SSC 2026 — Part 4: Respiratory Support

This section covers respiratory support in adults with sepsis, septic shock, and sepsis-associated acute respiratory distress syndrome (ARDS), including oxygenation monitoring, oxygen targets, noninvasive respiratory support strategies (high flow nasal cannula, noninvasive positive pressure ventilation), awake proning, lung-protective mechanical ventilation, PEEP strategy, prone positioning, neuromuscular blockade, and venovenous extracorporeal membrane oxygenation (VV-ECMO).

Respiratory failure is one of the most common organ dysfunctions in sepsis. Optimizing respiratory support — from noninvasive strategies through invasive mechanical ventilation — is central to improving outcomes. The 2026 guidelines introduce several new recommendations on oxygenation monitoring, oxygen targets, noninvasive respiratory support, and awake proning, while carrying over key principles of lung-protective ventilation from 2021.

1. Monitoring of Hypoxemia

Recommendation 65 — Oxygenation Monitoring

For adults with sepsis or septic shock and acute hypoxemic respiratory failure, we suggest measuring oxygenation by either pulse oximeter (SpO2) or arterial blood gas (SaO2) in conjunction with physical examination and clinical acumen.

Conditional recommendation, very low certainty evidence

Change from 2021: New recommendation.

Rationale: Arterial blood gas (ABG) analysis measuring PaO2 and SaO2 remains the gold standard for assessing oxygenation. However, pulse oximetry (SpO2) is widely available, noninvasive, and continuous, making it a practical and essential monitoring tool in all clinical settings.

The panel evaluated 45 observational studies examining the correlation between SpO2/FiO2 and PaO2/FiO2 ratios. Spearman correlation coefficients ranged from 0.5 to 0.8, supporting the use of SpO2/FiO2 as a surrogate for PaO2/FiO2 in clinical practice. This evidence underpins the new global ARDS consensus definition, which now formally allows SpO2-based measurements for ARDS diagnosis — a significant step toward equitable diagnosis in resource-limited settings where ABG analysis may not be readily available.¹

Important limitations of pulse oximetry:

The finding that patients with darker skin pigmentation experience nearly three-fold higher rates of occult hypoxemia (SpO2 overestimating true saturation) is a critical equity concern. Clinicians should maintain a lower threshold for ABG confirmation in these patients and when pulse oximetry values appear discordant with the clinical picture.

Remark: The SpO2/FiO2 ratio can substitute for PaO2/FiO2 in clinical assessment and ARDS diagnosis. ABG remains the gold standard when available. SpO2 is less accurate in patients in shock, those with darker skin tones, and at oxygen saturations below 90% or above 97%.

2. Oxygen Targets

Recommendation 66 — Oxygen Target Strategy

For adults with sepsis or septic shock and acute hypoxemic respiratory failure, we suggest titrating FiO2 to target either higher, more liberal oxygen levels or lower, conservative oxygen levels depending on patient factors and resource limitations.

Conditional recommendation, low certainty evidence

Change from 2021: New recommendation.

Rationale: The optimal oxygen target in critically ill patients with sepsis has been debated extensively. The panel systematically reviewed 10 RCTs enrolling a total of 24,022 patients comparing lower (conservative) versus higher (liberal) oxygenation targets.

Typical targets used in the reviewed trials:

Key findings from meta-analysis:

The primary outcome of short-term mortality showed little to no difference between lower and higher oxygen targets (RR 1.02; 95% CI, 0.99–1.05, low certainty). Neither strategy demonstrated a clear mortality advantage across the pooled data.

The panel acknowledged that the effect of oxygenation targets may not be uniform across all patients — some patients may benefit from lower targets (e.g., those at risk of oxygen toxicity, absorption atelectasis, or ventilator-induced lung injury from high FiO2), while others may benefit from higher targets (e.g., those with impaired oxygen delivery or at risk of tissue hypoxia).

Panel practice (“In our practice”): Panelists reported targeting SpO2 between 90% (IQR 90–92%) on the lower end and 96% (IQR 94–98%) on the upper end.

Remark: It is reasonable to target the lower end of the liberal range (SpO2 96–97%) or intermediate targets (SpO2 94–95%) even when resources are not limited. In resource-limited settings where continuous monitoring and ABG are less available, a wider acceptable range allows practical flexibility while avoiding the extremes of hypoxemia and hyperoxemia.

3. Noninvasive Respiratory Support

Recommendation 67 — HFNC Over Conventional Oxygen Therapy

For adults with sepsis or septic shock and acute hypoxemic respiratory failure, we suggest using high flow nasal cannula (HFNC) over conventional oxygen therapy.

Conditional recommendation, very low certainty evidence

Change from 2021: New recommendation.

Rationale: This recommendation pertains to patients with significant hypoxemia, defined as PaO2/FiO2 < 200 mm Hg or SpO2/FiO2 < 235. HFNC delivers heated, humidified oxygen at flow rates up to 60 L/min with titratable FiO2 up to 0.8–0.9. Compared with conventional oxygen delivery (nasal prongs, simple face mask, or non-rebreather mask), HFNC provides several physiological benefits:

• Improved matching of patient inspiratory flow demand
• Washout of nasopharyngeal dead space
• A modest degree of continuous positive airway pressure (2–5 cm H2O at high flow rates)
• Improved mucociliary clearance through heated humidification
• Enhanced patient comfort and tolerance

Conventional oxygen therapy via nasal prongs or face masks typically delivers a maximum FiO2 of 0.4–0.6, which may be insufficient for patients with moderate-to-severe hypoxemia. Non-rebreather masks can deliver higher FiO2 but at fixed flow rates that may not match patient demand.

Remark: HFNC is suggested over conventional oxygen therapy for patients with PaO2/FiO2 < 200 or SpO2/FiO2 < 235. HFNC can deliver FiO2 of 0.8–0.9 at flow rates up to 60 L/min.

Recommendation 68 — HFNC Over NIPPV as Initial Therapy

For adults with sepsis or septic shock and acute hypoxemic respiratory failure, we suggest using HFNC as initial therapy over noninvasive positive pressure ventilation (NIPPV).

Conditional recommendation, low certainty evidence

Change from 2021: New recommendation.

Rationale: The panel reviewed 11 RCTs enrolling 3,546 patients comparing HFNC with NIPPV as initial noninvasive respiratory support for acute hypoxemic respiratory failure.

Key findings from meta-analysis:

While point estimates consistently favored HFNC, confidence intervals included no difference for mortality and intubation. The possible reduction in intubation (RR 0.88) is clinically meaningful given the morbidity associated with invasive mechanical ventilation.

HFNC offers practical advantages over NIPPV including greater patient comfort, the ability to eat and speak during therapy, avoidance of gastric insufflation, reduced aspiration risk, and elimination of facial skin breakdown from mask interface pressure. These factors may improve adherence and tolerance, particularly during prolonged therapy.

Recommendation 69 — HFNC Over HFNC Alternating with NIPPV

For adults with sepsis or septic shock and acute hypoxemic respiratory failure, we suggest using HFNC over high flow nasal cannula alternating with noninvasive positive pressure ventilation.

Conditional recommendation, very low certainty evidence

Change from 2021: New recommendation.

Rationale: Some clinical practice involves alternating HFNC with periodic sessions of NIPPV in an attempt to combine the benefits of both modalities. The panel evaluated 7 RCTs comparing HFNC vs. NIPPV (2,465 patients) and 2 RCTs comparing HFNC vs. an alternating HFNC/NIPPV strategy (515 patients).

Key findings from meta-analysis (HFNC vs. alternating HFNC/NIPPV):

Although CIs were wide for mortality, point estimates favored HFNC alone. The increase in ventilator-free days (4 additional days) is a clinically important signal. The shorter hospital length of stay with HFNC alone further supports its use as a standalone therapy.

Practical advantages of HFNC alone over an alternating strategy:

• Simpler to administer and does not require switching between devices
• Improved patient comfort — avoids intermittent mask application
• Allows patients to eat, drink, and communicate continuously
• Eliminates risks associated with NIPPV (gastric insufflation, aspiration, facial skin damage)
• Reduces nursing workload and equipment requirements

4. Awake Proning

Recommendation 70 — Awake Proning in Non-Intubated Patients

For non-intubated adults with sepsis or septic shock and acute hypoxemic respiratory failure, we suggest a trial of awake proning.

Conditional recommendation, very low certainty evidence

Change from 2021: New recommendation.

Rationale: Awake (self) proning refers to placing non-intubated patients in the prone position to improve oxygenation and potentially reduce the need for intubation. The panel reviewed 17 RCTs enrolling 3,537 patients. Notably, all trials were conducted during the COVID-19 pandemic; applicability to non-COVID causes of acute hypoxemic respiratory failure requires further study.

Key findings from meta-analysis:

The reduction in intubation (RR 0.82) is the strongest and most consistent signal from the evidence. Two additional recent systematic reviews published outside the SSC evidence review confirmed the finding of reduced need for intubation and mechanical ventilation with awake proning.²

Practical considerations:

• Duration and frequency of proning sessions depend on patient tolerance
• Sedation should NOT be used to promote tolerance of proning in non-intubated patients
• Awake proning can be accomplished with minimal cost and no specialized equipment
• Represents an important equity consideration — particularly beneficial in low-resource settings where mechanical ventilation capacity may be limited
• May be combined with HFNC or NIPPV during proning sessions
• Monitoring during awake proning should include SpO2, respiratory rate, and assessment of patient comfort and work of breathing

The point estimate for serious adverse events (RR 2.54) warrants attention, although the wide confidence interval crossing 1.0 limits interpretation. Reported adverse events included pressure injuries, vomiting, and dislodgement of IV lines or monitoring equipment. Close monitoring during initial proning sessions is prudent.

5. Invasive Mechanical Ventilation

Recommendation 71 — Low Tidal Volume Ventilation for ARDS

For adults with sepsis-induced ARDS, we recommend using a low tidal volume ventilation strategy (6 mL/kg predicted body weight) over a high tidal volume strategy (> 10 mL/kg predicted body weight).

Strong recommendation, high certainty evidence

Change from 2021: Carryover.

Rationale: This landmark recommendation was carried over from 2021; please see the 2021 SSC guidelines for rationale.³ Low tidal volume ventilation remains the cornerstone of lung-protective ventilation. The seminal ARDS Network trial demonstrated a 22% relative reduction in mortality with 6 mL/kg vs. 12 mL/kg predicted body weight, establishing this as standard of care.

Low tidal volume ventilation reduces cyclic overdistension of aerated lung units, attenuates ventilator-induced lung injury (VILI), and limits biotrauma — the systemic release of inflammatory mediators from mechanically injured lung tissue. Tidal volumes should be calculated using predicted (ideal) body weight based on height and sex, not actual body weight, as lung size correlates with height rather than total body mass.

Predicted Body Weight Formulas:

• Males: 50 + 2.3 × (height in inches − 60) or 50 + 0.91 × (height in cm − 152.4)
• Females: 45.5 + 2.3 × (height in inches − 60) or 45.5 + 0.91 × (height in cm − 152.4)

Recommendation 72 — Tidal Volume Without ARDS

For adults with sepsis-associated hypoxemic respiratory failure without ARDS, we suggest using tidal volumes of 6–8 mL/kg ideal body weight (IBW) over lower tidal volumes (4 to < 6 mL/kg IBW).

Conditional recommendation, low certainty evidence

Change from 2021: New recommendation.

Rationale: While low tidal volume ventilation is well-established for ARDS, the optimal tidal volume for mechanically ventilated patients who do not meet ARDS criteria has been less clear. The panel reviewed 2 RCTs comparing low (4–6 mL/kg) vs. moderate (8–10 mL/kg) tidal volumes in general mechanically ventilated patients.

Key finding from meta-analysis:

Possibly no difference in 90-day mortality between low and moderate tidal volumes (RR 1.04; 95% CI, 0.89–1.22, low certainty).

The panel favored a tidal volume of 6–8 mL/kg IBW as the safest approach for several reasons:

1. ARDS is frequently under-recognized. Studies have demonstrated that 48.7–52.4% of patients meeting ARDS criteria are not identified by their treating clinicians. A tidal volume of 6–8 mL/kg provides a margin of safety for patients with early or unrecognized ARDS.
2. Very low tidal volumes (4–6 mL/kg) may require higher respiratory rates, increased sedation, and greater patient-ventilator dyssynchrony without demonstrated mortality benefit in non-ARDS patients.
3. A 6–8 mL/kg target aligns with the principle of avoiding volume-induced lung injury while maintaining adequate minute ventilation and patient comfort.

Remark: Clinicians should screen regularly for ARDS development in mechanically ventilated patients with sepsis, as tidal volumes should be reduced to 6 mL/kg IBW if ARDS criteria are met.

Recommendation 73 — Plateau Pressure Limits

For adults with sepsis-induced ARDS, we recommend an upper limit goal for plateau pressures of 30 cm H2O over higher plateau pressures.

Strong recommendation, high certainty evidence

Change from 2021: Carryover.

Rationale: This recommendation was carried over; please see the 2021 SSC guidelines for rationale.³ Plateau pressure reflects end-inspiratory alveolar distending pressure and is a surrogate for transpulmonary pressure and lung stress. Maintaining plateau pressure below 30 cm H2O limits alveolar overdistension, reduces barotrauma, and complements the low tidal volume strategy.

Plateau pressure should be measured using an inspiratory hold maneuver during volume-controlled ventilation. In patients with reduced chest wall compliance (e.g., morbid obesity, abdominal compartment syndrome, or large pleural effusions), elevated plateau pressures may not reflect true transpulmonary pressure, and esophageal manometry may be considered for a more accurate assessment.

Recommendation 74 — Higher PEEP for Moderate-Severe ARDS

For adults with sepsis-induced moderate-to-severe ARDS, we suggest using higher PEEP over lower PEEP.

Conditional recommendation, moderate certainty evidence

Change from 2021: Carryover.

Rationale: This recommendation was carried over; please see the 2021 SSC guidelines for rationale.³ Positive end-expiratory pressure (PEEP) prevents end-expiratory alveolar collapse, improves oxygenation, and reduces atelectrauma — the shear stress generated by cyclic opening and closing of atelectatic lung units.

Three major RCTs (ALVEOLI, LOV, ExPress) comparing higher vs. lower PEEP strategies in ARDS were pooled in an individual patient data meta-analysis, which demonstrated a mortality benefit with higher PEEP in patients with moderate-to-severe ARDS (PaO2/FiO2 ≤ 200 mm Hg). Conversely, higher PEEP showed no benefit (and a trend toward harm) in patients with mild ARDS.

ARDS Severity Classification (Berlin Definition):

Higher PEEP should be titrated carefully, with monitoring for adverse hemodynamic effects (reduced preload and cardiac output) and overdistension of already-aerated lung regions.

Recommendation 75 — Against Incremental PEEP Titration for Moderate-Severe ARDS

For adults with sepsis-induced moderate-to-severe ARDS, we recommend against using an incremental PEEP titration strategy (recruitment maneuver) for moderate-to-severe ARDS.

Strong recommendation, moderate certainty evidence

Change from 2021: Carryover.

Rationale: This recommendation was carried over; please see the 2021 SSC guidelines for rationale.³ Incremental (stepwise) PEEP titration strategies — involving sustained inflation or stepwise recruitment maneuvers followed by a decremental PEEP trial to identify “optimal PEEP” — were evaluated in the ART trial, which randomized 1,010 patients with moderate-to-severe ARDS.

The ART trial found that the recruitment maneuver and PEEP titration strategy increased 28-day mortality (55.3% vs. 49.3%; HR 1.20; 95% CI, 1.01–1.42), 6-month mortality, and was associated with more barotrauma, hemodynamic instability, and cardiac arrest. This led to a strong recommendation against this approach.

Standard PEEP tables (e.g., ARDSNet lower and higher PEEP/FiO2 tables) remain the recommended approach for setting PEEP in conjunction with clinical assessment of oxygenation, compliance, and hemodynamics.

Recommendation 76 — Prone Ventilation for Moderate-Severe ARDS

For adults with sepsis-induced moderate-to-severe ARDS, we suggest using prone ventilation for greater than 12 hours daily.

Conditional recommendation, moderate certainty evidence

Change from 2021: Carryover.

Rationale: This recommendation was carried over; please see the 2021 SSC guidelines for rationale.³ Prone positioning improves oxygenation through more homogeneous pleural pressure distribution, improved ventilation-perfusion matching, recruitment of dorsal atelectasis, and improved drainage of secretions. In moderate-to-severe ARDS, it also reduces VILI by creating a more uniform distribution of mechanical stress across the lung.

The landmark PROSEVA trial demonstrated that early prone positioning for at least 16 hours per day in patients with severe ARDS (PaO2/FiO2 < 150 mm Hg on FiO2 ≥ 0.6, PEEP ≥ 5 cm H2O) reduced 28-day mortality (16.0% vs. 32.8%; HR 0.39; 95% CI, 0.25–0.63) and 90-day mortality. Meta-analyses have consistently supported a mortality benefit when prone positioning is applied for more than 12 hours per day and combined with lung-protective ventilation.⁴

Implementation considerations:

• Requires trained staff and a standardized turning protocol
• Contraindications: spinal instability, open abdomen, unstable pelvic or long bone fractures
• Monitor for pressure injuries (face, chest, pelvis), endotracheal tube displacement, and loss of vascular access
• Aim for 16+ hours per day when feasible
• Continue until sustained improvement in oxygenation (PaO2/FiO2 ≥ 150 mm Hg in supine position)

Recommendation 77 — Intermittent NMBA Boluses Over Continuous Infusion

For adults with sepsis-induced moderate-to-severe ARDS, we suggest using intermittent NMBA boluses over continuous NMBA infusion.

Conditional recommendation, moderate certainty evidence

Change from 2021: Carryover.

Rationale: This recommendation was carried over; please see the 2021 SSC guidelines for rationale.³ Neuromuscular blocking agents (NMBAs) are used in ARDS to reduce patient-ventilator dyssynchrony, decrease oxygen consumption by respiratory muscles, and facilitate lung-protective ventilation when deep sedation alone is insufficient.

The ACURASYS trial (340 patients with early severe ARDS) showed reduced 90-day mortality with 48 hours of continuous cisatracurium infusion vs. placebo (31.6% vs. 40.7%; HR 0.68; 95% CI, 0.48–0.98). However, the larger ROSE trial (1,006 patients) found no mortality benefit with early continuous cisatracurium infusion (42.5% vs. 42.8%) when compared with a lighter sedation strategy, and the continuous infusion group had more ICU-acquired weakness and serious cardiovascular events.

Given the lack of demonstrated benefit and potential for harm with continuous infusion — and the ability of intermittent boluses to manage ventilator dyssynchrony on a targeted, as-needed basis — the panel suggests intermittent NMBA boluses as the preferred approach.

Clinical guidance:

• Use the lowest effective dose to achieve ventilator synchrony
• Train-of-four monitoring is recommended during NMBA use
• Ensure adequate sedation and analgesia prior to and during NMBA administration
• Reassess the need for NMBA at least daily with planned dose reductions or cessation

6. Venovenous ECMO

Recommendation 78 — VV-ECMO for Refractory ARDS

For adults with severe ARDS due to sepsis when conventional mechanical ventilation fails, we suggest using venovenous ECMO (VV-ECMO) in experienced centers with the infrastructure to support its use.

Conditional recommendation, low certainty evidence

Change from 2021: Carryover.

Rationale: This recommendation was carried over; please see the 2021 SSC guidelines for rationale.³ VV-ECMO provides extracorporeal gas exchange, allowing ultra-protective ventilation (or near-apneic ventilation) to minimize ongoing ventilator-induced lung injury in refractory severe ARDS.

The EOLIA trial randomized 249 patients with very severe ARDS (PaO2/FiO2 < 50 mm Hg for > 3 hours, or PaO2/FiO2 < 80 mm Hg for > 6 hours, or pH < 7.25 with PaCO2 ≥ 60 mm Hg for > 6 hours despite optimization) to early VV-ECMO vs. conventional ventilation. The primary outcome of 60-day mortality showed a non-significant trend favoring ECMO (35% vs. 46%; RR 0.76; 95% CI, 0.55–1.04; p = 0.09), though 28% of the control group crossed over to rescue ECMO. A post-hoc Bayesian analysis and subsequent meta-analyses including the EOLIA trial have demonstrated a likely mortality benefit.⁵

Prerequisites for VV-ECMO:

• Experienced ECMO center with adequate case volume and trained multidisciplinary team
• Infrastructure for cannulation, circuit management, anticoagulation monitoring, and complication management
• Patient selection: refractory hypoxemia despite optimized conventional management (prone positioning, neuromuscular blockade, higher PEEP)
• Timing: earlier initiation (before irreversible multi-organ failure) may improve outcomes
• Contraindications: irreversible underlying condition, severe multi-organ failure with futility, contraindication to anticoagulation

VV-ECMO is a resource-intensive intervention and is not universally available. Transfer to ECMO-capable centers should be considered early in the course of refractory ARDS when local expertise is unavailable. Regional ECMO referral networks and mobile ECMO retrieval teams have improved access in many healthcare systems.

Quick Reference: Respiratory Support Summary

RESPIRATORY SUPPORT — AT A GLANCE

A. MONITORING OF HYPOXEMIA
   ✓  Measure oxygenation with SpO2 or ABG + clinical acumen          [Conditional] NEW

B. OXYGEN TARGETS
   ✓  Titrate FiO2 to liberal or conservative target per patient factors [Conditional] NEW
      — Lower target ~90-93% SpO2; higher target ≥96% SpO2
      — No mortality difference (RR 1.02; 95% CI 0.99-1.05)

C. NONINVASIVE RESPIRATORY SUPPORT
   ✓  HFNC over conventional O2 therapy (PaO2/FiO2 <200)              [Conditional] NEW
   ✓  HFNC over NIPPV as initial therapy                              [Conditional] NEW
   ✓  HFNC alone over HFNC alternating with NIPPV                     [Conditional] NEW

D. AWAKE PRONING
   ✓  Trial of awake proning for non-intubated patients                [Conditional] NEW
      — Reduced intubation (RR 0.82); do NOT sedate to promote tolerance

E. INVASIVE MECHANICAL VENTILATION
   ✓✓ Low tidal volume (6 mL/kg PBW) for ARDS                         [Strong] CARRYOVER
   ✓  Tidal volume 6-8 mL/kg IBW without ARDS (over 4 to <6 mL/kg)   [Conditional] NEW
   ✓✓ Plateau pressure upper limit 30 cm H2O                          [Strong] CARRYOVER
   ✓  Higher PEEP over lower PEEP for moderate-severe ARDS             [Conditional] CARRYOVER
   ✗✗ Against incremental PEEP titration / recruitment maneuvers       [Strong against] CARRYOVER
   ✓  Prone ventilation >12 hours daily for moderate-severe ARDS       [Conditional] CARRYOVER
   ✓  Intermittent NMBA boluses over continuous NMBA infusion          [Conditional] CARRYOVER

F. VENOVENOUS ECMO
   ✓  VV-ECMO for refractory severe ARDS in experienced centers        [Conditional] CARRYOVER

KEY: ✓✓ = Strong recommendation ("we recommend")
     ✓  = Conditional recommendation ("we suggest")
     ✗✗ = Strong recommendation against
     GPS = Good practice statement
     NEW = New in 2026 | CARRYOVER = Carried over from 2021

References