Sepsis in the Emergency Department: Recognition and the First Hours

At least one in five adults who develop sepsis does not survive the admission.

Sepsis Before the Blood Pressure Falls

The clinician’s job at the front end is to separate infection from sepsis, and the dividing line is organ dysfunction, not hemodynamics. The Sepsis-3 consensus definition describes sepsis as life-threatening organ dysfunction caused by a dysregulated host response to infection. 

Hypotension appears nowhere in that sentence. A normotensive patient with a rising creatinine and new confusion has sepsis. A patient with a fever and a systolic of 118 who is mottled to the knees has sepsis. Waiting for a number to drop before applying the label is how the diagnosis gets made late.

Compensation is why the blood pressure lags. Young patients maintain perfusion pressure through tachycardia and vasoconstriction until they cannot, then fall off quickly. Patients on beta blockers cannot mount the tachycardia that would otherwise signal the problem. Chronically hypertensive patients who live at 165 systolic are functionally hypotensive at 115, and an absolute threshold misses them entirely.

Screening tools should be chosen with this in mind. The 2026 Surviving Sepsis Campaign international guidelines make a strong recommendation to use NEWS, NEWS2, MEWS, or SIRS criteria rather than qSOFA to screen hospitalized patients. qSOFA is reasonably specific but poorly sensitive, and a tool requiring two of three fairly late findings is a poor instrument for catching illness early. The same guidelines state the principle plainly: sepsis is a clinical diagnosis and should not be ruled in or ruled out using a single biomarker.

In practice, a handful of triage findings should prompt a deliberate second look rather than a routine queue position:

• Respiratory rate above 22, which is the earliest abnormal vital sign in most sepsis presentations and the one most often recorded inaccurately
• Fever or hypothermia, with hypothermia carrying the worse prognosis and drawing the less urgent response
• Tachycardia out of proportion to the fever, or that persists after antipyretics
• Any new alteration in mental status, including the older adult family describes as quieter than usual
• A drop of 40 mm Hg or more from a documented baseline systolic pressure, even if the current value looks acceptable

One set of vital signs at triage is a photograph, not a trajectory. Repeating them within an hour on any patient with suspected infection catches the group whose compensation is failing, while there is room to intervene.

Organ Dysfunction: What to Watch For

Once infection is suspected, the assessment turns to whether any organ system is failing. The findings are ordinary values and bedside observations, and their significance comes from being new rather than dramatic.

• Neurologic: new confusion, agitation, lethargy, or an unexplained fall. In older adults this is frequently the only presenting sign, and it precedes hemodynamic change by hours.
• Respiratory: a rising respiratory rate, a new or increasing oxygen requirement, or a falling ratio of arterial oxygen to inspired oxygen fraction.
• Cardiovascular: absolute or relative hypotension, a narrowing pulse pressure, prolonged capillary refill, cool extremities, or mottling over the knees and shins.
• Renal: a creatinine above baseline or urine output under 0.5 mL/kg/hr, which is easy to miss without a deliberate decision to measure it.
• Hepatic: a new or rising bilirubin, particularly without an obvious biliary source.
• Hematologic: thrombocytopenia, a rising INR without anticoagulation, or laboratory evidence of consumption.

The bedside findings deserve as much weight as the laboratory values. Mottling and prolonged capillary refill reflect the peripheral circulation directly and are abnormal in patients whose central numbers still look acceptable. A patient with a normal blood pressure, cool knees, and a capillary refill of four seconds is not a reassuring patient.

Children compensate more effectively than adults, and hypotension in a child is a late and ominous finding. The 2024 Phoenix criteria for pediatric sepsis and septic shock replaced the older SIRS-based pediatric definitions with a score built on four organ systems: respiratory, cardiovascular, coagulation, and neurologic. A Phoenix Sepsis Score of at least 2 points in a child with suspected infection identifies life-threatening organ dysfunction, and septic shock is the subset with at least 1 cardiovascular point. The practical message is that pediatric sepsis is identified through organ dysfunction and perfusion findings, not a blood pressure threshold that will not be crossed until very late.

Lactate, Cultures, and the First Dose of Antibiotics

Lactate belongs in the initial workup of any patient with suspected sepsis, and the 2026 guidelines suggest both measuring it and using serial values to guide resuscitation. What has changed is the interpretation. Lactate is a marker of illness severity, not a measurement of intravascular volume. Elevations come from impaired hepatic clearance, catecholamine-driven aerobic glycolysis, beta agonist therapy, and metformin as often as from hypoperfusion. The current guidance specifically advises against treating an elevated lactate as an automatic indication for more fluid, a meaningful correction to a decade of reflexive bolusing. Two points follow. A normal lactate does not exclude sepsis. And a lactate that fails to clear on repeat is a reason to reassess the patient and the source, not to hang another liter.

Blood cultures should be drawn before antibiotics whenever obtaining them will not meaningfully delay the first dose. Two sets from separate sites remains the standard, with source-directed cultures added as the presentation dictates. When the patient is unstable and access is difficult, antibiotics come first.

Antibiotic timing is where the 2026 guidance is most useful for emergency practice, because it stratifies rather than issuing one deadline. For probable or definite septic shock, antibiotics should be given immediately, ideally within one hour. For probable sepsis without shock, the recommendation remains within one hour. For possible sepsis without shock, where the diagnosis is genuinely uncertain, the guidance is to investigate rapidly and start antibiotics within three hours if concern persists. That third tier gives defensible room to work up an undifferentiated patient rather than reaching for broad-spectrum coverage on suspicion alone. The hour-1 bundle published in 2018 collapsed the prior 3-hour and 6-hour elements into a single immediate set, and the debate it started about whether the evidence supports that rigidity has not settled.

Documentation matters alongside the clinical decision, since the CMS SEP-1 bundle is the measure most departments are scored against. Treat it as a floor describing adequate care, and record the reasoning whenever a specific patient needs something different.

Empiric coverage is chosen from the suspected source, the local antibiogram, and patient-specific risk for resistant organisms. Doses should be weight-based, and the first dose should generally not be reduced for renal impairment even when subsequent dosing will be.

None of this substitutes for source control. An abscess, an obstructed urinary tract, a necrotizing soft tissue infection, or an infected line will not respond to antibiotics while the source remains in place. Identifying a drainable source and starting the consultation immediately is often the highest-value action of the encounter.

Fluid Resuscitation: How Much, and What Kind

The 30 mL/kg figure traces back to the 2001 early goal-directed therapy trial. The large trials that followed found protocolized care performed no better than usual care, and most of the original targets have been abandoned. The volume figure survived.

The 2026 guidelines carry a conditional recommendation for at least 30 mL/kg of intravenous crystalloid within the first three hours for sepsis-induced hypoperfusion or septic shock. The conditional grading and the acknowledged low certainty of evidence are part of that recommendation and should be read alongside it. The number is a starting point for a typical patient, not a mandate. Thirty milliliters per kilogram in a dialysis-dependent patient with an ejection fraction of 25 percent is a different intervention than the same volume in a previously healthy adult, and judgment about that difference is appropriate rather than deviant.

On fluid choice, the guidance favors balanced crystalloids over 0.9 percent saline. The SMART trial randomized more than 15,000 critically ill adults and found a lower rate of major adverse kidney events at 30 days with balanced crystalloids, with the effect most apparent in patients with sepsis. Traumatic brain injury remains the practical exception, where saline is still preferred.

What happens after the initial bolus was addressed directly by the CLOVERS trial, which randomized patients with sepsis-induced hypotension after 1 to 3 liters of crystalloid to a restrictive fluid strategy with earlier vasopressors or a liberal fluid strategy. The restrictive arm received roughly 2 liters less over 24 hours. Death before discharge home by day 90 occurred in 14.0 percent of the restrictive group and 14.9 percent of the liberal group, and the trial stopped early for futility. The useful reading is not that one strategy won. It is that fluid beyond the initial resuscitation offers no demonstrated benefit, and that starting norepinephrine early rather than continuing to bolus is a safe choice.

After the initial volume, the decision to give more fluid should rest on an assessment rather than a protocol step. A passive leg raise with a measure of stroke volume response, focused cardiac and lung ultrasound, and serial examination of perfusion all inform the question. Static measures such as central venous pressure do not. Equally important is recognizing when enough has been given: worsening oxygenation, new crackles, or increasing work of breathing during resuscitation are signals to stop bolusing and start vasopressors, not to push through.

Septic Shock and When to Escalate

Septic shock is sepsis with persistent hypotension requiring vasopressors to maintain a mean arterial pressure of at least 65 mm Hg, together with a lactate above 2 mmol/L, despite adequate fluid resuscitation. Reported mortality exceeds 40 percent, which sets the urgency for everything that follows.

Norepinephrine is the first-line agent, targeting a mean arterial pressure of 65 mm Hg. Current guidance supports starting it through a peripheral line rather than delaying while central access is obtained, using a proximal large-bore catheter above the wrist with confirmed placement and frequent site checks. Vasopressin is added when the norepinephrine requirement climbs, epinephrine for inadequate response, and hydrocortisone for patients with ongoing vasopressor needs. Throughout, the source question should be revisited, because rising pressor requirements in a patient with an undrained source is a surgical problem, not a pharmacologic one.

Certain findings should prompt a change in level of care rather than another round of the same:

• A norepinephrine requirement that is rising rather than plateauing over the first hour or two
• A lactate that climbs or fails to fall on serial measurement despite adequate resuscitation
• Deteriorating mental status, or a patient whose airway is becoming a question
• A rising oxygen requirement or new hypoxemia during fluid resuscitation
• A source requiring drainage or debridement that has not yet been addressed

Disposition is usually the intensive care unit, and the request should be made early rather than after a period of watchful waiting. Boarding is the reality in most departments, and sepsis care does not pause when the admission order is entered. Serial lactates, reassessment of perfusion, and vasopressor titration remain the responsibility of whoever is at the bedside.

Sepsis Awareness Month and World Sepsis Day on September 13 keep attention on a condition whose outcomes depend heavily on the first few hours, most often in an emergency department. The tools have improved: better screening instruments, clearer stratification of antibiotic urgency, a more honest account of what lactate does and does not tell us, and permission to individualize fluid volume. What has not changed is that the decisive move is still recognition, made early, in a patient who does not yet look as sick as they are. That work happens one patient at a time, on every shift, long after the observance ends.

References

Centers for Disease Control and Prevention. (n.d.). About sepsis. U.S. Department of Health and Human Services. Retrieved August 2026, from https://www.cdc.gov/sepsis/about/index.html

Centers for Disease Control and Prevention. (n.d.). Caring for patients with sepsis. U.S. Department of Health and Human Services. Retrieved August 2026, from https://www.cdc.gov/sepsis/hcp/clinical-care/

Centers for Medicare & Medicaid Services. (2025). Severe sepsis and septic shock: Management bundle (composite measure), version 5.17a. Hospital Inpatient Quality Reporting Program. Retrieved August 2026, from https://www.qualityreportingcenter.com/globalassets/iqr2025-events/iqr040125/april-2025_sep1_npc_vfinal_508.pdf

Evans, L., Rhodes, A., Alhazzani, W., Antonelli, M., Coopersmith, C. M., French, C., … Levy, M. (2021). Surviving Sepsis Campaign: International guidelines for management of sepsis and septic shock 2021. Critical Care Medicine, 49(11), e1063-e1143. https://doi.org/10.1097/CCM.0000000000005337

Levy, M. M., Evans, L. E., & Rhodes, A. (2018). The Surviving Sepsis Campaign bundle: 2018 update. Critical Care Medicine, 46(6), 997-1000. https://doi.org/10.1097/CCM.0000000000003119

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. Advance online publication. https://doi.org/10.1097/CCM.0000000000007075

Rivers, E., Nguyen, B., Havstad, S., Ressler, J., Muzzin, A., Knoblich, B., Peterson, E., & Tomlanovich, M. (2001). Early goal-directed therapy in the treatment of severe sepsis and septic shock. New England Journal of Medicine, 345(19), 1368-1377. https://doi.org/10.1056/NEJMoa010307

Schlapbach, L. J., Watson, R. S., Sorce, L. R., Argent, A. C., Menon, K., Hall, M. W., … Society of Critical Care Medicine Pediatric Sepsis Definition Task Force. (2024). International consensus criteria for pediatric sepsis and septic shock. JAMA, 331(8), 665-674. https://doi.org/10.1001/jama.2024.0179

Semler, M. W., Self, W. H., Wanderer, J. P., Ehrenfeld, J. M., Wang, L., Byrne, D. W., … Rice, T. W. (2018). Balanced crystalloids versus saline in critically ill adults. New England Journal of Medicine, 378(9), 829-839. https://doi.org/10.1056/NEJMoa1711584

Shapiro, N. I., Douglas, I. S., Brower, R. G., Brown, S. M., Exline, M. C., Ginde, A. A., … National Heart, Lung, and Blood Institute Prevention and Early Treatment of Acute Lung Injury Clinical Trials Network. (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

Singer, M., Deutschman, C. S., Seymour, C. W., Shankar-Hari, M., Annane, D., Bauer, M., … Angus, D. C. (2016). The third international consensus definitions for sepsis and septic shock (Sepsis-3). JAMA, 315(8), 801-810. https://doi.org/10.1001/jama.2016.0287