Introduction

Urgency and uncertainty frequently coexist in the management of adult sepsis. Patients with vasopressor-dependent septic shock may experience clinical decline if effective antimicrobial therapy or source control is delayed. Conversely, hemodynamically stable patients with noninfectious sepsis mimics may be adversely affected by unnecessary antimicrobials, fluid administration, invasive monitoring, or diagnostic anchoring. [1,6,7]

These presentations should not be treated as equivalent merely because both trigger an electronic alert, satisfy systemic inflammatory response criteria, or have an elevated lactate concentration. Sepsis is a clinical syndrome of infection-associated organ dysfunction, not a diagnosis established by a score or laboratory result. [1,2]

The 2026 Surviving Sepsis Campaign formalizes this distinction by separating possible sepsis without shock from probable or definite sepsis and septic shock. The guideline preserves immediate antimicrobial treatment for the highest-risk presentations while allowing a brief, structured diagnostic assessment for selected patients with possible sepsis and no shock. [1]

The central clinical query is not solely whether sepsis should be treated early, but rather how to maintain rapid intervention for patients most likely to benefit while limiting unnecessary or poorly targeted treatment when diagnostic or physiologic uncertainty persists.

Why the Topic Matters

Sepsis programs have made delayed recognition and treatment visible clinical failures. Standardized protocols can improve reliability, but they can also encourage clinicians and institutions to treat process completion as the principal outcome.

The Centers for Medicare & Medicaid Services currently includes the Severe Sepsis and Septic Shock: Management Bundle, commonly called SEP-1, in the Hospital Value-Based Purchasing Safety Domain for fiscal years 2026-2031. SEP-1 therefore remains operationally and financially consequential for United States hospitals. [4]

A joint position paper from infectious diseases, emergency medicine, hospital medicine, pediatric infectious diseases, healthcare epidemiology, and antimicrobial-stewardship societies concluded that SEP-1 implementation increased measured processes such as lactate testing, broad-spectrum antimicrobial use, and standardized fluid administration without consistent evidence of reduced mortality. The authors recommended replacing the all-or-nothing measure with more clinically meaningful and electronically measurable outcomes. [5]

That position does not establish that individual bundle elements are ineffective. It shows that more complete process documentation does not necessarily produce proportional improvement in patient-centered outcomes.

A 2026 multicenter observational study proposed the concept of sepsis diagnostic excellence, balancing rapid recognition of true sepsis against avoidance of unnecessary treatment in patients without sepsis. Hospitals with stronger balanced diagnostic performance had lower adjusted odds of death or hospice discharge, although the hospital-level observational design cannot establish causality. [20]

The practical objective is not to eliminate protocols, but to integrate diagnostic revision, source-control assessment, individualized resuscitation, antimicrobial de-escalation, and treatment-safety review within the protocol framework.

Background and Pathophysiologic Framework

Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection. Under the Sepsis-3 framework, acute organ dysfunction is operationalized as an increase of at least 2 points in the Sequential Organ Failure Assessment score. Septic shock is clinically identified by vasopressor dependence to maintain a mean arterial pressure of at least 65 mm Hg and a serum lactate concentration greater than 2 mmol/L despite adequate volume resuscitation. [2]

These definitions characterize severity and prognosis but do not establish infection by themselves. Acute organ-system dysfunction can also accompany hemorrhage, cardiogenic shock, pulmonary embolism, pancreatitis, adrenal crisis, medication toxicity, severe inflammatory disease, or other noninfectious syndromes.

Moreover, sepsis is heterogeneous. The original SENECA analysis derived four clinical phenotypes with different patterns of illness severity, inflammation, chronic disease, renal dysfunction, pulmonary dysfunction, hepatic dysfunction, and shock. [17]

A 2026 external validation across independent European cohorts found limited consistency in the prevalence and characteristics of the original phenotypes. This weakens the premise that the four-category SENECA taxonomy is broadly generalizable. Clinical phenotyping remains a research strategy rather than a validated method for assigning routine bedside treatment. [19]

In current clinical practice, sepsis is most appropriately conceptualized as a time-dependent working diagnosis linked to a dynamic physiologic state. Infection probability, organ dysfunction, perfusion, treatment response, fluid tolerance, source-control requirements, and alternative diagnoses should be continually reassessed as new information becomes available.

Where Undertreatment Occurs

Delayed Antimicrobial Therapy in Septic Shock

The strongest rationale for immediate antimicrobial therapy applies to septic shock. The 2026 guideline strongly recommends administering antimicrobials immediately, ideally within 1 hour, in patients with possible, probable, or definite septic shock. The certainty supporting the timing threshold is very low, and the recommendation should not be interpreted as proof of a uniform causal mortality increase for each hour of delay in every sepsis presentation. [1]

In a retrospective analysis of 49,331 adults treated under New York State sepsis regulations, each additional hour to completion of the initial bundle was associated with greater risk-adjusted in-hospital mortality. Longer time to antimicrobial administration showed a similar association, whereas time to completion of the initial fluid bolus did not show the same independent relationship. Residual confounding remains possible because recognition time, severity, diagnostic certainty, and treatment selection were not randomized. [7]

The study supports avoiding preventable antimicrobial delays in cases of shock and high-probability sepsis. It does not demonstrate that all hemodynamically stable patients with possible infection derive benefit from identical timing of antimicrobial administration.

Delayed Source Control

Antimicrobials cannot substitute for drainage, debridement, removal of an infected device, relief of an infected obstruction, or correction of a perforated or ischemic source.

The 2026 guideline recommends rapid identification of an anatomical source requiring intervention and conditionally suggests early source control rather than delayed intervention, ideally within 6 hours when feasible. The certainty of evidence is very low. [1]

In an observational cohort of 4,962 adults with sepsis who underwent source-control procedures, intervention within 6 hours was associated with lower adjusted odds of 90-day mortality than intervention performed at 6-36 hours. The adjusted odds ratio was 0.71, with a 95% confidence interval of 0.63-0.80. The association varied by anatomical source and procedure type and cannot prove that the timing difference itself caused the outcome difference. [8]

Source-control planning should be initiated promptly and conducted in parallel with antimicrobial and hemodynamic management when an actionable source is suspected.

Failure to Recognize Deterioration

Sepsis can evolve across the emergency department, medical ward, procedural suite, and intensive care unit. A single reassuring blood pressure reading, a transient response to fluid, a low, quick Sequential Organ Failure Assessment score, or a normal initial lactate does not establish clinical stability.

Repeated assessment is particularly important when mental status, respiratory function, urine output, skin perfusion, blood pressure, lactate, vasopressor requirements, or other organ-function measures worsen. Clinical trajectory is more informative than a one-time designation of “screen positive” or “screen negative.”

Where Overtreatment Occurs

Treating Possible Sepsis as Definite Infection

The 2026 guideline does not impose the same antimicrobial timetable on every patient with likely sepsis.

For probable or definite sepsis without shock, the guideline strongly recommends immediate antimicrobial therapy, ideally within 1 hour, although the supporting certainty is very low. For possible sepsis without shock, it conditionally supports a rapid, time-limited investigation and antimicrobial administration within 3 hours if concern for infection persists. When shock is absent and infection is unlikely, antimicrobials may be deferred with close monitoring. [1]

This distinction matters because suspected sepsis is not synonymous with confirmed bacterial infection. In a 2025 multicenter retrospective cohort of 600 adults treated for suspected sepsis in seven emergency departments, structured adjudication classified 189 patients, or 31.5%, as having possible but less likely bacterial infection or definitely no bacterial infection. [6]

Retrospective adjudication benefits from information unavailable to the initial clinician. The study should not be interpreted as proof that the initial empiric decisions were inappropriate. It instead demonstrates the need for a formal diagnostic checkpoint after early test results and treatment response become available.

Excessively Broad Antimicrobial Therapy

Empiric treatment should cover likely pathogens according to the suspected source, severity, previous microbiology, recent healthcare exposure, local resistance patterns, immune status, allergy history, and risk of resistant organisms.

The 2026 guideline recommends empiric multidrug-resistant organism coverage when risk is high and advises against routine resistant-organism coverage when risk is low. It also recommends antimicrobial de-escalation when microbiologic results identify the pathogen and conditionally supports de-escalation when clinical improvement occurs even if no pathogen is identified. [1]

In the 2025 suspected-sepsis cohort, 325 of 411 patients adjudicated as having definite or probable bacterial infection, or 79.1%, received a regimen considered excessively broad when assessed retrospectively. Potential antimicrobial complications occurred in 104 of 600 patients, or 17.3%, within 90 days. Resistant infection or colonization occurred in 48 patients, or 8.0%. These outcomes were observational and cannot all be attributed directly to antimicrobial exposure. [6]

The findings reinforce the need to reassess indication, spectrum, route, dose, microbiologic data, source control, and planned duration. They should not be used to criticize appropriately cautious initial empiric treatment.

Continuing Therapy After the Diagnosis Changes

A frequent stewardship failure is not the first antimicrobial dose but failure to stop or narrow treatment after the working diagnosis changes.

The guideline recommends reassessing and stopping empiric antimicrobials when a noninfectious diagnosis becomes more likely. It also favours a shorter duration of therapy rather than a longer one when source control is adequate and clinical response allows. [1]

An initial sepsis diagnosis should therefore remain revisable. A full antimicrobial course should not continue solely because treatment was started under uncertainty.

Treating Lactate as a Fluid Prescription

Lactate measurement provides prognostic and physiologic information, but an elevated value is not a direct prescription for repeated fluid administration.

The 2026 guideline conditionally suggests serial lactate measurements during resuscitation. It explicitly cautions that fluid should not be continued merely until lactate normalizes. [1]

Persistent lactate elevation may represent ongoing hypoperfusion, adrenergic stimulation, impaired clearance, seizures, medication effects, or other mechanisms. Fluid decisions should incorporate the full clinical state rather than the lactate value alone.

Are We Targeting the Wrong Metrics?

qSOFA as a Screening Gate

The quick Sequential Organ Failure Assessment was developed as a parsimonious prognostic tool rather than a definitive diagnostic test.

A 2023 meta-analysis of 57 studies reported the following pooled sensitivity and specificity estimates for sepsis identification:

• SIRS: 0.85 and 0.41
• qSOFA: 0.42 and 0.98
• NEWS: 0.71 and 0.85

The high specificity of qSOFA identifies a concerning subgroup, but its low pooled sensitivity makes it unsuitable as the sole method for excluding sepsis. [3]

The 2026 guideline strongly recommends NEWS, NEWS2, MEWS, or SIRS over qSOFA when one screening tool is used in hospitalized adults. Screening results must still be interpreted alongside infection probability, organ dysfunction, comorbidity, and clinical trajectory. [1]

Lactate as a Resuscitation Endpoint

The ANDROMEDA-SHOCK trial randomized 424 adults with early septic shock to a resuscitation strategy targeting capillary refill time or one targeting lactate reduction.

Twenty-eight-day mortality was 34.9% in the peripheral-perfusion group and 43.4% in the lactate-targeted group. The hazard ratio was 0.75, with a 95% confidence interval of 0.55-1.02 and a P value of .06. The primary mortality comparison was not statistically significant. The peripheral-perfusion group had a modestly lower mean Sequential Organ Failure Assessment score at 72 hours. [12]

The trial does not establish that capillary refill should replace lactate. It supports using capillary refill as a rapid adjunct and avoiding laboratory normalization as the sole resuscitation objective.

Fluid Volume as a Quality Metric

Fluid volume is an intervention, not a direct measure of successful resuscitation.

The 2026 guideline conditionally suggests at least 30 mL/kg of intravenous crystalloid within the first 3 hours for sepsis-induced hypoperfusion or septic shock. The recommendation is supported by low-certainty evidence and includes an explicit requirement for individualization and frequent reassessment. Actual body weight is generally used, although adjusted or ideal body weight may be considered in patients with a body mass index above 30 kg/m². [1]

The recommendation should not become an automatic instruction to continue fluid despite worsening oxygenation, venous congestion, pulmonary edema, or absent hemodynamic response.

Three major randomized trials inform treatment after initial recognition:

• CLOVERS: Among 1,563 adults with sepsis-induced hypotension who had already received 1-3 liters of fluid, death before discharge home by day 90 occurred in 14.0% under a restrictive strategy prioritizing vasopressors and 14.9% under a liberal-fluid strategy. The difference was not statistically significant. [9]
• CLASSIC: Among 1,554 adults with septic shock who had already received at least 1 liter of fluid, 90-day mortality was 42.3% with a restrictive strategy and 42.1% with standard fluid management. The difference was not statistically significant. [10]
• ARISE FLUIDS: In the 2026 trial, 963 participants were included in the intention-to-treat analysis comparing an early vasopressor and restricted-fluid strategy with a higher-fluid and later-vasopressor strategy. Median days alive and out of hospital through day 90 were 76 days in both groups. Pulmonary edema occurred less frequently with the early vasopressor strategy, but the primary clinical outcome did not differ. [18]

These trials do not establish universal equivalence between every fluid and vasopressor protocol. They show that no tested routine strategy was superior across the full enrolled population and support repeated assessment of perfusion, fluid responsiveness, fluid tolerance, and vasopressor need.

Rigid Physiologic Protocols

The PRISM individual-patient-data meta-analysis combined 3,723 participants from the ProCESS, ARISE, and ProMISe trials.

Ninety-day mortality was 24.9% with protocolized early goal-directed therapy and 25.4% with contemporary usual care. Protocolized treatment increased intensive-care and cardiovascular-support use and increased costs without improving survival. [13]

The findings do not argue against rapid recognition, antimicrobials, source control, or hemodynamic support. They show that a specific invasive, fixed-target protocol did not outperform contemporary clinician-directed care.

Bundle Completion

Bundle adherence may improve reliability, but an all-or-nothing process measure cannot fully represent diagnostic accuracy, treatment appropriateness, contraindications, patient response, or treatment-related harm.

Clinically meaningful measurement should include:

• Time to effective therapy in shock
• Diagnostic accuracy
• Time to indicated source control
• Appropriate initial antimicrobial spectrum
• Timely narrowing or discontinuation
• Organ-support exposure
• Fluid-related complications
• Antimicrobial complications
• Functional recovery
• Mortality and hospice disposition

Process metrics remain useful, but they should not displace outcome and balancing measures.

Patient Selection and Treatment Urgency

Clinical category	Initial antimicrobial approach	Essential reassessment
Possible, probable, or definite septic shock	Administer immediately, ideally within 1 hour. [1]	Confirm source, antimicrobial activity, dose, perfusion response, and source-control needs.
Probable or definite sepsis without shock	Administer immediately, ideally within 1 hour. [1]	Reassess whether organ dysfunction is infection related and narrow treatment when possible.
Possible sepsis without shock	Conduct a rapid, time-limited investigation; treat within 3 hours if infection concern persists. [1]	Monitor trajectory and reconsider noninfectious causes during the diagnostic interval.
Low likelihood of infection without shock	Defer antimicrobials with close monitoring. [1]	Escalate promptly if infection probability or organ dysfunction increases.

This framework should not become a reason to postpone treatment in a deteriorating patient while attempting to achieve diagnostic certainty.

Clinical Presentation

Sepsis should be suspected when a plausible or documented infection is accompanied by acute organ dysfunction.

Clinically important manifestations may include:

• New hypotension or vasopressor requirement
• Acute hypoxemic respiratory dysfunction
• Altered mental status
• Acute kidney injury or oliguria
• Coagulopathy or thrombocytopenia
• Hepatic dysfunction
• Abnormal peripheral perfusion
• Rising lactate in the appropriate clinical setting

No individual manifestation establishes sepsis. Baseline organ function, chronic disease, medications, recent procedures, immune status, and competing diagnoses must be considered.

The absence of fever, leukocytosis, elevated lactate, or a high qSOFA score does not independently exclude clinically important infection-associated organ dysfunction.

Diagnostic Considerations

Estimate Infection Probability

Initial assessment should identify:

• A plausible anatomical source
• Recent healthcare or antimicrobial exposure
• Previous microbiology and resistant organisms
• Immune status
• Recent procedures or indwelling devices
• Epidemiologic exposures
• Major noninfectious alternatives

Possible, probable, and definite infection are clinical categories rather than definitive laboratory states.

Obtain Cultures Without Creating Harmful Delay

The 2026 guideline strongly recommends obtaining blood cultures as soon as possible and ideally before antimicrobials. The supporting certainty is low.

Culture collection should not create a clinically significant delay in a patient with septic shock or high-probability sepsis. Source-directed cultures and rapid diagnostic tests should be obtained when their results can reasonably influence management. [1]

Assess Organ Dysfunction

The Sequential Organ Failure Assessment score organizes respiratory, coagulation, hepatic, cardiovascular, neurologic, and renal dysfunction. Its components should be interpreted relative to baseline because chronic abnormalities may not represent an acute sepsis-related change. [2]

Use Lactate in Context

The guideline conditionally suggests measuring lactate in adults with possible, probable, or definite sepsis or septic shock.

Lactate can help identify risk and monitor trajectory, but it should not be used as a stand-alone diagnostic test or automatic trigger for additional fluid. [1]

Evaluate Source Control Early

Imaging and procedural consultation should be prioritized when the suspected source may require drainage, debridement, device removal, relief of obstruction, or another definitive intervention.

Source-control planning and physiologic stabilization often need to proceed concurrently.

Reopen the Diagnosis

A formal diagnostic time-out should occur after early laboratory, microbiology, imaging, and response data become available.

The clinician should ask:

1. Does infection remain the leading diagnosis?
2. Is the presumed source still plausible?
3. Is the empiric antimicrobial regimen active but broader than necessary?
4. Is source control adequate?
5. Does another diagnosis better explain the syndrome?
6. Can antimicrobial therapy be stopped, narrowed, or shortened?

Therapeutic Considerations

Empiric Antimicrobial Therapy

Initial antimicrobial selection should reflect:

• Suspected source
• Illness severity
• Local susceptibility patterns
• Previous microbiologic results
• Recent antimicrobial exposure
• Healthcare-associated exposure
• Immune status
• Organ function
• Allergy history
• Risk factors for resistant pathogens

Empiric resistant-organism coverage is appropriate when individual and local risk is high. Routine coverage for multidrug-resistant pathogens is not supported when risk is low. [1]

Administration and Pharmacokinetics

The guideline strongly recommends a loading dose followed by prolonged infusion of beta-lactam maintenance therapy when clinically and operationally appropriate. The evidence certainty is moderate.

Therapeutic drug monitoring may be considered in selected patients when altered pharmacokinetics, extracorporeal support, renal replacement therapy, augmented clearance, toxicity risk, or a narrow therapeutic target makes standard dosing unreliable. [1]

Specific antimicrobial doses remain drug-, source-, pathogen-, organ-function-, and institution-dependent and are outside this review’s scope.

De-escalation and Duration

The guideline recommends de-escalation when microbiology identifies the causative pathogen and conditionally supports narrowing in clinically improving patients even when no pathogen is recovered.

Shorter courses are generally preferred when source control is adequate and clinical response supports discontinuation. Duration should remain source-specific rather than determined by the sepsis label alone. [1]

Procalcitonin

Procalcitonin should not replace clinical assessment when deciding whether to start antimicrobials.

The guideline conditionally suggests procalcitonin plus clinical evaluation to support discontinuation when source control is adequate and optimal treatment duration remains unclear. [1]

A 2023 meta-analysis of 26 randomized trials involving 9,048 critically ill patients found that procalcitonin-guided strategies reduced antimicrobial duration by a mean of 1.79 days, with a 95% confidence interval of 0.92-2.65 fewer days. Recurrent infection was more frequent in procalcitonin-guided groups, with an odds ratio of 1.36 and a 95% confidence interval of 1.10-1.68. Protocols, populations, adherence, and risk of bias varied substantially, and immunocompromised patients were commonly excluded. [14]

Procalcitonin should therefore support, rather than replace, clinical judgment.

Initial Fluid Resuscitation

Crystalloids are recommended as first-line resuscitation fluid. Balanced crystalloids are conditionally suggested over saline, with moderate-certainty evidence. Saline is preferred in patients with traumatic brain injury. [1]

In a secondary sepsis analysis of the SMART trial, balanced crystalloids were associated with lower adjusted 30-day in-hospital mortality than saline. Because this was a secondary subgroup analysis from a pragmatic cluster-randomized trial, it should not be interpreted as a definitive sepsis-specific mortality trial. [11]

For sepsis-induced hypoperfusion or septic shock, the guideline conditionally suggests at least 30 mL/kg within the first 3 hours. The volume and rate should be individualized according to perfusion, cardiac function, pulmonary status, kidney function, chronic blood pressure, body habitus, illness severity, and response to each treatment phase. [1]

Fluids After the Initial Phase

Following initial resuscitation, the guideline does not favor a universally liberal or restrictive strategy. Either approach may be reasonable according to the patient’s hemodynamics, comorbidities, fluid tolerance, and available monitoring.

Additional fluid should be considered only when there is a reasonable expectation of improving circulation or organ perfusion without disproportionate risk of congestion.

Dynamic Assessment

Dynamic measures are conditionally preferred over physical examination or static variables alone when estimating whether another fluid bolus will increase cardiac output.

Potential methods include:

• Passive leg raising with measured response
• A small fluid challenge with measured response
• Stroke-volume or pulse-pressure variation in appropriately selected patients
• Echocardiographic assessment when expertise is available

Fluid responsiveness does not automatically establish that fluid is required. A patient may show an increase in cardiac output yet have limited fluid tolerance because of pulmonary edema, right ventricular dysfunction, elevated venous pressure, or tissue congestion.

Vasopressors

Norepinephrine is the recommended first-line vasopressor for septic shock. [1]

The guideline conditionally suggests starting vasopressors through an appropriate peripheral intravenous catheter rather than delaying treatment until central venous access is established. Evidence is insufficient to prescribe one universal peripheral catheter location, maximum dose, or maximum duration. Local protocols and close site monitoring remain necessary.

For patients with persistent hypotension after an initial fluid bolus, vasopressors should be started rather than repeatedly administering fluid without evidence of benefit. In unstable shock, fluid and vasopressors may be initiated concurrently according to the clinical situation.

Blood Pressure Targets

An initial mean arterial pressure target of 65 mm Hg is strongly recommended for most adults with septic shock.

For adults aged 65 years or older, the guideline conditionally suggests an initial range of 60-65 mm Hg rather than routinely targeting a higher pressure. This is not a directive to accept inadequate perfusion. The target should be individualized according to mental status, urine output, skin perfusion, chronic hypertension, cardiac or cerebrovascular disease, arrhythmia risk, and vasopressor toxicity. [1]

Corticosteroids

The 2026 guideline conditionally suggests intravenous corticosteroids in adults with septic shock. The evidence certainty is low. The recommendation applies to established shock rather than uncomplicated sepsis without vasopressor dependence. [1]

In ADRENAL, 3,800 mechanically ventilated adults with septic shock received hydrocortisone 200 mg per day by continuous intravenous infusion or placebo for up to 7 days. Hydrocortisone did not significantly reduce 90-day mortality but accelerated shock resolution. [15]

In APROCCHSS, adults with severe septic shock received hydrocortisone 50 mg intravenously every 6 hours plus fludrocortisone 50 micrograms orally once daily for 7 days or placebo. Ninety-day mortality was 43.0% with corticosteroid therapy and 49.1% with placebo, corresponding to a relative risk of 0.88 and a 95% confidence interval of 0.78-0.99. Hyperglycemia was more frequent with corticosteroid therapy. [16]

These trial regimens are evidence-based septic shock strategies but are not specifically approved by the FDA for septic shock. The guideline does not establish one universally required corticosteroid formulation or regimen for every patient.

Therapies Without Established Routine Benefit

The guideline suggests against routine sepsis-specific use of:

• Intravenous vitamin C
• Intravenous immunoglobulin
• Blood-purification techniques
• Vitamin D

These interventions should not displace effective antimicrobial therapy, source control, appropriate hemodynamic support, or evidence-based organ support. [1]

Safety Considerations

Antimicrobial Safety

Broad empiric coverage may be necessary in septic shock but can contribute to:

• Renal, hepatic, hematologic, neurologic, or cardiac toxicity
• Drug interactions
• Clostridioides difficile infection
• Microbiome disruption
• Resistant-organism selection
• Secondary infection
• Line-related or infusion-related complications

Safety assessment should include organ function, allergy history, concomitant medications, electrocardiographic risk when relevant, therapeutic drug monitoring requirements, and cumulative duration.

Fluid Safety

Potential consequences of excessive fluid include:

• Pulmonary edema
• Impaired gas exchange
• Venous congestion
• Tissue edema
• Delayed de-resuscitation
• Worsening right ventricular loading conditions

Potential consequences of inadequate resuscitation include persistent hypoperfusion, worsening organ dysfunction, and delayed shock reversal.

The relevant question is not whether fluid is inherently beneficial or harmful. It is whether the next increment is likely to improve perfusion more than it increases congestion or other complications.

Vasopressor Safety

Vasopressor therapy requires monitoring for:

• Excessive vasoconstriction
• Digital, mesenteric, or other ischemia
• Arrhythmia
• Extravasation
• Inadequate cardiac output despite an acceptable arterial pressure

Peripheral administration requires a functioning catheter, appropriate site selection under local protocol, frequent inspection, and rapid response to suspected extravasation.

Corticosteroid Regulatory and Safety Considerations

Hydrocortisone sodium succinate labeling does not include a specific septic shock indication. Current labeling identifies systemic fungal infection and hypersensitivity to the product or its components as contraindications. It includes warnings related to infection risk, endocrine effects, gastrointestinal complications, psychiatric effects, fluid and electrolyte disturbances, and hyperglycemia. [21]

Fludrocortisone labeling covers mineralocorticoid replacement indications rather than septic shock. Its mineralocorticoid activity can contribute to sodium retention, edema, hypertension, hypokalemia, and fluid imbalance. Blood pressure, volume status, glucose, and electrolytes should be monitored when it is used. [22]

The absence of a specific FDA septic shock indication does not make the trial-supported regimens erroneous. It means they should be recognized as off-label uses informed by clinical evidence and professional guidelines.

Diagnostic Safety

Premature closure on sepsis can delay recognition of pulmonary embolism, hemorrhage, cardiogenic shock, adrenal crisis, medication toxicity, pancreatitis, or another noninfectious syndrome.

The opposing error is withholding treatment from a patient with evolving infection because early cultures, imaging, or biomarkers are nondiagnostic.

Diagnostic safety depends on repeated reassessment rather than confidence in the initial label.

Special Populations

Older Adults

Older adults may have chronic organ dysfunction, frailty, reduced physiologic reserve, polypharmacy, and greater susceptibility to treatment complications.

The guideline’s suggested initial mean arterial pressure range of 60-65 mm Hg for adults aged 65 years or older is intended to avoid unnecessary exposure to higher vasopressor targets. Perfusion must still be assessed individually.

Heart Failure

Heart failure does not automatically preclude initial sepsis resuscitation. It does strengthen the rationale for:

• Smaller, reassessed fluid increments
• Dynamic testing
• Early vasopressor consideration when appropriate
• Evaluation of pulmonary edema and venous congestion
• Bedside cardiac assessment when available

The presence of heart failure should prompt individualization, not automatic fluid withholding or automatic administration of a fixed volume without reassessment.

Chronic Kidney Disease

Chronic kidney disease complicates interpretation of creatinine-based organ dysfunction and increases concern about volume accumulation.

Neither chronic kidney disease nor dialysis dependence proves that all fluid is harmful. Both increase the importance of dynamic assessment, fluid-tolerance evaluation, medication-dose review, and early recognition of persistent shock.

Traumatic Brain Injury

Balanced crystalloids are generally conditionally favored in sepsis, but the 2026 guideline favors saline in patients with traumatic brain injury. This exception should be recognized when choosing resuscitation fluid. [1]

Immunocompromised Patients

Immunocompromised patients may have atypical presentations, broader pathogen possibilities, and greater consequences from delayed effective treatment.

Evidence for procalcitonin discontinuation is less generalizable to this population because immunocompromised patients were commonly excluded from randomized studies. [14]

Chronic Organ Dysfunction

SOFA abnormalities must be interpreted against baseline in patients with chronic renal, hepatic, respiratory, neurologic, or hematologic disease. A chronic abnormality should not automatically be attributed to sepsis, while a relatively small acute change may still be clinically significant in a patient with limited reserve.

Practical Clinical Approach

1. Identify Immediate Instability

Determine whether the patient has septic shock, rapidly progressive organ dysfunction, respiratory failure, altered mental status, or another indication for immediate escalation.

2. Estimate Infection Probability

Classify the syndrome as possible, probable, or definite infection while documenting the suspected source and important noninfectious alternatives.

3. Match Antimicrobial Urgency to Risk

Treat shock and high-probability sepsis immediately. Use a short, structured diagnostic interval only in selected patients without shock when infection remains uncertain and close monitoring is available.

4. Obtain Cultures and Source-Directed Testing

Collect blood cultures and appropriate source specimens as soon as possible, ideally before antimicrobials, without delaying treatment in shock.

5. Select Source- and Risk-Appropriate Empiric Therapy

Choose the initial regimen according to source, severity, prior microbiology, local resistance, immune status, organ function, and resistant-organism risk.

6. Evaluate Source Control Immediately

Determine whether drainage, debridement, device removal, relief of obstruction, or another intervention is required.

7. Resuscitate and Reassess

Use crystalloid for initial resuscitation, generally favoring balanced solutions except in traumatic brain injury. Reassess perfusion and fluid tolerance after each phase.

8. Start Vasopressors When Indicated

Use norepinephrine as first-line therapy when hypotension persists, or additional fluid is unlikely to produce net benefit.

9. Integrate Multiple Perfusion Measures

Use blood pressure, mental status, urine output, skin perfusion, capillary refill, lactate trajectory, cardiac function, dynamic responsiveness, congestion, and organ-function trends.

10. Perform a Mandatory Diagnostic and Therapeutic Time-Out

Reassess:

• Infection probability
• Source
• Antimicrobial activity and spectrum
• Microbiology
• Source control
• Fluid balance
• Vasopressor need
• Corticosteroid indication
• Competing diagnoses
• Stop date or de-escalation plan

What to Measure Instead of a Single Sepsis Endpoint

Limited metric	Why it can mislead	More useful clinical frame
qSOFA alone	High specificity but low pooled sensitivity for sepsis identification. [3]	Combine a validated warning score with infection probability, organ dysfunction, and trajectory.
Lactate normalization	Persistent elevation does not establish that additional fluid will improve perfusion.	Interpret the trend with capillary refill, hemodynamics, organ function, and fluid tolerance.
Total fluid administered	Volume delivered does not establish improved perfusion or appropriate resuscitation.	Measure response to each treatment phase and evidence of congestion or persistent hypoperfusion.
Mean arterial pressure alone	An acceptable pressure may coexist with impaired perfusion or excessive vasoconstriction.	Use pressure as one component of a broader perfusion assessment.
Initial antimicrobial breadth	Appropriate empiric coverage may become unnecessarily broad after clarification.	Measure time to effective therapy and time to appropriate narrowing or discontinuation.
Bundle completion	Process adherence may not capture diagnostic accuracy, contraindications, treatment harm, or longitudinal care.	Pair process measures with source-control, organ-support, stewardship, adverse-event, and outcome measures.

Abbreviations: qSOFA, quick Sequential Organ Failure Assessment.

Clinical Implications

Contemporary sepsis care should be fastest where the expected benefit of immediate intervention is greatest. This includes septic shock, convincing infection-associated organ dysfunction, and rapid clinical deterioration.

The same pathway should preserve diagnostic discretion when shock is absent, and infection is only one of several plausible explanations. That discretion must be time-limited, documented, and supported by close monitoring.

Initial treatment should be viewed as a reversible working plan rather than a permanent commitment to a full antimicrobial course, fixed spectrum, predetermined fluid volume, or isolated physiologic target.

Institutional programs should evaluate not only whether initial processes were completed but also whether:

• True sepsis was recognized promptly
• Noninfectious syndromes were identified
• Effective therapy was delivered in shock
• Source control occurred when needed
• Antimicrobials were narrowed or stopped
• Fluid and vasopressor treatment matched the patient’s physiology
• Treatment complications were minimized
• Patient-centered outcomes improved

Limitations of the Evidence

Many 2026 Surviving Sepsis Campaign recommendations are conditional or supported by low- or very low-certainty evidence. Recommendation strength should not be confused with precise knowledge of the optimal intervention for every patient.

Evidence linking antimicrobial delay to mortality is largely observational and may be affected by illness severity, recognition time, treatment selection, and diagnostic accuracy. These limitations are particularly relevant when extrapolating findings from septic shock to stable patients with uncertain infection.

Source-control timing evidence is also observational. Earlier intervention may identify hospitals and patients with faster recognition, better access to procedures, or more treatable sources.

CLOVERS, CLASSIC, and ARISE FLUIDS compared strategies rather than proving that all liberal, restrictive, or early vasopressor approaches are interchangeable. CLOVERS and CLASSIC enrolled patients after initial fluid had already been administered.

The balanced-crystalloid evidence includes a secondary sepsis subgroup analysis rather than a definitive sepsis-only mortality trial.

ANDROMEDA-SHOCK did not reach statistical significance for its primary mortality outcome.

Antimicrobial-overtreatment estimates relied on retrospective adjudication and should not be used to judge initial decisions with hindsight.

Procalcitonin protocols were heterogeneous, and immunocompromised patients were frequently excluded.

SEP-1 studies evaluate a complex quality program and cannot isolate the causal contribution of each component.

The original SENECA phenotype taxonomy showed limited reproducibility in a 2026 external validation study and should not be used to assign treatment.

Future Directions

Subsequent research should prioritize prospective identification of treatment-responsive states rather than repeatedly estimating average treatment effects across heterogeneous sepsis populations.

Important priorities include:

• Rapid methods for distinguishing bacterial infection from noninfectious syndromes
• Individualized integration of fluid responsiveness and fluid tolerance
• Identification of patients who benefit from earlier vasopressors
• Better evidence in heart failure, chronic kidney disease, frailty, and immune compromise
• Prospective testing of reproducible sepsis subtypes
• Outcome-focused quality measures
• Balanced measures that capture both delayed and unnecessary treatment
• Reliable electronic identification without excessive alert burden

Phenotype models, biomarkers, and artificial-intelligence systems should not direct treatment until prospective studies demonstrate that their use improves clinically meaningful outcomes.

 

Article Type:

Evidence-Based Clinical Review

Target Audience:

Physicians, clinical pharmacists, advanced practice clinicians, nurses involved in acute and critical care, researchers, and medically sophisticated healthcare readers

Estimated Reading Time:

Approximately 28 minutes at 200 words per minute, excluding references and publication metadata

References

1. Prescott, H. C., Antonelli, M., Alhazzani, W., et al. (2026). Surviving Sepsis Campaign: International guidelines for management of sepsis and septic shock 2026. Critical Care Medicine, 54(4), 725-812. https://doi.org/10.1097/CCM.0000000000007075 PMID: 41869847.
2. Singer, M., Deutschman, C. S., Seymour, C. W., et al. (2016). The Third International Consensus Definitions for Sepsis and Septic Shock. JAMA, 315(8), 801-810. https://doi.org/10.1001/jama.2016.0287 PMID: 26903338.
3. Qiu, X., Lei, Y. P., & Zhou, R. X. (2023). SIRS, SOFA, qSOFA, and NEWS in the diagnosis of sepsis and prediction of adverse outcomes: A systematic review and meta-analysis. Expert Review of Anti-infective Therapy, 21(8), 891-900. https://doi.org/10.1080/14787210.2023.2237192 PMID: 37450490.
4. Centers for Medicare & Medicaid Services. (2026). Hospital Value-Based Purchasing Program: Fiscal year 2026-2031 measures. Accessed August 2, 2026. https://qualitynet.cms.gov/inpatient/hvbp/measures
5. Rhee, C., Strich, J. R., Chiotos, K., et al. (2024). Improving sepsis outcomes in the era of pay-for-performance and electronic quality measures: A joint IDSA/ACEP/PIDS/SHEA/SHM/SIDP position paper. Clinical Infectious Diseases, 78(3), 505-513. https://doi.org/10.1093/cid/ciad447
6. Shappell, C. N., Yu, T., Klompas, M., et al. (2025). Frequency of antibiotic overtreatment and associated harms in patients presenting with suspected sepsis to the emergency department: A retrospective cohort study. Clinical Infectious Diseases, 80(6), 1197-1207. https://doi.org/10.1093/cid/ciaf118 PMID: 40231968.
7. Seymour, C. W., Gesten, F., Prescott, H. C., et al. (2017). Time to treatment and mortality during mandated emergency care for sepsis. New England Journal of Medicine, 376(23), 2235-2244. https://doi.org/10.1056/NEJMoa1703058 PMID: 28528569.
8. Reitz, K. M., Kennedy, J., Li, S. R., et al. (2022). Association between time to source control in sepsis and 90-day mortality. JAMA Surgery, 157(9), 817-826. https://doi.org/10.1001/jamasurg.2022.2761 PMID: 35830181.
9. Shapiro, N. I., Douglas, I. S., Brower, R. G., et al. (2023). Early restrictive or liberal fluid management for sepsis-induced hypotension. New England Journal of Medicine, 388(6), 499-510. https://doi.org/10.1056/NEJMoa2212663 PMID: 36688507.
10. Meyhoff, T. S., Hjortrup, P. B., Wetterslev, J., et al. (2022). Restriction of intravenous fluid in ICU patients with septic shock. New England Journal of Medicine, 386(26), 2459-2470. https://doi.org/10.1056/NEJMoa2202707 PMID: 35709019.
11. Brown, R. M., Wang, L., Coston, T. D., et al. (2019). Balanced crystalloids versus saline in sepsis: A secondary analysis of the SMART clinical trial. American Journal of Respiratory and Critical Care Medicine, 200(12), 1487-1495. https://doi.org/10.1164/rccm.201903-0557OC PMID: 31454263.
12. Hernández, G., Ospina-Tascón, G. A., Damiani, L. P., et al. (2019). Effect of a resuscitation strategy targeting peripheral perfusion status versus serum lactate levels on 28-day mortality among patients with septic shock: The ANDROMEDA-SHOCK randomized clinical trial. JAMA, 321(7), 654-664. https://doi.org/10.1001/jama.2019.0071 PMID: 30772908.
13. PRISM Investigators. (2017). Early, goal-directed therapy for septic shock: A patient-level meta-analysis. New England Journal of Medicine, 376(23), 2223-2234. https://doi.org/10.1056/NEJMoa1701380 PMID: 28320242.
14. Papp, M., Kiss, N., Baka, M., et al. (2023). Procalcitonin-guided antibiotic therapy may shorten length of treatment and may improve survival: A systematic review and meta-analysis. Critical Care, 27, 394. https://doi.org/10.1186/s13054-023-04677-2 PMID: 37833778.
15. Venkatesh, B., Finfer, S., Cohen, J., et al. (2018). Adjunctive glucocorticoid therapy in patients with septic shock. New England Journal of Medicine, 378(9), 797-808. https://doi.org/10.1056/NEJMoa1705835 PMID: 29347874.
16. Annane, D., Renault, A., Brun-Buisson, C., et al. (2018). Hydrocortisone plus fludrocortisone for adults with septic shock. New England Journal of Medicine, 378(9), 809-818. https://doi.org/10.1056/NEJMoa1705716 PMID: 29490185.
17. Seymour, C. W., Kennedy, J. N., Wang, S., et al. (2019). Derivation, validation, and potential treatment implications of novel clinical phenotypes for sepsis. JAMA, 321(20), 2003-2017. https://doi.org/10.1001/jama.2019.5791 PMID: 31104070.
18. Peake, S. L., Macdonald, S. P. J., Howe, B. D., et al.; ARISE FLUIDS Investigators, ANZICS Clinical Trials Group, and ACEM Clinical Trials Network. (2026). Vasopressors or fluids in early septic shock. New England Journal of Medicine. Advance online publication. https://doi.org/10.1056/NEJMoa2516225 PMID: 42274006.
19. Yoon, C. H., Sjöholm, D., Wickstrøm, K. E., et al. (2026). External validation of clinical sepsis phenotypes across independent cohorts. JAMA Network Open, 9(6), e2616134. https://doi.org/10.1001/jamanetworkopen.2026.16134 
    PMID: 42223936.
20. Kowalkowski, M., Birken, S., Obermiller, C., et al. (2026). Sepsis diagnostic excellence and hospital outcomes. Clinical Infectious Diseases. Advance online publication, ciag382. https://doi.org/10.1093/cid/ciag382 PMID: 42376964.
21. DailyMed. (2026). Hydrocortisone sodium succinate injection: Current prescribing information. U.S. National Library of Medicine. Accessed August 2, 2026. https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=hydrocortisone%20sodium%20succinate
22. DailyMed. (2026). Fludrocortisone acetate tablets: Current prescribing information. U.S. National Library of Medicine. Accessed August 2, 2026. https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=fludrocortisone%20acetate