Sepsis Management in ICU

Learning Objectives

• State the Sepsis-3 definitions of sepsis and septic shock and explain why they replaced the old SIRS-based definitions
• Calculate a qSOFA score at the bedside and explain how it differs from the full SOFA score
• List the components of the Surviving Sepsis Campaign hour-1 bundle in the order they should be executed
• Explain the physiologic rationale for lactate-guided fluid resuscitation
• Justify norepinephrine as the first-line vasopressor in septic shock and know when to add a second agent
• Describe the role and timing of source control in a patient who is not responding to antibiotics and fluids
• Identify common ICU complications of sepsis and how they are prevented or managed

Quick Answer

Sepsis is life-threatening organ dysfunction caused by a dysregulated host response to infection; septic shock is sepsis with circulatory and cellular/metabolic abnormalities severe enough to substantially raise mortality. In the ICU, the modern approach is protocolized and time-sensitive: draw blood cultures and lactate, give broad-spectrum antibiotics within one hour of recognition, resuscitate with 30 mL/kg of crystalloid for hypotension or lactate ≥4 mmol/L, start norepinephrine if MAP stays below 65 mmHg despite fluids, and re-measure lactate to confirm the patient is actually improving rather than just looking better on paper. Source control — draining an abscess, removing an infected line — is done as soon as it is safe, because no antibiotic can sterilize a walled-off collection of pus.

Why Sepsis Definitions Changed: SIRS to Sepsis-3

For years, sepsis was taught as "infection + SIRS criteria" (temperature, heart rate, respiratory rate, white cell count). The problem was that SIRS is present in huge numbers of patients who are not actually in danger — a person jogging into the emergency department with a viral cold can meet two SIRS criteria. SIRS was sensitive but not specific for the thing that actually kills patients: organ dysfunction.

In 2016, the Sepsis-3 task force redefined things around outcome, not inflammation:

• Sepsis = life-threatening organ dysfunction caused by a dysregulated host response to infection, operationalized as an acute increase of ≥2 points in the SOFA (Sequential Organ Failure Assessment) score in a patient with suspected or confirmed infection.
• Septic shock = a subset of sepsis with circulatory and cellular/metabolic dysfunction severe enough to increase mortality substantially. Clinically this means: the patient needs vasopressors to maintain MAP ≥ 65 mmHg, AND has a serum lactate > 2 mmol/L, despite adequate fluid resuscitation.

Think of it this way: SIRS asked "does this look like inflammation?" Sepsis-3 asks "is this infection actually damaging organs?" The second question is the one that predicts death, so it is the one guidelines now anchor on.

qSOFA — the bedside screening tool

The full SOFA score needs labs (creatinine, bilirubin, platelets, PaO2/FiO2) that take time to come back. For quick screening outside the ICU — in the emergency department, on the ward, in triage — clinicians use qSOFA (quick SOFA), which needs nothing but a clinical exam:

1. Respiratory rate ≥ 22/min
2. Altered mentation (any change in GCS from baseline)
3. Systolic blood pressure ≤ 100 mmHg

A score of ≥2 flags a patient at higher risk of poor outcome from suspected infection and should prompt escalation of care and formal SOFA/organ-function assessment. It is important to know: qSOFA is a prognostic screening tool, not a diagnostic criterion for sepsis itself — a patient can have sepsis with a qSOFA of 0 or 1, especially early on. Do not use a low qSOFA to rule out sepsis; use it to flag who needs closer monitoring.

Clinical Presentation

Sepsis can present dramatically or subtly, which is exactly why it kills people who are triaged as "not that sick." Look for:

• Fever or, in the elderly and immunosuppressed, hypothermia (a worse prognostic sign — it suggests the immune system cannot mount a febrile response)
• Tachycardia and tachypnea (early compensatory signs, often present before hypotension)
• Hypotension — a late sign; by the time blood pressure drops, compensatory mechanisms are failing
• Altered mental status — confusion or lethargy, especially in elderly patients, may be the only early clue
• Signs of end-organ hypoperfusion: mottled skin, delayed capillary refill, oliguria
• Localizing signs of the source infection (cough and hypoxia for pneumonia, flank pain for pyelonephritis, abdominal pain for peritonitis, erythema for cellulitis)

The teaching point worth remembering: an elderly patient with new confusion and no fever can still be septic. Vital signs lag behind the underlying process — never wait for a classic fever-tachycardia-hypotension triad before you start working up sepsis.

Diagnosis and Initial Workup

Diagnosis in the ICU is really two parallel tasks happening at once: (1) confirming organ dysfunction and severity, and (2) finding the source of infection.

• Severity assessment: SOFA score using labs (creatinine, bilirubin, platelet count, PaO2/FiO2, GCS, MAP/vasopressor requirement)
• Source identification: blood cultures (two sets, from different sites, before antibiotics whenever this does not meaningfully delay treatment), urine culture, sputum culture, imaging (chest X-ray, CT abdomen/pelvis) guided by the suspected source
• Lactate: a marker of tissue hypoperfusion and anaerobic metabolism; a starting point, and — critically — something you re-check (typically at 2–4 hours) to see if resuscitation is working
• Inflammatory markers: procalcitonin and CRP can support the diagnosis and, increasingly, guide antibiotic de-escalation, but they are adjuncts, not substitutes for clinical judgment

The Surviving Sepsis Campaign Hour-1 Bundle

This is the single most testable piece of sepsis management, and the name is deliberately literal: these actions should all be initiated within one hour of sepsis or septic shock being recognized, not completed sequentially over the day.

The four hour-1 actions, in the order clinicians usually execute them:

1. Measure lactate. Remeasure if the initial value is above 2 mmol/L — a single number tells you little; the trend tells you whether resuscitation is working.
2. Obtain blood cultures before antibiotics — but only if this can be done without meaningfully delaying antibiotic administration. Never let culture logistics push antibiotics past the one-hour mark.
3. Administer broad-spectrum antibiotics. Every hour of delay in septic shock has been associated with measurably higher mortality in observational data — this is the single strongest argument for urgency in sepsis care.
4. Begin rapid administration of 30 mL/kg crystalloid for hypotension or lactate ≥ 4 mmol/L. If hypotension persists during or after this fluid challenge, start vasopressors to maintain MAP ≥ 65 mmHg.

Antibiotics and Source Control

Empiric antibiotics must be broad-spectrum and cover the most likely pathogens for the suspected source, local resistance patterns, and patient-specific risk factors (recent hospitalization, prior resistant organisms, immunosuppression). Once cultures return, de-escalate to the narrowest effective agent — this is not optional housekeeping, it is core antimicrobial stewardship that reduces resistance and side effects like C. difficile colitis.

Source control is the step trainees most often under-appreciate. No dose of antibiotic can penetrate an abscess wall, sterilize an infected prosthetic joint, or clear a necrotic gallbladder. If there is a drainable collection, an infected device, or necrotic tissue, it needs to be drained, removed, or debrided — ideally within the first 6–12 hours once the patient is stable enough for the intervention. A patient who keeps decompensating despite "appropriate" antibiotics almost always has a source control problem, not an antibiotic choice problem.

Fluid Resuscitation: Why Lactate-Guided, Not ScvO2-Guided

Older protocols (Early Goal-Directed Therapy, EGDT) targeted a central venous oxygen saturation (ScvO2) of 70–80% using a fixed algorithm of fluids, transfusion, and inotropes. Large trials (ProCESS, ARISE, ProMISe) later showed that protocolized EGDT was no better than usual clinical care with lactate monitoring — and usual care was simpler and cheaper. Current Surviving Sepsis Campaign guidance therefore favors dynamic, lactate-guided resuscitation over a fixed ScvO2 target:

• Initial bolus: 30 mL/kg balanced crystalloid within the first 3 hours for hypotension or lactate ≥ 4 mmol/L
• Reassess with dynamic measures of fluid responsiveness (passive leg raise, pulse pressure variation, bedside ultrasound of the IVC) rather than static numbers like CVP alone — a normal CVP does not tell you whether more fluid will help
• Track lactate clearance: a falling lactate over 2–4 hours is reassuring; a static or rising lactate despite fluids means shock is ongoing and you need to reassess the diagnosis, the source, or escalate support
• Avoid indiscriminate over-resuscitation — excess fluid causes pulmonary edema, abdominal compartment syndrome, and worse outcomes in ARDS; "enough fluid to restore perfusion," not "as much fluid as possible," is the goal

Vasopressors and Refractory Shock

Norepinephrine is the first-line vasopressor in septic shock. It provides potent alpha-adrenergic vasoconstriction with modest beta-1 inotropic support, raising MAP without the excessive tachycardia and arrhythmia risk seen with dopamine (dopamine is no longer recommended as first-line and is associated with more arrhythmias).

• If MAP remains below 65 mmHg on high-dose norepinephrine, add vasopressin (fixed low dose, roughly 0.03 units/min) as a second agent — it works through a different receptor pathway and allows some reduction in catecholamine dose.
• Epinephrine can be added or substituted when a third agent is needed or when there is a component of myocardial depression.
• For patients who remain hypotensive despite adequate fluids and vasopressors (refractory septic shock), give IV hydrocortisone (typically 200 mg/day). This does not reverse shock immediately but can shorten the duration of vasopressor dependence — it treats relative adrenal insufficiency that critical illness can unmask.
• Dobutamine is added, not substituted, when there is evidence of persistent hypoperfusion with adequate volume and blood pressure but signs of myocardial dysfunction (low cardiac output despite normal MAP on vasopressors).

Organ Support and ICU-Level Monitoring

Sepsis frequently produces multi-organ dysfunction, and ICU management runs in parallel with the source-directed treatment above:

• Respiratory failure: lung-protective mechanical ventilation (low tidal volume, ~6 mL/kg predicted body weight) if ARDS develops, which is common in sepsis
• Acute kidney injury: renal replacement therapy (usually CRRT in hemodynamically unstable patients) for refractory hyperkalemia, acidosis, fluid overload, or uremia
• Coagulopathy: sepsis can trigger disseminated intravascular coagulation (DIC); monitor platelets, fibrinogen, and coagulation studies
• Glycemic control: target moderate glucose control (roughly 140–180 mg/dL) — tight control below 110 mg/dL increases hypoglycemia risk without proven mortality benefit
• Continuous monitoring: arterial line for real-time blood pressure, central line for vasopressors, early warning scores (e.g., NEWS2) to catch deterioration on the ward before ICU transfer becomes an emergency

Prevention and Stewardship

• Early recognition protocols and sepsis screening tools embedded in triage and nursing assessments
• Strict infection-prevention bundles for catheter-associated UTIs, central-line-associated bloodstream infections, and ventilator-associated pneumonia — since ICU-acquired infections are themselves a major cause of secondary sepsis
• Antimicrobial stewardship (de-escalation, appropriate duration) to reduce resistance without compromising initial empiric coverage
• Vaccination (influenza, pneumococcus) to reduce the burden of the infections that most commonly progress to sepsis in vulnerable populations

Key Terms

Term	Definition	Related Concept
Sepsis (Sepsis-3)	Life-threatening organ dysfunction caused by a dysregulated host response to infection, defined as an acute SOFA increase ≥2	SOFA score, dysregulated host response
Septic shock	Sepsis with vasopressor requirement to maintain MAP ≥65 mmHg plus lactate >2 mmol/L despite adequate fluids	Vasopressors, lactate clearance
qSOFA	Bedside screen (RR ≥22, altered mentation, SBP ≤100) used to flag high-risk infected patients outside the ICU	Sepsis screening, SOFA
SOFA score	Sequential Organ Failure Assessment — six-organ system score (respiratory, coagulation, liver, cardiovascular, CNS, renal) used to define sepsis and track severity	ICU severity scoring
Hour-1 bundle	Surviving Sepsis Campaign bundle: lactate, blood cultures, antibiotics, 30 mL/kg fluids, all initiated within one hour of recognition	Time-to-antibiotic mortality
Lactate clearance	Serial fall in serum lactate over hours, used to confirm resuscitation is restoring tissue perfusion	Anaerobic metabolism, shock resolution
Source control	Physical removal/drainage of the infection focus (abscess drainage, line removal, debridement)	Antibiotic failure, refractory sepsis
Norepinephrine	First-line vasopressor in septic shock; alpha-predominant with mild inotropic effect	Vasopressin, refractory shock
Refractory septic shock	Persistent hypotension despite adequate fluids and vasopressors, often treated with hydrocortisone	Relative adrenal insufficiency
CRRT	Continuous Renal Replacement Therapy — dialysis for hemodynamically unstable septic patients with AKI	Acute kidney injury, fluid balance

Common Mistakes

Misconception: A patient can't have sepsis unless they meet SIRS criteria (fever, tachycardia, elevated white count).

Why it's wrong: SIRS criteria are neither required nor part of the current Sepsis-3 definition. They are nonspecific — many non-infected patients (post-surgical, post-exercise, anxious) meet SIRS criteria — and many genuinely septic patients, especially elderly or immunosuppressed ones, do not mount a SIRS response at all.

Correct understanding: Sepsis is defined by acute organ dysfunction (SOFA increase ≥2) in the setting of suspected or confirmed infection. qSOFA is used only as a bedside prognostic screen outside the ICU, not as a diagnostic requirement, and a low qSOFA does not exclude sepsis.

---

Misconception: The goal of fluid resuscitation is to give as much crystalloid as possible until blood pressure normalizes.

Why it's wrong: Aggressive, unguided fluid administration causes pulmonary edema, worsens ARDS, and can precipitate abdominal compartment syndrome. Trials comparing protocolized EGDT (which pushed large fixed fluid volumes) against usual lactate-guided care showed no mortality benefit for the aggressive fixed-volume approach.

Correct understanding: Give an initial 30 mL/kg crystalloid bolus for hypotension or lactate ≥4 mmol/L, then reassess using dynamic measures of fluid responsiveness and trend the lactate. Stop escalating fluids once perfusion is restored, even if blood pressure is not yet perfectly normalized — vasopressors, not more fluid, are the next step.

---

Misconception: Dopamine and norepinephrine are interchangeable first-line vasopressors in septic shock.

Why it's wrong: Randomized trial data (the SOAP II trial) showed dopamine is associated with significantly more cardiac arrhythmias than norepinephrine, with no mortality benefit, and a subgroup analysis suggested worse outcomes in cardiogenic shock patients.

Correct understanding: Norepinephrine is the first-line vasopressor for septic shock in current guidelines. Vasopressin is added as a second agent to reduce catecholamine dose in patients not responding to norepinephrine alone; dopamine is reserved for select circumstances (e.g., low risk of tachyarrhythmia, bradycardia) and is not first-line.

Comparison and Connections

Feature	Sepsis	Septic Shock	SIRS (older concept)
Core definition	Organ dysfunction (SOFA ↑≥2) from infection	Sepsis + vasopressor need for MAP ≥65 + lactate >2 despite fluids	≥2 of: fever/hypothermia, tachycardia, tachypnea, abnormal WBC
Specificity for danger	Moderate–high (tied to organ damage)	High (identifies the sickest subgroup)	Low — many non-infected patients qualify
Requires infection	Yes	Yes	No — can occur from trauma, pancreatitis, burns
Bedside screening tool	qSOFA (≥2 flags risk)	Vasopressor requirement + lactate	SIRS criteria (now largely retired from diagnosis)
Mortality association	Elevated	Substantially higher than sepsis alone	Poorly correlated with actual mortality

Practice Questions

Recall

1. What three variables make up the qSOFA score, and what cutoff for each counts as a positive point? Answer guidance: Respiratory rate ≥22/min, altered mentation (any change from baseline GCS), and systolic blood pressure ≤100 mmHg. A total score of ≥2 flags higher risk and should prompt closer monitoring or ICU-level assessment.

2. Name the four core actions in the Surviving Sepsis Campaign hour-1 bundle. Answer guidance: Measure (and remeasure if elevated) lactate; obtain blood cultures before antibiotics if feasible; administer broad-spectrum antibiotics; begin rapid 30 mL/kg crystalloid for hypotension or lactate ≥4 mmol/L, with vasopressors added if MAP stays below 65 mmHg.

Understanding

3. Why did Sepsis-3 move away from SIRS criteria as the definition of sepsis? Answer guidance: SIRS is highly sensitive but nonspecific — many non-infected conditions (surgery, exercise, anxiety) trigger SIRS, while some genuinely septic patients (elderly, immunosuppressed) never meet SIRS criteria. Sepsis-3 anchors the definition on acute organ dysfunction (SOFA increase ≥2), which correlates far better with the outcome that actually matters: mortality from infection.

4. Explain why lactate is measured serially rather than just once at presentation. Answer guidance: A single lactate tells you the patient has some degree of tissue hypoperfusion or anaerobic metabolism at that moment, but not whether treatment is working. Remeasuring at 2–4 hours shows the trend — a falling lactate (clearance) indicates resuscitation is restoring perfusion, while a static or rising lactate means the patient is still in shock and needs re-evaluation of fluids, vasopressors, source control, or the underlying diagnosis.

Application

5. A 68-year-old man presents with new confusion, respiratory rate 26/min, and systolic BP 92 mmHg. He is afebrile with a normal white cell count. Does he have a qSOFA-positive picture, and what should happen next? Answer guidance: Yes — he scores 3/3 on qSOFA (altered mentation, RR ≥22, SBP ≤100), despite lacking classic SIRS features like fever or leukocytosis. This should prompt immediate escalation: full sepsis workup (lactate, blood cultures, SOFA-relevant labs), search for an infectious source, and initiation of the hour-1 bundle if infection is confirmed or strongly suspected — his lack of fever does not rule out sepsis.

6. A septic shock patient has received 30 mL/kg crystalloid and remains hypotensive with MAP 58 mmHg. What is the next step, and what would you add if that step alone is insufficient? Answer guidance: Start norepinephrine as the first-line vasopressor to raise MAP to at least 65 mmHg. If MAP remains inadequate despite escalating norepinephrine doses, add fixed low-dose vasopressin as a second agent; if shock persists despite both agents and adequate volume, consider IV hydrocortisone for refractory shock and reassess for an uncontrolled source of infection.

Analysis

7. A patient with septic shock from a perforated diverticular abscess is on appropriate broad-spectrum antibiotics, adequate fluids, and norepinephrine, but remains hypotensive with a rising lactate at 6 hours. What is the most likely explanation, and what should be done? Answer guidance: The most likely problem is inadequate source control — antibiotics cannot sterilize a walled-off abscess or ongoing peritoneal contamination. Rather than simply escalating vasopressors or switching antibiotics, the team should pursue urgent source control (percutaneous drainage or surgical intervention) once the patient is stable enough, since persistent hemodynamic failure despite "appropriate" medical therapy is the classic sign of an uncontrolled source.

8. Compare the rationale for using dynamic measures of fluid responsiveness (passive leg raise, IVC ultrasound) versus static measures (CVP) when deciding whether to give more fluid in septic shock. Answer guidance: Static measures like a single CVP reading reflect a snapshot of venous pressure but do not reliably predict whether the heart will respond to more volume with increased stroke volume — a normal or even elevated CVP can occur in both fluid-responsive and fluid-unresponsive patients. Dynamic measures (passive leg raise producing a transient "auto-bolus," respiratory variation in IVC diameter, pulse pressure variation) directly test how the cardiovascular system responds to a change in preload, making them more accurate for guiding further fluid administration and reducing the risk of harmful over-resuscitation.

FAQ

Is a high lactate always caused by tissue hypoperfusion from shock? No. Lactate can rise from causes unrelated to hypoperfusion, including beta-agonist use (such as high-dose albuterol or epinephrine), liver dysfunction (impaired lactate clearance), certain malignancies, and metformin accumulation in renal failure. In sepsis specifically, lactate elevation is thought to reflect a combination of tissue hypoperfusion and accelerated aerobic glycolysis driven by catecholamines and inflammation — it is not purely an anaerobic phenomenon. Regardless of mechanism, a persistently elevated or rising lactate in a septic patient is a reliable marker that something is still wrong and needs to be addressed.

Why is norepinephrine preferred over dopamine when both raise blood pressure? The SOAP II trial directly compared the two as first-line vasopressors in shock and found no difference in overall mortality, but significantly more arrhythmic events (mainly atrial fibrillation) with dopamine, along with a signal toward worse outcomes in the cardiogenic shock subgroup. Norepinephrine achieves comparable blood pressure support with a more favorable side-effect profile, which is why virtually every current guideline lists it as first-line.

If antibiotics are started within the hour, why do some septic patients still die or deteriorate? Antibiotics only address the infection through the bloodstream — they cannot penetrate abscesses, remove infected hardware, or debride necrotic tissue. A patient who continues to deteriorate despite "on-time, appropriate" antibiotics almost always needs source control, has a resistant organism not covered by the empiric regimen, or has progressed to irreversible multi-organ failure. Time-to-antibiotic matters enormously on a population level, but it is only one piece of a bundle that also depends on adequate resuscitation and source control.

How is septic shock different from other types of shock like cardiogenic or hypovolemic shock? Septic shock is a form of distributive shock: peripheral vasodilation drops systemic vascular resistance, so cardiac output is often normal or even elevated (warm, flushed extremities early on) rather than low, which is the opposite hemodynamic pattern from cardiogenic shock (low cardiac output, high SVR, cold extremities) or hypovolemic shock (low cardiac output, high SVR from compensatory vasoconstriction). This is exactly why vasopressors, not just fluids or inotropes, are central to septic shock management — the fundamental problem is loss of vascular tone, not primarily pump failure or volume loss.

When should hydrocortisone be given in septic shock, and does it save lives? Hydrocortisone is reserved for patients with septic shock who remain hemodynamically unstable despite adequate fluid resuscitation and vasopressor therapy — so-called refractory shock — rather than being given to everyone with sepsis. Trial evidence shows it shortens the duration of shock and vasopressor requirement, but the effect on mortality has been inconsistent across studies. It is best thought of as a tool to help wean vasopressors sooner in the sickest patients, not a universal sepsis treatment.

Quick Revision

• Sepsis-3: sepsis = infection + SOFA increase ≥2; septic shock = sepsis + vasopressor need for MAP ≥65 + lactate >2 mmol/L despite fluids
• qSOFA (RR ≥22, altered mentation, SBP ≤100; score ≥2 = high risk) is a bedside screening tool, not a diagnostic criterion — a low score does not exclude sepsis
• SIRS criteria are no longer part of the sepsis definition; they are sensitive but nonspecific
• Hour-1 bundle: measure/remeasure lactate, draw blood cultures before antibiotics if feasible, give antibiotics within 1 hour, give 30 mL/kg crystalloid for hypotension or lactate ≥4
• Every hour of delayed antibiotics in septic shock is linked to worse outcomes — do not wait for cultures to give antibiotics
• Fluid resuscitation is lactate- and dynamic-response-guided, not fixed-ScvO2-guided (EGDT protocols showed no benefit over usual care in ProCESS/ARISE/ProMISe)
• Norepinephrine is the first-line vasopressor; add vasopressin as a second agent before escalating norepinephrine indefinitely; avoid dopamine due to arrhythmia risk
• Hydrocortisone is for refractory shock (persistent hypotension despite fluids and vasopressors), not for all septic patients
• Source control (drainage, debridement, device removal) is mandatory whenever a controllable focus exists — antibiotics alone cannot sterilize an abscess
• De-escalate antibiotics once culture and sensitivity results are available — this is core stewardship, not an afterthought
• Watch for multi-organ complications: ARDS (lung-protective ventilation), AKI (CRRT if unstable), DIC (coagulation monitoring)
• Elderly and immunosuppressed patients may present with confusion or hypothermia instead of fever — don't anchor on classic vital sign patterns

Related Topics

Prerequisites: Basic infectious disease principles, shock physiology (distributive vs. cardiogenic vs. hypovolemic), cardiovascular and renal physiology, pharmacology of vasopressors and antibiotics

Related Topics: Intensive Care Unit Procedures, Acute Respiratory Distress Syndrome (ARDS), Disseminated Intravascular Coagulation, Antimicrobial Stewardship, Shock and Hemodynamic Monitoring

Next Topics: ARDS Management in ICU, Renal Replacement Therapy in ICU, Multi-Organ Dysfunction Syndrome, Antibiotic Stewardship Principles