Skip to content

Surviving Sepsis Campaign 2026: adult sepsis and septic shock

International SCCM/ESICM Current

Surviving Sepsis Campaign: international guidelines for management of sepsis and septic shock 2026

Prescott HC, Antonelli M, et al. Intensive Care Med. 2026;52:863–936
DOI 10.1007/s00134-026-08361-1 · Summary checked against the source: 2026-11

Read or download the original

The 2026 guideline has 107 numbered statements. Citations give the strength, the certainty of evidence where it is clear in the source, and the statement number with its page in Table 1. Strength: strong = "we recommend"; conditional = "we suggest" (called weak before 2026).

At a glance

  1. Sepsis is life-threatening acute organ dysfunction due to infection (Sepsis-3); septic shock is the subset with circulatory dysfunction and higher mortality. It's a clinical diagnosis: no single biomarker rules it in or out. [GPS · #5, p. 864]
  2. Screen with NEWS, NEWS2, MEWS or SIRS, not qSOFA alone. [Strong, moderate · #4, p. 864]
  3. Antimicrobials within 1 hour for septic shock and for probable or definite sepsis. Possible sepsis without shock: rapid assessment, then antibiotics within 3 hours if concern persists. [#16–18, p. 865]
  4. At least 30 mL/kg crystalloid in the first 3 hours for hypoperfusion or shock, reassessing all the time. [Conditional, low · #10, p. 864]
  5. Noradrenaline first, started peripherally rather than waiting for a central line. MAP 65 mm Hg (60–65 in people aged 65 or over). [#12–14, #53, p. 864–866]
  6. Balanced crystalloids over saline; crystalloids alone over adding albumin; never starches. [#43–46, p. 866]
  7. IV hydrocortisone in septic shock. [Conditional, low · #79, p. 868]
  8. New in 2026: HFNC first for hypoxaemic respiratory failure, a trial of awake proning, and active fluid removal after resuscitation. [p. 867–868]

Screening and early management

Strong recommendations and good practice [p. 864]

  • Performance improvement programmes with sepsis screening, standard operating procedures and quality improvement. [Strong, moderate · #1]
  • NEWS, NEWS2, MEWS or SIRS over qSOFA as a single screening tool. [Strong, moderate · #4]
  • Blood cultures as soon as possible, ideally before antimicrobials. [Strong, low · #7]
  • Sepsis and septic shock are emergencies: start treatment and resuscitation immediately. [GPS · #9]
  • Initial MAP target 65 mm Hg over higher targets (in practice, a range within about 5 mm Hg). [Strong, moderate · #13]

Conditional recommendations [p. 864]

  • Measure lactate. Individualise fluids after the first bolus by following the fall in lactate, not by giving fluid until lactate normalises. [Conditional · #8]
  • ≥30 mL/kg IV crystalloid in the first 3 hours for sepsis-induced hypoperfusion or shock. Use ideal or adjusted body weight if BMI >30. [Conditional, low · #10]
  • Hypotension: crystalloid bolus first, then vasopressor if hypotension persists. Unstable shock (very low BP, mottling, cyanosis, altered mentation): give fluid and vasopressor together. [Conditional · #11]
  • Start vasopressors peripherally rather than delay until central access is secured. [Conditional · #12]
  • Age ≥65: initial MAP 60–65 mm Hg over higher ranges. [Conditional · #14]
  • Admit to ICU within 6 hours if ICU care is needed. [Conditional, low · #15]
  • A "code sepsis" or "sepsis huddle" protocol. [Conditional · #2]

Infection management

Situation Antimicrobial timing Grade
Possible, probable or definite septic shock Immediately, ideally within 1 h of recognition Strong · #16
Probable or definite sepsis Immediately, ideally within 1 h Strong · #17
Possible sepsis without shock Rapid investigation; antimicrobials within 3 h if concern persists Conditional · #18
Low likelihood of infection, no shock Defer antimicrobials and monitor closely Conditional · #20
Septic shock with >60 min to hospital Give antimicrobials in the ambulance (with a screening system in place) Conditional · #21

Adapted from SSC 2026 Table 1, statements 16–21 [p. 865]

Strong and good practice [p. 865]

  • Find the source fast: rapidly evaluate for a source needing emergency source control, ideally within 6 hours. [GPS · #23; Conditional · #24]
  • Prolonged infusion of beta-lactams for maintenance, after a loading dose. [Strong, moderate · #33]
  • De-escalate once the microbiology and susceptibilities are known. [Strong · #36]
  • Keep reassessing; stop empirical antimicrobials if a non-infectious cause is found. [GPS · #35]

Conditional [p. 865]

  • High risk of an MDR organism: cover it empirically. Low risk: don't. [Conditional · #25–26]
  • No routine empirical antifungals; consider them case by case (immunosuppression, prolonged antibiotics or admission, intra-abdominal source). [Conditional against · #27]
  • Anaerobic cover only with risk factors (intra-abdominal, gynaecological, necrotising soft tissue, head and neck, CNS abscess). [Conditional · #28–29]
  • Shorter courses after adequate source control; use procalcitonin plus clinical evaluation to decide when to stop. [Conditional · #39–40]
  • Use clinical evaluation alone, not procalcitonin, to decide whether to start. [Conditional · #22]
  • Therapeutic drug monitoring case by case. [Conditional · #34]

Haemodynamic management

Fluids

Statement Grade
Crystalloids first line for resuscitation Strong, moderate · #43
Balanced crystalloids over 0.9% saline (saline for traumatic brain injury) Conditional · #44
Crystalloids alone over crystalloids plus albumin Conditional, moderate · #45
Starches Strong against, high · #46
Gelatin Conditional against · #47
After 30 mL/kg with persistent hypoperfusion: liberal or restrictive strategy, depending on the patient Conditional, low · #48
Dynamic measures (passive leg raise, SV, SVV, PPV) over examination or static measures Conditional · #49
Serial lactate and capillary refill time to guide resuscitation Conditional · #51–52

Adapted from SSC 2026 Table 1, statements 43–52 [p. 866]

Vasopressors and inotropes

flowchart TD
    A["Septic shock: hypotension after fluids<br/>(unstable shock: fluids and vasopressor together)"] --> B["Noradrenaline first line, may start peripherally<br/>MAP 65 mm Hg (60–65 if aged ≥65)"]
    B --> C{"Escalating noradrenaline<br/>dose?"}
    C -->|Yes| D["Add vasopressin<br/>Conditional"]
    D --> E{"MAP still inadequate?"}
    E -->|Yes| F["Add adrenaline<br/>Conditional"]
    B --> G{"Cardiac dysfunction with persistent<br/>hypoperfusion despite fluids and BP?"}
    G -->|Yes| H["Add an inotrope:<br/>dobutamine + noradrenaline<br/>or adrenaline alone"]
    B --> I["Refractory shock on vasopressors:<br/>IV hydrocortisone, Conditional"]
Built from SSC 2026 statements 53–61 and 79 [p. 866, 868]; the guideline has no figure for this.

Statement Grade
Noradrenaline first line over dopamine (high certainty), adrenaline or selepressin Strong · #53
Noradrenaline first line over vasopressin or angiotensin II Conditional · #55 (downgraded from strong)
Terlipressin Conditional against · #54
Add vasopressin on escalating noradrenaline Conditional · #56
Add adrenaline if MAP still inadequate on noradrenaline + vasopressin Conditional · #57
Cardiac dysfunction: either noradrenaline or adrenaline first line (noradrenaline if tachyarrhythmia, adrenaline if bradycardia) Conditional · #58
Cardiac dysfunction with persistent hypoperfusion: inotrope plus vasopressor Conditional · #60
Levosimendan Conditional against · #62
Beta-blockers (esmolol, landiolol) as a treatment for septic shock Conditional against · #64
Methylene blue, midodrine Insufficient evidence · #59, #63
Invasive arterial monitoring if vasopressors are escalating or BP readings unreliable; either invasive or non-invasive otherwise Conditional · #42

Adapted from SSC 2026 Table 1, statements 42–64 [p. 866]

Respiratory support

Statement Grade
Measure oxygenation by SpO₂ or SaO₂ with clinical assessment Conditional · #65
HFNC over conventional oxygen for acute hypoxaemic respiratory failure (P/F <200) Conditional · #67
HFNC over NIV as initial therapy Conditional · #68
Awake proning trial if not intubated Conditional · #70
ARDS: tidal volume 6 mL/kg predicted body weight (not >10 mL/kg) Strong, high · #71
Without ARDS: 6–8 mL/kg over 4 to <6 mL/kg Conditional · #72
Plateau pressure ≤30 cm H₂O Strong, high · #73
Moderate–severe ARDS: higher PEEP Conditional, moderate · #74
Incremental PEEP titration (recruitment) strategy Strong against, moderate · #75
Prone ventilation >12 h a day Conditional, moderate · #76
Intermittent NMBA boluses over a continuous infusion Conditional, moderate · #77
VV-ECMO for severe ARDS when ventilation fails, in an experienced centre Conditional · #78

Adapted from SSC 2026 Table 1, statements 65–78 [p. 867]

Additional therapies

Strong [p. 868]

  • Restrictive transfusion strategy over liberal (threshold Hb 7 g/dL in the 2021 guideline). [Strong, moderate · #90]
  • Start insulin at glucose ≥180 mg/dL (10 mmol/L). [Strong, moderate · #92]
  • Pharmacological VTE prophylaxis, with LMWH over UFH. [Strong, moderate · #97–98]

Conditional for [p. 868]

  • IV corticosteroids in septic shock (2021: hydrocortisone 200 mg/day when on noradrenaline ≥0.25 µg/kg/min for ≥4 h). [Conditional, low · #79]
  • Stress ulcer prophylaxis with a PPI if there are GI bleeding risk factors. [Conditional, moderate · #87]
  • Active fluid removal (diuretics or ultrafiltration) after the acute resuscitation phase. [Conditional · #89]
  • Early enteral nutrition within 72 hours. [Conditional · #91]
  • Bicarbonate only for pH ≤7.2 with AKI (AKIN stage 2–3). [Conditional · #96]
  • AKI needing RRT: continuous or intermittent, either. [#94]

Conditional against [p. 868]

  • Antipyretics or cooling for fever, to improve outcomes. [Conditional against · #80]
  • IV vitamin C, IV immunoglobulin, vitamin D, probiotics. [Conditional against · #81–82, #85, #88]
  • Blood purification (haemoperfusion, high-volume haemofiltration, plasma exchange) and polymyxin B haemoperfusion. [Conditional against · #83–84]
  • RRT without a definite indication. [Conditional against · #93]
  • Bicarbonate for hypoperfusion-related lactic acidaemia. [Conditional against · #95]

Goals of care

  • Discuss goals of care and prognosis with patients and families. [GPS · #100, p. 869]
  • Early (within 72 h) rather than late goals-of-care discussions. [Conditional · #101, p. 869]
  • Integrate palliative care principles; no routine formal palliative consultation for everyone. [GPS · #106; Conditional against · #107, p. 869]

Clinical pearls

  • Sepsis-3: organ dysfunction = SOFA rise ≥2. Septic shock = vasopressors for MAP ≥65 plus lactate >2 mmol/L despite adequate fluids.
  • qSOFA (RR ≥22, altered mentation, SBP ≤100) predicts death but is not a stand-alone screening tool.
  • Hour-1 bundle (SSC 2018): lactate, blood cultures, broad-spectrum antibiotics, 30 mL/kg crystalloid for hypotension or lactate ≥4, vasopressors for MAP <65.
  • Noradrenaline → vasopressin → adrenaline. Dobutamine for cardiac dysfunction.
  • ARDS: 6 mL/kg, plateau ≤30, prone >12 h; no incremental recruitment manoeuvres.

Classes and levels of evidence

How SSC 2026 grades its recommendations. Adapted from SSC 2026 Methods and Table 2 [p. 870–871]. Hover over any class or level cited on this page to see what it means.

Strength Meaning Wording used
Strong Most, if not all, patients should receive the intervention we recommend
Conditional The balance of benefit and harm varies with patient values, circumstances or resources; most would want it, but many would not we suggest
Conditional against Most patients should probably not receive the intervention, but the balance can vary we suggest against
Strong against Most, if not all, patients should not receive the intervention we recommend against
Good practice statement Ungraded: clear, necessary practice whose net benefit is unequivocal and not suited to formal grading clinicians should …
Level Meaning
High certainty Very confident that the true effect lies close to the estimate
Moderate certainty Moderately confident; the true effect is likely close to the estimate but may be substantially different
Low certainty Limited confidence; the true effect may be substantially different
Very low certainty Very little confidence; the true effect is likely substantially different

SSC grades certainty of evidence and strength of recommendation with GRADE. "We recommend" means a strong recommendation; "we suggest" means a conditional one (called weak before 2026). Good practice statements are ungraded statements whose net benefit is unequivocal. Where the panel found too little evidence it says so, sometimes adding an "in our practice" note describing what panellists do.