Surviving Sepsis 2026: what changed since 2021

A 72-year-old man arrives at the emergency department with three days of abdominal pain, fever and vomiting. He has hypertension, weighs 78 kg and does not have heart failure. Blood pressure is 82/48 mmHg, heart rate is 124 per minute, respiratory rate is 26 per minute and he is confused. His knees are mottled and capillary refill time is over five seconds. Lactate is 4.6 mmol/L. The CT scan shows perforated appendicitis with a pelvic collection. The team asks what changed in the guideline since the last time they studied the topic. The scene is fictional and was written for teaching.

This text uses the case as a clinical reasoning scenario for students and health professionals. It compares the 2021 international Surviving Sepsis Campaign guideline with the 2026 update for adults, point by point, always from the original texts. The focus is what changes for the physician at the bedside: screening, antimicrobial timing, fluids, pressure target, vasopressors, corticosteroids, oxygen and the new SOFA-2 score. It does not replace your institution's protocol. If you play the case of the day on NEXO, you will recognize the challenge: deciding with clues that arrive in sequence.

Key points

• The 2026 guideline has 129 statements, 46 of them new, and updates the 2021 recommendations.
• Antimicrobials remain immediate, preferably within one hour, in shock and in probable or definite sepsis.
• In possible sepsis without shock, the rapid investigation may last up to three hours before the antimicrobial.
• Prolonged beta-lactam infusion went from a suggestion to a strong recommendation.
• In people aged 65 or older, the guideline suggests an initial mean arterial pressure target of 60 to 65 mmHg.
• After the first 30 mL/kg, a liberal or restrictive fluid strategy can be chosen according to the patient and the service.

Why this is a reasoning scenario

Sepsis is an emergency, and the 2026 guideline opens the resuscitation section with a good practice statement: treatment and resuscitation should begin immediately (2026 recommendations on the campaign page). That has not changed. What changed was the level of detail. The campaign page says the update reinforces the pillars of care, which are early recognition, timely treatment of the infection and hemodynamic resuscitation, and adds more clarity and nuance for real decisions (2026 recommendations on the campaign page).

The scenario is good for training because the man in the scene brings together several new decisions. He has unstable shock, he is over 65 and he has a source that needs surgical control. Each of these points gained its own recommendation in 2026. The post on signs, symptoms and history helps value mottled skin, confusion and slow capillary refill, and the post on clinical decision-making reminds us that each test needs to change a decision.

What the scene teaches

The scene trains six habits. The first is recognizing shock through the physical exam, not only through blood pressure. The second is collecting blood cultures and giving the antimicrobial without delay. The third is deciding the initial fluid volume by weight and by response. The fourth is starting the vasopressor through a peripheral vein if needed. The fifth is choosing the pressure target by age. The sixth is calling surgery early to control the source.

What is not at stake

The text does not discuss sepsis in children, which has its own 2026 guideline, nor details of antibiotics for each focus. Nursing screening was covered in another blog text. The post on diagnostic uncertainty helps in the opposite case to the scene, when the infection is only a possibility.

How to use the comparison

A new guideline does not make everything done before wrong. Many recommendations were kept, some gained strength and few changed direction. Knowing which is which avoids two traps: abandoning what still holds and keeping a habit that the new evidence weakened.

What the 2026 guideline is

The 2026 guideline was published by the Society of Critical Care Medicine and the European Society of Intensive Care Medicine, in the journals Critical Care Medicine and Intensive Care Medicine (2026 recommendations on the campaign page). The panel had 69 people from 23 countries, and 38 percent work or have worked in low- or middle-income countries (2026 recommendations on the campaign page). The guideline covers adults with sepsis in the hospital, in transport before the hospital and shortly after discharge (2026 recommendations on the campaign page).

What the 2021 guideline was

The 2021 guideline had the same structure of strong recommendations, weak recommendations and good practice statements (Evans et al., 2021). A review of what it brought that was new highlighted the end of qSOFA as a single screening tool, the use of capillary refill time as an adjuvant in resuscitation and a new block of recommendations on discharge (Prescott and Ostermann, 2023). The same review notes that several points were left without a recommendation in 2021, such as the fluid strategy after the initial phase (Prescott and Ostermann, 2023). Some of these gaps were filled in 2026 (2026 recommendations on the campaign page).

SOFA-2

In parallel, the SOFA score was revised. A 2023 article noted that the original SOFA was created in 1996 and that parts of it no longer reflected current practice (Moreno et al., 2023). SOFA-2 keeps the 0 to 4 scale per organ and updates the respiratory, cardiovascular and renal domains, including renal replacement therapy or eligibility for it when it is not available (Lobo et al., 2025). For the brain, treatment of delirium now counts as a sign of dysfunction (Lobo et al., 2025). A study of 74615 adults with suspected infection showed agreement between the two scores in 89.62 percent of patients, and the authors suggest using both during the transition until new sepsis criteria are validated (Zhu et al., 2026). Another study, in sepsis linked to pneumonia, found prognostic performance of SOFA-2 comparable to that of the old score (Wei et al., 2026).

The table summarizes what changed between the two guidelines on the main points.

Topic	2021 guideline	2026 guideline	Source
Screening	Against qSOFA as a single tool, compared with SIRS, NEWS or MEWS	NEWS, NEWS2, MEWS or SIRS instead of qSOFA as a single tool	(Evans et al., 2021) (2026 recommendations on the campaign page)
Antimicrobial in possible sepsis without shock	Rapid investigation and antimicrobial within three hours if suspicion persists	Kept, with a scheme that separates possible, probable and definite sepsis	(Evans et al., 2021) (2026 antimicrobial timing scheme)
Antimicrobial before the hospital	No specific recommendation	Suggests giving it in the ambulance in septic shock when hospital assessment would take more than 60 minutes	(2026 recommendations on the campaign page)
Beta-lactam	Suggests prolonged infusion	Recommends prolonged infusion, strong recommendation	(Evans et al., 2021) (2026 recommendations on the campaign page)
Albumin	Suggests albumin in those who received large volumes of crystalloid	Suggests crystalloid alone instead of crystalloid with supplemental albumin	(Evans et al., 2021) (2026 recommendations on the campaign page)
Fluid after 30 mL/kg	Insufficient evidence for liberal versus restrictive	Suggests liberal or restrictive according to the patient and the service	(Evans et al., 2021) (2026 recommendations on the campaign page)
Pressure target	Initial mean arterial pressure of 65 mmHg	Kept, and 60 to 65 mmHg in those aged 65 or older	(Evans et al., 2021) (2026 recommendations on the campaign page)
Corticosteroid	Suggests intravenous corticosteroid with ongoing need for vasopressor	Suggests intravenous corticosteroid in septic shock, with low certainty	(Evans et al., 2021) (2026 recommendations on the campaign page)

Recognize and treat the infection

Screening

In 2021, the guideline recommended not using qSOFA, compared with SIRS, NEWS or MEWS, as the sole screening tool, with a strong recommendation and moderate-quality evidence (Evans et al., 2021). In 2026, the direction is the same, written in positive form: use NEWS, NEWS2, MEWS or SIRS instead of qSOFA (2026 recommendations on the campaign page). The guideline adds that sepsis is a clinical diagnosis and should not be confirmed or ruled out by a single biomarker or test (2026 recommendations on the campaign page). It also suggests sepsis code protocols or a rapid bedside huddle after a positive screen (2026 recommendations on the campaign page). A meta-analysis of 22 studies with 19580 patients linked sepsis alert systems in the emergency department to a lower risk of death, with a relative risk of 0.81 (Kim et al., 2024).

Blood cultures and lactate

The 2026 guideline recommends collecting blood cultures as soon as possible and, preferably, before the antimicrobial (2026 recommendations on the campaign page). It suggests measuring lactate in adults with possible, probable or definite sepsis (2026 recommendations on the campaign page). In the man in the scene, both happen in parallel with venous access, without delaying treatment.

Antimicrobial timing

In 2021, the guideline recommended immediate antimicrobial, preferably within one hour, in possible shock or high-probability sepsis (Evans et al., 2021). In possible sepsis without shock, it suggested rapid investigation and antimicrobial within three hours if the suspicion persisted (Evans et al., 2021). In 2026, the scheme became clearer. In possible, probable or definite septic shock, the recommendation is immediate antimicrobial, preferably within one hour (2026 recommendations on the campaign page). In probable or definite sepsis without shock, also within one hour (2026 recommendations on the campaign page). In possible sepsis without shock, the suggestion is a short period of rapid investigation and antimicrobial within three hours if the suspicion continues (2026 recommendations on the campaign page). With low probability of infection and no shock, the suggestion is to defer the antimicrobial and monitor closely (2026 recommendations on the campaign page).

The 2026 scheme defines probable sepsis as high suspicion, in which sepsis is the most likely diagnosis, and possible sepsis as moderate suspicion, in which another diagnosis is also likely (2026 antimicrobial timing scheme). The rapid assessment includes history, examination, tests for infectious and non-infectious causes and immediate treatment of conditions that mimic sepsis (2026 antimicrobial timing scheme). The man in the scene is in shock, so there is no room to wait: the antimicrobial comes in the first hour.

Source control and choice of antimicrobial

The 2026 guideline suggests early source control, preferably within six hours of diagnosis, when there is a focus that needs to be drained or removed (2026 recommendations on the campaign page). Without risk factors for anaerobes, it suggests a regimen without anaerobic coverage; with an abdominal focus, deep gynecological infection, necrotizing soft tissue infection, head and neck infection or central nervous system abscess, it suggests including coverage (2026 recommendations on the campaign page). In the man in the scene, perforated appendicitis with a collection calls for anaerobic coverage and drainage or surgery on the same day. The guideline suggests not using procalcitonin to decide whether to start the antimicrobial, but suggests using it with the clinical assessment to decide when to stop, if the ideal duration is not clear (2026 recommendations on the campaign page).

Prolonged infusion and de-escalation

In 2021, prolonged infusion of a beta-lactam for maintenance, after an initial dose, was a weak suggestion with moderate evidence (Evans et al., 2021). In 2026, it became a strong recommendation (2026 recommendations on the campaign page). The guideline also recommends de-escalation when there is a microbiological diagnosis with a susceptibility profile, suggests de-escalation even with negative cultures, and suggests shorter courses when the source was well controlled (2026 recommendations on the campaign page). The post about differential diagnosis helps with the good practice of reassessing and stopping the antimicrobial when another cause appears (2026 recommendations on the campaign page).

Regulatory status: FDA and EMA

This text is about a clinical guideline, not a new drug. So the check for FDA and EMA approval does not apply to the central topic. The drugs mentioned, such as norepinephrine, vasopressin, epinephrine, hydrocortisone and beta-lactams, have long been in use, and this text does not discuss any newly approved drug. The only mention of a regulatory status is XueBiJing, which the guideline suggests not using outside the jurisdictions where it is approved (2026 recommendations on the campaign page). Terlipressin, intravenous vitamin C, immunoglobulins and polymyxin B hemoperfusion appear with a suggestion against use (2026 recommendations on the campaign page).

Guideline dates

The 2021 version was published in Intensive Care Medicine in 2021 (Evans et al., 2021). The 2026 version is on the campaign page, which provides the filterable list of recommendations, the quick resuscitation guide and the antimicrobial timing scheme, all consulted on October 24, 2026 (2026 recommendations on the campaign page) (2026 quick resuscitation guide) (2026 antimicrobial timing scheme). The post about interprofessional clinical reasoning helps bring this update to nursing, pharmacy and medicine at the same time.

Fluids, pressure and vasopressors

Initial volume

In 2021, the guideline suggested at least 30 mL/kg of intravenous crystalloid in the first three hours, with low-quality evidence (Evans et al., 2021). In 2026, the suggestion remains, also with low certainty (2026 recommendations on the campaign page). The remarks are new: consider the patient's characteristics, reassess frequently to avoid too little or too much fluid, and calculate using actual weight, or adjusted or ideal weight when the body mass index is above 30 kg/m² (2026 recommendations on the campaign page). For the man in the scene, at 78 kg, that comes to about 2340 mL in the first three hours, with reassessment after each bolus.

Fluid and vasopressor together

The 2026 guideline suggests a crystalloid bolus followed by a vasopressor if hypotension persists (2026 recommendations on the campaign page). The remark says that, in unstable shock, giving a vasopressor together with fluid may be justified, case by case (2026 recommendations on the campaign page). Signs of unstable shock include very low pressure, mottled skin, grayish appearance, cyanosis or falling saturation, tachycardia and altered mental status (2026 recommendations on the campaign page). The man in the scene has almost all of them. The quick guide carries the same idea: consider norepinephrine together with fluid when there is life-threatening hypoperfusion (2026 quick resuscitation guide).

Type of fluid

The 2026 guideline recommends crystalloids as the first choice and suggests balanced crystalloids instead of 0.9 percent saline in initial resuscitation, now with moderate certainty (2026 recommendations on the campaign page). In 2021, the same suggestion had low-quality evidence (Evans and colleagues, 2021). In head trauma, the 2026 guideline suggests saline (2026 recommendations on the campaign page). On albumin, the direction changed: in 2021 there was a suggestion to use albumin in those who received large volumes of crystalloid (Evans and colleagues, 2021); in 2026, the suggestion is crystalloid alone, with the caveat that albumin may be appropriate after large volumes or in cirrhosis (2026 recommendations on the campaign page). Starches remain prohibited with a strong recommendation (2026 recommendations on the campaign page).

After the first 30 mL/kg

In 2021, there was not enough evidence to choose between a liberal and a restrictive strategy after initial resuscitation (Evans and colleagues, 2021). In 2026, the suggestion is to use either one according to the patient and the service, taking into account conditions such as heart failure and the availability of monitored beds (2026 recommendations on the campaign page). The guideline suggests dynamic measures, such as passive leg raising or stroke volume response to a bolus, instead of physical examination or static measures alone (2026 recommendations on the campaign page). It suggests serial lactate and capillary refill time as support, and notes that fluid should not continue until lactate normalizes (2026 recommendations on the campaign page). A review on microcirculation notes that capillary refill is simple, but subjective and sensitive to the environment (Serova and colleagues, 2026).

Pressure target

In 2021, the recommendation was an initial mean arterial pressure of 65 mmHg instead of higher targets, a strong recommendation (Evans and colleagues, 2021). In 2026, it remains, with the observation that a range of about 5 mmHg is reasonable (2026 recommendations on the campaign page). The new element is the suggestion of an initial target of 60 to 65 mmHg in those aged 65 or older (2026 recommendations on the campaign page). The man in the scene is 72 years old, and the target becomes this range.

Vasopressors

Norepinephrine remains the first choice, with a strong recommendation against dopamine, epinephrine or selepressin as first line (2026 recommendations on the campaign page). The 2026 guideline suggests starting the vasopressor through a peripheral vein instead of waiting for central access, as the 2021 one already suggested (2026 recommendations on the campaign page) (Evans and colleagues, 2021). In the analysis of the CLOVERS study, 490 of 582 patients started the vasopressor through a peripheral vein, and there was a peripheral complication in 3 of them, 0.6 percent, all from mild extravasation (Munroe and colleagues, 2025). A meta-analysis of 11 studies with adults estimated adverse events in 1.8 percent of peripheral infusions, mostly mild (Owen and colleagues, 2021).

With increasing doses of norepinephrine, the 2026 guideline suggests adding vasopressin, with moderate certainty (2026 recommendations on the campaign page). In 2021, the suggestion was to add vasopressin instead of increasing norepinephrine, and the panel usually started norepinephrine between 0.25 and 0.5 μg/kg/min (Evans and colleagues, 2021). If pressure remains low with norepinephrine and vasopressin, the suggestion is to add epinephrine (2026 recommendations on the campaign page). In cardiac dysfunction with persistent hypoperfusion, the guideline suggests dobutamine with norepinephrine or epinephrine alone, and suggests not using levosimendan or a short-acting beta-blocker (2026 recommendations on the campaign page). A review on hemodynamic management notes that the effectiveness of vasopressors depends on adequate fluid management (Kotani and colleagues, 2025).

The table summarizes the 2026 sequence for the man in the scene.

Timing	2026 approach	Strength	Source
First hour	Blood culture, lactate, antimicrobial and start of balanced crystalloid	Strong for blood culture and antimicrobial	(2026 recommendations on the campaign page)
Unstable shock	Consider norepinephrine together with the fluid, through a peripheral vein if needed	Conditional	(2026 recommendations on the campaign page) (2026 quick resuscitation guide)
Up to three hours	At least 30 mL/kg of crystalloid, with reassessment	Conditional	(2026 recommendations on the campaign page)
Target	Mean arterial pressure of 60 to 65 mmHg, because the patient is 65 or older	Conditional	(2026 recommendations on the campaign page)
Norepinephrine rising	Add vasopressin and then epinephrine	Conditional	(2026 recommendations on the campaign page)
Up to six hours	Source control and admission to intensive care	Conditional	(2026 recommendations on the campaign page)

Corticosteroid, lungs and care after the acute phase

Before considering the corticosteroid

• Confirm that the shock persists despite adequate fluid.
• Check the norepinephrine dose and whether it is rising.
• Review whether the source has been controlled.
• Think about other causes of shock.
• Consider an arterial line if the vasopressor dose is moderate to high or increasing.
• Combine glycemic control, because the corticosteroid can raise blood glucose.

Corticosteroid

In 2021, the suggestion was intravenous corticosteroid in patients with a continuous need for a vasopressor, with moderate evidence (Evans and colleagues, 2021). The remark cited hydrocortisone 200 mg per day, as 50 mg every six hours or continuous infusion, starting with norepinephrine or epinephrine at 0.25 μg/kg/min or more for at least four hours (Evans and colleagues, 2021). In 2026, the suggestion is intravenous corticosteroid in septic shock, now with low certainty (2026 recommendations on the campaign page). The 2026 quick guide mentions hydrocortisone, with or without fludrocortisone, if the need for norepinephrine persists (2026 quick resuscitation guide).

Oxygen and ventilation

In 2021, there was not enough evidence to recommend conservative oxygen targets (Evans and colleagues, 2021). In 2026, the suggestion is to adjust the oxygen fraction to more liberal or more conservative targets according to the patient and the resources, and the panel reports a saturation target between 90 and 96 percent in practice (2026 recommendations on the campaign page). The guideline suggests high-flow nasal cannula instead of conventional oxygen when the ratio between arterial oxygen pressure and inspired fraction is below 200, and as initial therapy instead of noninvasive ventilation (2026 recommendations on the campaign page). It suggests awake prone positioning in patients who are not intubated (2026 recommendations on the campaign page). In acute respiratory distress syndrome, the recommendation remains a tidal volume of 6 mL/kg and a plateau pressure of up to 30 cmH2O (2026 recommendations on the campaign page).

Other points that changed or stayed the same

Insulin still starts with blood glucose of 180 mg/dL or more, as in 2021 (2026 recommendations on the campaign page) (Evans and colleagues, 2021). The 2026 guideline suggests not using antipyretics to improve outcomes, suggests active fluid removal after the acute phase, recommends a restrictive transfusion strategy and suggests enteral nutrition in the first 72 hours (2026 recommendations on the campaign page). It suggests discussing goals of care in the first 72 hours (2026 recommendations on the campaign page). On methylene blue in refractory shock, there is not enough evidence to recommend it (2026 recommendations on the campaign page).

After intensive care

The 2026 guideline suggests a transition program on discharge from intensive care, structured handover and medication reconciliation with a pharmacist (2026 recommendations on the campaign page). It suggests offering follow-up services after critical illness, physical rehabilitation for those who spent more than 48 hours on mechanical ventilation and mental health support (2026 recommendations on the campaign page). The post about diagnostic pause helps to stop before discharge and review what is still missing.

How NEXO helps train this reasoning

On NEXO, the case of the day is fictional, with a real clinical presentation and up to six progressive clues, and you propose the diagnosis by looking for the ICD-10 code. Sepsis is a good exercise because a lactate, a pressure or a CT scan change the approach, and time counts. The post about learning based on clinical cases explains why this format fixes knowledge.

Three exercises with cases

Change the shock. Say what changes if the pressure were 112/70 mmHg, with no mottled skin, with moderate suspicion of urinary infection.

Change the age. Say what the pressure target would be if he were 52 years old.

Change the organ. Say what changes in the fluid if he had cirrhosis and had already received large volumes.

The post about simulation of care helps train the first hour as a team.

The limits of use of the game are described on the NEXO Economy page. If you research teaching of clinical reasoning, the page for Researcher explains how to get in touch. Professionals who write or review cases find information on Curator and on Creators, and institutions can learn about the proposal on the partnerships page.

Frequently asked questions

Is the 30 mL/kg rule over?

No. The 2026 guideline still suggests at least 30 mL/kg in the first three hours, with frequent reassessment (2026 recommendations on the campaign page).

Is the antimicrobial still within the first hour?

Yes in shock and in probable or definite sepsis. In possible sepsis without shock, up to three hours (2026 antimicrobial timing framework).

What is the pressure target for older adults?

The suggestion is 60 to 65 mmHg for those aged 65 or older (2026 recommendations on the campaign page).

Is albumin still used?

The suggestion is crystalloid alone, with albumin possible after large volumes or in cirrhosis (2026 recommendations on the campaign page).

Does SOFA-2 replace SOFA?

It updates the score, and a large study suggests using both during the transition (Zhu and colleagues, 2026).

Final checklist

• I recognized the shock by physical examination.
• I collected blood culture and measured lactate.
• I gave the antimicrobial within the first hour.
• I calculated the fluid by weight.
• I started norepinephrine without waiting for central access.
• I adjusted the pressure target to age.
• I called surgery to control the source.
• I planned de-escalation and follow-up after discharge.

Want to practice reasoning with progressive clues? Play the case of the day.

References

1. Evans L, Rhodes A, Alhazzani W, et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021. PMID 34599691.
2. Kim HJ, Ko RE, Lim SY, et al. Sepsis Alert Systems, Mortality, and Adherence in Emergency Departments: A Systematic Review and Meta-Analysis. JAMA Netw Open. 2024. PMID 39037814.
3. Kotani Y, Ryan N, Udy AA, et al. Haemodynamic management of septic shock. Burns Trauma. 2025. PMID 39816212.
4. Lobo SM, Roepke RML, Myatra SN, et al. Using the SOFA 2.0 score: a quick guide for clinicians and researchers. Crit Care Sci. 2025. PMID 41379119.
5. Moreno R, Rhodes A, Piquilloud L, et al. The Sequential Organ Failure Assessment (SOFA) Score: has the time come for an update? Crit Care. 2023. PMID 36639780.
6. Munroe ES, Co IN, Douglas I, et al. Peripheral Vasopressor Use in Early Sepsis-Induced Hypotension. JAMA Netw Open. 2025. PMID 40864467.
7. Owen VS, Rosgen BK, Cherak SJ, et al. Adverse events associated with administration of vasopressor medications through a peripheral intravenous catheter: a systematic review and meta-analysis. Crit Care. 2021. PMID 33863361.
8. Prescott HC, Ostermann M. What is new and different in the 2021 Surviving Sepsis Campaign guidelines. Med Klin Intensivmed Notfmed. 2023. PMID 37286842.
9. Serova V, Klibus M, Marcinkevics Z, et al. Microcirculation Monitoring in Septic Shock: Focused Review. Medicina (Kaunas). 2026. PMID 41752745.
10. Society of Critical Care Medicine. Antibiotic Timing Framework for Adults. Surviving Sepsis Campaign, 2026. Accessed on 24/10/2026. Official source.
11. Society of Critical Care Medicine. Quick Guide for Resuscitation and Hemodynamic Support in Adults With Sepsis and Septic Shock. Surviving Sepsis Campaign, 2026. Accessed on 24/10/2026. Official source.
12. Society of Critical Care Medicine. Surviving Sepsis Campaign Adult Guidelines | SCCM. Accessed on 10/24/2026. Official source.
13. Wei S, Wang T, Li C, et al. SOFA-2 score predicts mortality in pneumonia-associated sepsis: a retrospective cohort study. Crit Care. 2026. PMID 42045961.
14. Zhu H, Li P, Wang B, et al. Impact of the updated SOFA-2 score on sepsis diagnosis and prognosis: a retrospective multicenter cohort study. Crit Care. 2026. PMID 42092945.