gas exchange, acceptable ventricular filling, preserved or improved stroke volume, and improving tissue perfusion. These observations explain why respiratory support can form part of circulatory support while still requiring close reassessment of the right heart and systemic pressure.

42.4 Right Ventricular Outflow Load and Pericardial Constraint Can Mimic Volume Depletion

A right ventricle facing acute outflow obstruction or high pulmonary vascular load may dilate while its forward stroke volume falls. Less blood then reaches the left heart. The left ventricle can look small even though the right heart and systemic veins are congested.

Pulmonary embolic obstruction, hypoxic vascular constriction, hypercapnic acidemia, and overdistended lung can contribute different components of that load. Their coexistence matters. A ventilation change that improves alveolar oxygen may reduce chemical vasoconstrictor stimulation while increasing mechanical vascular compression elsewhere.

As the right ventricle dilates, septal displacement can restrict left ventricular filling. Pericardial constraint adds another interaction because the chambers share limited surrounding space. Higher venous pressure may then produce little additional useful left ventricular preload. Adding volume can increase right ventricular wall stress and regurgitation rather than restore forward flow.

Jardin and colleagues combined pressure measurements and echocardiography during PEEP changes in patients with ARDS. Their observations linked falling output with altered septal geometry and reduced left ventricular dimensions. At the greatest pressures studied, restoring transmural filling pressure with volume did not restore output. This supports ventricular interaction as one possible cause of apparent left-heart underfilling, within the limits of that small historical experiment. 4

Pericardial tamponade is a separate obstructive mechanism. The clinically important pressure is the chamber pressure relative to the surrounding constraint, and positive pressure can further reduce the margin for filling. However, a raised central venous pressure and respiratory variation do not diagnose tamponade without the relevant clinical and imaging findings.

Assessment should therefore compare the two ventricles, septal behavior, venous congestion, forward flow, and the respiratory pressure pattern. A small left ventricle downstream from right ventricular failure requires a different interpretation from two small ventricles with limited venous supply. Neither pattern should be converted into an automatic fluid instruction without evaluating the mechanism and urgency.

42.5 The Intubation Transition Can Expose Limited Cardiovascular Reserve

Intubation changes several circulatory supports at once. Induction may reduce sympathetic tone, increase venous capacitance, and depress myocardial performance. Loss of inspiratory muscle activity removes negative-pressure assistance to venous return. Positive-pressure inflation then changes filling and pulmonary vascular load.

A patient who maintained output through intense effort and adrenergic stimulation may deteriorate rapidly when these compensations disappear. Another patient may improve because unloading reduces left ventricular stress and respiratory oxygen demand. The direction cannot be inferred from the airway procedure alone.

The international INTUBE cohort documented frequent major peri-intubation events, especially cardiovascular instability, among critically ill adults. Because the study was observational, it describes risk during the combined procedure rather than isolating the causal effect of positive pressure, an induction agent, or a specific clinical practice. 5

A plausible reduction in venous return does not establish that all patients benefit from preventive fluid. PREPARE II compared a standardized crystalloid bolus with no bolus during intubation with planned positive-pressure ventilation. Among 1,065 analyzed patients, it did not demonstrate a significant reduction in cardiovascular collapse. Patients for whom fluid was already required or contraindicated were excluded. The result therefore concerns the tested preventive strategy and does not remove separate indications for volume replacement. 6

Before-and-after observations should capture the actual transition: pressure, pulse or measured flow, effort, ventricular findings when available, and the timing of medication and inflation. An increase in central venous pressure with falling pressure and output may reflect transmitted pressure, impaired right ventricular ejection, or both.

Abrupt deterioration also requires assessment for mechanical complications, including tension pneumothorax, airway obstruction, and severe dynamic hyperinflation. These can require urgent action rather than an extended preload-testing sequence. Physiological reasoning should speed recognition of the relevant failure, not delay it while seeking one definitive monitor value.

42.6 Improved Gas Exchange Can Coexist with Lower Systemic Oxygen Delivery

Resuscitation must consider the oxygen content entering the systemic circulation and the flow carrying it. A higher PaO₂ or saturation addresses only part of that balance. In shock, a reduction in output can offset an improvement in pulmonary oxygen transfer.

Consider an illustrative lung with two regions. One returns blood containing 12 mL O₂/dL and receives 40% of pulmonary flow. The other returns blood containing 18 mL O₂/dL and receives 60%. These contents are assumed fixed only to isolate the effect of redistribution.

Initial content is 0.40 × 12 + 0.60 × 18 = 15.6 mL O₂/dL. If the poorer region receives 20% and the better region 80%, content becomes 0.20 × 12 + 0.80 × 18 = 16.8 mL O₂/dL. Blood mixes by oxygen content; oxygen tensions cannot be averaged in the same way because of the nonlinear dissociation relationship.

At output of 4.5 L/min, initial systemic oxygen delivery is 15.6 × 4.5 × 10 = 702 mL O₂/min. The factor of 10 converts liters of blood to deciliters. At unchanged output, the improved mixture provides 16.8 × 4.5 × 10 = 756 mL O₂/min. If output instead falls to 3.5 L/min, delivery becomes 16.8 × 3.5 × 10 = 588 mL O₂/min.

This is an arithmetic possibility, not a prediction that better matching necessarily reduces output. Actual regional contents also change with mixed venous oxygen, flow, recruitment, and transit time.

Mekontso Dessap and colleagues demonstrated a related clinical divergence in a small severe-ARDS study: ventilation strategies improved oxygenation and recruitment while hypercapnic acidemia, right ventricular dilation, and reduced cardiac index developed. Several variables changed together, so the study cannot isolate one cause. It does show why improved oxygenation must be interpreted with the circulation. 7

42.7 Fluid Responsiveness, Fluid Need, and Fluid Tolerance Are Separate Questions

Fluid responsiveness asks whether increasing preload is likely to increase stroke volume or output. Fluid need asks whether increasing that flow is likely to address an important perfusion deficit. Fluid tolerance asks what adverse effects may accompany the additional intravascular volume and pressure.

A patient may remain responsive despite adequate perfusion. Additional fluid can then increase output without providing a meaningful benefit. Another patient may have a strong potential flow response but limited pulmonary or venous tolerance. The expected increase in flow must be weighed against the risk of edema, congestion, ventricular dilation, and impaired downstream organ perfusion.

Respiratory disease sharpens that tradeoff. Increasing pulmonary vascular pressures can worsen lung water when the barrier is injured. Greater right ventricular filling can aggravate septal interaction if the ventricle cannot eject. Higher abdominal pressure can further impede return and organ venous drainage. A favorable stroke volume response does not establish that these consequences are acceptable.

The 2026 Surviving Sepsis Campaign conditionally favors dynamic measures over examination or static measures alone for fluid guidance, with low-certainty evidence. This supports testing likely flow benefit while retaining clinical assessment. It does not turn responsiveness into an obligation to give fluid, nor establish that all forms of dynamic monitoring improve outcomes equally. 8

Tolerance should be reassessed during and after an intervention. A rising venous pressure with little sustained flow increase suggests a different balance from improved flow with stable chamber geometry and improving perfusion. Changes in gas exchange, lung findings, right ventricular size, and systemic venous congestion provide complementary evidence, each with its measurement limits.

No single reassuring variable proves tolerance. Preserved arterial saturation does not exclude additional lung water, and a noncongested venous examination does not exclude pulmonary edema. The assessment must match the organ at risk and the mechanism by which the intervention could cause harm. The same patient can have different responsiveness, need, and tolerance later in the resuscitation.

42.8 Respiratory Variation Can Reflect Preload Sensitivity or Right Ventricular Stress

Controlled inflation creates a recurring cardiovascular challenge. When filling changes produce substantial stroke volume changes, respiratory pulse pressure or stroke volume variation can help predict preload responsiveness. The inference requires an interpretable challenge and a suitable measurement.

Spontaneous effort changes the direction and amplitude of pressure swings. Arrhythmia creates beat-to-beat variation unrelated to the ventilator. Low tidal volume or limited transmission to the pleural space can produce a weak signal despite preload responsiveness. Changing respiratory rate or heart rate alters how many ejections occur during each respiratory phase.

A large signal can also have a different explanation. If inflation increases pulmonary vascular load, a vulnerable right ventricle may eject less during inspiration. The delayed reduction in left ventricular filling can generate arterial variation. Additional volume may then worsen right ventricular dilation rather than correct the limiting outflow problem.

Mahjoub and colleagues evaluated mechanically ventilated patients selected for high pulse pressure variation. Some did not increase stroke volume after volume expansion, and right ventricular tissue Doppler findings helped identify these false-positive patterns. The study supports evaluating right ventricular function before interpreting a large variation as a fluid indication; it does not establish one Doppler cutoff as a universal diagnostic rule. 9

De Backer and colleagues also demonstrated tidal-volume dependence of the predictive performance of pulse pressure variation. Their findings explain why a small signal during protective ventilation may be inconclusive. They do not justify increasing tidal volume routinely or transferring a threshold between different respiratory conditions. 10

The practical distinction is between an invalid or insufficient test and a true absence of preload responsiveness. Both may display a low number. A high number also requires a mechanism. When rhythm, effort, tidal challenge, or right ventricular performance makes interpretation uncertain, the clinician should use another suitable assessment rather than force the signal into a universal fluid rule.

42.9 Reversible Tests Require a Valid Maneuver and Direct Assessment of Flow

Passive leg raising briefly redistributes venous blood toward the thorax without adding external fluid. A rise in stroke volume or output suggests that the circulation can convert the temporary increase in preload into additional flow. The effect should be captured promptly and should reverse when the patient returns to the original position.

Monnet and colleagues studied 71 patients considered for volume expansion, including patients with spontaneous activity or arrhythmia. Changes in Doppler aortic flow during leg raising predicted the subsequent flow response better than relying on the associated arterial pulse pressure change. This supports measuring flow during the maneuver rather than using blood pressure alone. 11

The test requires stable concurrent treatment and a measurement able to capture the transient response. Pain, coughing, active movement, changing support, or a moving Doppler sampling position can alter the apparent result. Raised abdominal pressure can prevent the intended preload transfer. Beurton and colleagues found false-negative leg-raising responses in patients with intra-abdominal hypertension; a negative result is therefore not universally conclusive. 12

End-expiratory occlusion tests a different mechanism. Briefly withholding the next insufflation removes recurring inspiratory impediments to filling while maintaining the end-expiratory state. If output rises, the response can indicate preload sensitivity. Monnet and colleagues demonstrated predictive value in 34 mechanically ventilated patients with shock using a short occlusion and pulse-contour flow assessment. 13

An interpretable occlusion requires that the patient tolerate the pause without disruptive respiratory effort. The monitor must resolve a small, brief change; a heavily averaged output display or an imprecise repeated echocardiographic measurement may miss it. Ventilator disconnection is not equivalent because it changes the end-expiratory pressure and lung volume.

Both tests examine the current circulation under their conditions. They do not establish the duration of a subsequent fluid response or its safety. Their most useful role is to reduce uncertainty about likely flow benefit while the clinician separately evaluates perfusion need and tolerance.

42.10 Respiratory Interventions Can Unload the Right Ventricle When the Mechanism Fits

The respiratory goal during shock extends beyond raising arterial oxygen. Recruitment may reduce hypoxic vascular constriction and restore perfused exchange surfaces. Excessive inflation may compress vessels and increase right ventricular load. Changes in carbon dioxide and pH modify vascular tone while changes in mean pleural pressure modify venous return.

The net effect depends on recruitability, regional distension, blood volume distribution, and ventricular reserve. There is no airway pressure that identifies the favorable balance in every patient. Monitoring should determine whether the chosen respiratory strategy improves gas exchange with acceptable cardiac geometry, flow, and perfusion.

Prone positioning can alter several components together. More even lung inflation and improved ventilation can reduce regional overdistension and chemical vasoconstrictor stimulation. Position also changes abdominal and venous conditions, so an individual hemodynamic response cannot be assumed from oxygenation alone.

Vieillard-Baron and colleagues evaluated 42 patients with severe ARDS before and after prolonged prone ventilation. In those with acute cor pulmonale, right ventricular enlargement and septal abnormalities improved alongside changes in airway pressure and carbon dioxide. This before-and-after physiological study supports possible right ventricular unloading; it cannot isolate one mediator or establish a mortality effect from unloading itself. 14

PROSEVA tested early prolonged prone sessions against supine care within a protective ventilation strategy. Its analysis included 466 randomized patients with severe ARDS and showed lower mortality with proning. The trial enrolled intubated patients with PaO₂/FiO₂ below 150 mm Hg while FiO₂ was at least 0.6 and PEEP at least 5 cm H₂O, confirmed after stabilization. The benefit concerns that severe-ARDS strategy and population; it is not evidence that proning treats every cause of shock or that right ventricular unloading alone explains survival. 15

For a patient with shock and lung injury, respiratory optimization should therefore test a reasoned mechanism while preserving the broader treatment strategy. Gas exchange, expiratory emptying, right ventricular response, systemic pressure, and flow all contribute to the assessment. Improvement in one should not conceal deterioration in another.

42.11 The Limiting Mechanism Changes During Resuscitation

A resuscitation maneuver changes the conditions under which the next maneuver will act. Fluid can raise upstream filling pressure initially, then accumulate as interstitial edema. Restoring vascular tone can change venous recruitment and arterial ejection load together. Respiratory unloading can reduce oxygen demand while altering preload and afterload.

The dose and timing of each intervention therefore matter even when its mechanism is reasonable. A favorable immediate response may fade as fluid redistributes or as vascular tone changes. A previously responsive heart can become less responsive after filling increases. Improved lung recruitment can later coexist with overdistension as lung mechanics evolve.

CLOVERS compared restrictive and liberal early resuscitation strategies in 1,563 patients with sepsis-induced hypotension after initial fluid treatment. The restrictive strategy used less fluid and earlier vasopressors. The trial did not demonstrate a difference in its primary mortality outcome before discharge home by day 90. Because both fluid and vasopressor strategies differed, the result cannot isolate the effect of one fluid amount or establish equivalence for every phenotype. Fluid overload and severe nonseptic depletion were excluded. 16

Perfusion measurements help determine whether macrocirculatory changes are useful. Capillary refill, skin findings, mentation, urine output, and lactate trends describe different parts of the response. None uniquely measures cardiac output or proves one cause of hypoperfusion. Lactate can remain elevated because of production and clearance processes other than ongoing oxygen supply limitation.

The 2026 Surviving Sepsis Campaign conditionally supports serial lactate guidance and capillary refill as an adjunct to other perfusion measures, with low-certainty evidence. Its remarks favor individualized further fluid assessment rather than continuing fluid solely until lactate normalizes. These are combined clinical assessments, not substitutes for identifying mechanical or cardiac failure. 8

Ongoing reassessment should therefore ask whether the current limitation remains venous supply, vascular tone, ventricular ejection, external constraint, oxygen carriage, or regional distribution. Mixed dysfunction is common, and the dominant component can change within the same resuscitation.

42.12 Compare Predicted Benefit with the Sustained Measured Response

Consider an illustrative patient with persistent poor perfusion whose measured output is 4 L/min and arterial oxygen content is 15.5 mL O₂/dL. A technically valid leg-raising test increases output to 4.6 L/min, with return to baseline afterward. The relative increase is (4.6 − 4.0) ÷ 4.0 × 100 = 15%. This describes the observed response, not a universal treatment threshold.

Suppose fluid is subsequently administered after the clinical assessment, and sustained output becomes 4.3 L/min. Content falls to 14.8 mL O₂/dL because the measured hemoglobin concentration and oxygen carriage have changed. Right atrial pressure rises, and reassessment shows greater right ventricular dilation with new septal flattening.

Systemic oxygen delivery is:

CaO₂ is arterial content in mL O₂/dL, output is in L/min, and DO₂ is in mL O₂/min. Initial delivery is 4.0 × 15.5 × 10 = 620 mL O₂/min. The sustained value is 4.3 × 14.8 × 10 = 636.4 mL O₂/min, approximately 636. Output increased by 0.3 ÷ 4.0 × 100 = 7.5%, but the delivery increase is only about 2.6% because content fell.

The test and intervention are not identical events. The maneuver created a transient preload change without hemodilution. Subsequent vascular conditions, respiratory loading, and redistribution may differ. A smaller sustained response does not automatically prove that the test was invalid, although measurement error must be considered.

More importantly, the new right-heart findings change the next decision. A modest increase in delivery must be interpreted with congestion, ventricular interaction, and actual perfusion. Continuing the same intervention solely because the original test was positive ignores the new state.

A useful resuscitation record therefore pairs the suspected mechanism with baseline measurements, the intended physiological change, the observed immediate response, and the sustained response with adverse effects. This makes each intervention a testable clinical action rather than an accumulation of treatments attached to an unchanged shock label.

References

1. Monnet X, Messina A, Greco M, et al. ESICM guidelines on circulatory shock and hemodynamic monitoring 2025. Intensive Care Med. 2025;51:1971–2012.

2. Fessler HE, Brower RG, Wise RA, Permutt S. Effects of positive end-expiratory pressure on the canine venous return curve. Am Rev Respir Dis. 1992;146:4–10.

3. Naughton MT, Rahman MA, Hara K, Floras JS, Bradley TD. Effect of continuous positive airway pressure on intrathoracic and left ventricular transmural pressures in patients with congestive heart failure. Circulation. 1995;91:1725–1731.

4. Jardin F, Farcot JC, Boisante L, Curien N, Margairaz A, Bourdarias JP. Influence of positive end-expiratory pressure on left ventricular performance. N Engl J Med. 1981;304:387–392.

5. Russotto V, Myatra SN, Laffey JG, et al. Intubation practices and adverse peri-intubation events in critically ill patients from 29 countries. JAMA. 2021;325:1164–1172.

6. Russell DW, Casey JD, Gibbs KW, et al. Effect of fluid bolus administration on cardiovascular collapse among critically ill patients undergoing tracheal intubation: a randomized clinical trial. JAMA. 2022;328:270–279.

7. Mekontso Dessap A, Charron C, Devaquet J, et al. Impact of acute hypercapnia and augmented positive end-expiratory pressure on right ventricle function in severe acute respiratory distress syndrome. Intensive Care Med. 2009;35:1850–1858.

8. Society of Critical Care Medicine and European Society of Intensive Care Medicine. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock 2026. Official guideline recommendations. 2026.

9. Mahjoub Y, Pila C, Friggeri A, et al. Assessing fluid responsiveness in critically ill patients: false-positive pulse pressure variation is detected by Doppler echocardiographic evaluation of the right ventricle. Crit Care Med. 2009;37:2570–2575.

10. De Backer D, Heenen S, Piagnerelli M, Koch M, Vincent JL. Pulse pressure variations to predict fluid responsiveness: influence of tidal volume. Intensive Care Med. 2005;31:517–523.

11. Monnet X, Rienzo M, Osman D, et al. Passive leg raising predicts fluid responsiveness in the critically ill. Crit Care Med. 2006;34:1402–1407.

12. Beurton A, Teboul JL, Girotto V, et al. Intra-abdominal hypertension is responsible for false negatives to the passive leg raising test. Crit Care Med. 2019;47:e639–e647.

13. Monnet X, Osman D, Ridel C, Lamia B, Richard C, Teboul JL. Predicting volume responsiveness by using the end-expiratory occlusion in mechanically ventilated intensive care unit patients. Crit Care Med. 2009;37:951–956.

14. Vieillard-Baron A, Charron C, Caille V, Belliard G, Page B, Jardin F. Prone positioning unloads the right ventricle in severe ARDS. Chest. 2007;132:1440–1446.

15. Guérin C, Reignier J, Richard JC, et al. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med. 2013;368:2159–2168.

16. Shapiro NI, Douglas IS, Brower RG, et al.; NHLBI PETAL Clinical Trials Network. Early restrictive or liberal fluid management for sepsis-induced hypotension. N Engl J Med. 2023;388:499–510.