Residency · Residency · Critical Care

Sepsis and Septic Shock: Surviving Sepsis Campaign

Definitions and Epidemiology

Sepsis-3 Definitions (2016)

The Third International Consensus Definitions for Sepsis and Septic Shock, published in 2016, represented a fundamental reconceptualization of sepsis. Sepsis is now defined as life-threatening organ dysfunction caused by a dysregulated host response to infection, explicitly abandoning the prior systemic inflammatory response syndrome (SIRS)-based definition that had been the clinical standard since 1991. Organ dysfunction is operationally identified as an acute change in the Sequential Organ Failure Assessment (SOFA) score of 2 points or more, with a baseline of zero assumed for patients without known pre-existing organ dysfunction. This definitional shift reflects the understanding that sepsis is not merely the presence of infection with an inflammatory response, but rather the transition point at which the host response becomes maladaptive and begins causing organ injury.

Septic shock is defined as a subset of sepsis in which the circulatory and cellular metabolic abnormalities are sufficiently profound to substantially increase mortality. The operational criteria require both a vasopressor requirement to maintain a mean arterial pressure of 65 mmHg or greater and a serum lactate level exceeding 2 mmol/L despite adequate volume resuscitation. This dual requirement distinguishes septic shock from simple sepsis-associated hypotension and identifies a population with mortality rates of 40-50%, compared to approximately 10% for sepsis without shock.

SOFA Score Components

The SOFA score evaluates six organ systems, each scored from 0 to 4: respiration (PaO2/FiO2 ratio), coagulation (platelet count), liver (bilirubin), cardiovascular (mean arterial pressure and vasopressor requirements), central nervous system (Glasgow Coma Scale), and renal function (creatinine and urine output). The total score ranges from 0 to 24, with higher scores indicating greater organ dysfunction and higher mortality.

Organ SystemScore 0Score 1Score 2Score 3Score 4
Respiration (PaO₂/FiO₂)≥400<400<300<200 (with respiratory support)<100 (with respiratory support)
Coagulation (Platelets ×10³/µL)≥150<150<100<50<20
Liver (Bilirubin, mg/dL)<1.21.2–1.92.0–5.96.0–11.9>12.0
CardiovascularMAP ≥70MAP <70Dopamine <5 or dobutamine (any)Dopamine 5.1–15 or epi/norepi ≤0.1Dopamine >15 or epi/norepi >0.1
CNS (GCS)1513–1410–126–9<6
Renal (Creatinine, mg/dL)<1.21.2–1.92.0–3.43.5–4.9 or UO <500 mL/day>5.0 or UO <200 mL/day

The quick SOFA (qSOFA) was proposed as a bedside screening tool incorporating three easily assessable clinical variables: respiratory rate of 22 or greater, altered mentation, and systolic blood pressure of 100 mmHg or less. A qSOFA score of 2 or more should prompt further evaluation for sepsis, including formal SOFA scoring and appropriate clinical investigation. However, subsequent validation studies have demonstrated that qSOFA has poor sensitivity (approximately 51%) for sepsis identification, and it is therefore not recommended as the sole screening tool. Its primary utility lies in identifying patients at higher risk of poor outcomes who warrant closer evaluation, rather than as a diagnostic instrument.

Epidemiology

The global burden of sepsis is staggering. An estimated 48.9 million sepsis cases occurred worldwide in 2017, resulting in approximately 11 million sepsis-related deaths, representing roughly 20% of all global deaths. ICU mortality from sepsis has declined from approximately 45% to 25% over the past two decades, reflecting improvements in recognition, protocolized early management, lung-protective ventilation, and overall critical care delivery. The most common sources of infection leading to sepsis are pulmonary (35-50%), abdominal (20-30%), urogenital (10-15%), and skin and soft tissue (5-10%).

Pathophysiology

Immune Dysregulation

The pathophysiology of sepsis reflects a complex interplay between the invading pathogen and the host immune response. The initial hyperinflammatory phase begins with recognition of pathogen-associated molecular patterns (PAMPs) by pattern recognition receptors, including toll-like receptors (TLRs) and NOD-like receptors, on innate immune cells. This recognition triggers activation of the nuclear factor kappa-B (NF-kB) signaling pathway, which initiates transcription of pro-inflammatory cytokines including tumor necrosis factor-alpha, interleukin-1, interleukin-6, and interleukin-8. The resulting cytokine storm propagates the systemic inflammatory cascade far beyond the local site of infection.

Concurrent with the hyperinflammatory response, the immune system paradoxically develops a state of immunosuppression termed immunoparalysis. This is characterized by downregulation of monocyte human leukocyte antigen-DR (HLA-DR) expression, massive lymphocyte apoptosis, and T-cell exhaustion with upregulation of inhibitory checkpoint receptors. It is critically important to recognize that the transition from hyperinflammation to immunosuppression is not a purely sequential process; rather, these opposing immune states coexist to varying degrees across different patients and at different time points in the same patient, creating a highly heterogeneous and dynamic immunological landscape.

Endothelial and Microcirculatory Dysfunction

Endothelial activation and injury are central to the pathophysiology of sepsis-associated organ dysfunction. Degradation of the endothelial glycocalyx, a protective carbohydrate-rich layer that lines the vascular endothelium, leads to dramatically increased vascular permeability and capillary leak. Endothelial activation promotes expression of tissue factor on the endothelial surface, initiating the coagulation cascade and producing widespread microvascular thrombosis that constitutes the pathological basis of disseminated intravascular coagulation (DIC). Simultaneously, overproduction of nitric oxide by inducible nitric oxide synthase (iNOS) produces profound vasodilation and vasoplegia that is often refractory to conventional vasopressor therapy.

At the microcirculatory level, sepsis produces a characteristic pattern of heterogeneous flow, in which stopped-flow capillaries exist immediately adjacent to hyperperfused capillaries. This microcirculatory shunting results in areas of tissue hypoxia despite what may appear to be adequate macrocirculatory parameters (blood pressure, cardiac output, oxygen delivery), explaining the phenomenon of hemodynamic incoherence that is a hallmark of septic shock.

Organ Dysfunction Mechanisms

Sepsis-associated organ dysfunction affects virtually every organ system through a combination of macrocirculatory and microcirculatory derangements, direct inflammatory injury, and metabolic dysregulation. Septic cardiomyopathy, present in approximately 30% of patients with septic shock, manifests as biventricular global hypokinesis with left ventricular dilation and is characteristically reversible with resolution of the septic insult. Pulmonary injury produces acute respiratory distress syndrome through either direct (pneumonia) or indirect (non-pulmonary sepsis) mechanisms. Acute kidney injury in sepsis is not purely a consequence of hypoperfusion, as was historically believed, but involves complex interactions between hemodynamic alterations, tubular inflammatory injury, and mitochondrial metabolic reprogramming. Hepatic dysfunction manifests as cholestasis, ischemic hepatitis, and impaired drug metabolism. Sepsis-associated encephalopathy produces diffuse cerebral dysfunction with EEG abnormalities and contributes to ICU delirium. Hematologic dysfunction presents as DIC with thrombocytopenia, prolonged coagulation times, elevated D-dimer, and low fibrinogen.

<image>Detailed pathophysiology diagram of sepsis showing the cascade from pathogen invasion to organ dysfunction. Start with bacteria/fungi releasing PAMPs, interacting with innate immune cells (macrophages, neutrophils) via toll-like receptors. Show branching pathways: (1) cytokine release leading to systemic inflammation; (2) endothelial activation leading to glycocalyx shedding and capillary leak; (3) coagulation cascade activation leading to microvascular thrombosis; (4) mitochondrial dysfunction leading to cellular energy failure. Terminal boxes showing effects on each organ system (heart, lungs, kidneys, liver, brain). Include both pro-inflammatory and immunosuppressive arms of the response.</image>

Surviving Sepsis Campaign Guidelines (2021 Update)

Hour-1 Bundle

The Surviving Sepsis Campaign Hour-1 Bundle represents the critical time-sensitive interventions that should be initiated within the first hour of sepsis recognition. Serum lactate should be measured immediately, with remeasurement within 2-4 hours if the initial level exceeds 2 mmol/L, as lactate kinetics provide essential information about the trajectory of tissue perfusion. Blood cultures, at minimum two sets including both aerobic and anaerobic bottles, should be obtained before antibiotic administration, though this should never delay antibiotic therapy when culture procurement cannot be accomplished rapidly.

Broad-spectrum antibiotics covering the most likely pathogens based on the suspected source of infection must be administered within one hour of sepsis recognition, as each hour of delay in appropriate antibiotic therapy is associated with incremental increases in mortality. Rapid crystalloid administration should be initiated for patients with hypotension or lactate levels of 4 mmol/L or greater. The 2021 guideline update notably suggests individualization of this volume rather than mandating a fixed 30 mL/kg for all patients, reflecting growing recognition that one-size-fits-all fluid prescriptions may cause harm through excessive fluid administration in some patients. Vasopressors should be applied if hypotension persists during or after fluid resuscitation to maintain a MAP of 65 mmHg or greater.

Antimicrobial Therapy

Empiric antimicrobial therapy should provide broad-spectrum coverage targeting the most likely pathogens based on the suspected source of infection, local microbiological epidemiology, and patient risk factors. Coverage for methicillin-resistant Staphylococcus aureus (MRSA) with vancomycin (dosed at 15-20 mg/kg with an AUC/MIC target of 400-600) or linezolid should be included when risk factors are present, including prior MRSA colonization or infection, recent hospitalization, and indwelling central lines. Antipseudomonal coverage with agents such as piperacillin-tazobactam, cefepime, meropenem, or ceftazidime should be included for nosocomial infections, though routine double antipseudomonal coverage is not recommended. Empiric antifungal coverage with an echinocandin should be considered in patients with prior Candida colonization, total parenteral nutrition, prolonged broad-spectrum antibiotic exposure, immunosuppression, or recent abdominal surgery.

De-escalation of antimicrobial therapy to narrower-spectrum agents based on culture results should be pursued within 24-48 hours whenever feasible. Duration of therapy should generally be limited to 7-8 days for most infections, as studies comparing 8-day versus 15-day courses in ventilator-associated pneumonia demonstrated equivalent mortality with significantly less antibiotic exposure in the shorter-course group. Procalcitonin-guided discontinuation, as evaluated in the PRORATA and SAPS trials, represents an evidence-based strategy for reducing antibiotic duration while maintaining clinical safety.

Source Control

Identification and control of the infectious source represents one of the most critical interventions in sepsis management and should be accomplished within 6-12 hours of identification whenever possible. Specific source control measures include drainage of abscesses, debridement of necrotic tissue, and removal of infected devices such as central venous catheters, prosthetic joints, or implantable cardiac devices. The fundamental principle is that no amount of antibiotics will sterilize an undrained abscess, making source control the definitive therapeutic intervention.

A notable exception is infected pancreatic necrosis, for which the current evidence supports delayed intervention until the necrosis has become walled off, ideally beyond 4 weeks from onset, using a step-up approach beginning with percutaneous drainage rather than primary open surgical necrosectomy. For acute cholecystitis in patients who are not surgical candidates, percutaneous cholecystostomy tube placement provides effective source control.

Hemodynamic Management

Fluid Resuscitation

Balanced crystalloids are preferred over normal saline for fluid resuscitation in sepsis based on the SMART trial evidence. Initial resuscitation should consist of 1-2 liters of crystalloid with subsequent reassessment, rather than reflexive administration of 30 mL/kg to all patients. The CLOVERS trial of 2023 compared a restrictive, vasopressor-first strategy with a liberal, fluid-first strategy in early sepsis and found no difference in 90-day mortality, providing reassurance that a more restrained approach to initial fluid resuscitation is safe and potentially beneficial by avoiding the downstream complications of fluid overload.

Albumin may be used as an adjunctive resuscitation fluid in sepsis. The ALBIOS trial demonstrated a non-significant trend toward mortality reduction when albumin was used to maintain serum albumin levels above 30 g/L, though the study was not powered to detect a definitive mortality difference. Hydroxyethyl starch must be avoided in sepsis, as both the 6S and CHEST trials demonstrated increased rates of acute kidney injury, renal replacement therapy, and (in the 6S trial) mortality with its use.

Vasopressor Therapy

Norepinephrine is the first-line vasopressor for septic shock, a strong recommendation supported by extensive evidence. Vasopressin at a fixed dose of 0.03 U/min is recommended as the preferred second-line agent to reduce norepinephrine dose. The VASST trial demonstrated no overall mortality benefit of adding vasopressin to norepinephrine, but a subgroup analysis suggested possible benefit in patients with less severe septic shock (norepinephrine requirement less than 15 mcg/min). The VANISH trial comparing vasopressin to norepinephrine as a first-line agent similarly found no mortality difference but reported less renal replacement therapy requirement with vasopressin. Epinephrine serves as an alternative second-line agent.

Angiotensin II (Giapreza), a non-adrenergic vasopressor acting on the AT1 receptor, demonstrated improved mean arterial pressure response in the ATHOS-3 trial of refractory vasodilatory shock and should be considered in patients with vasoplegia that is unresponsive to catecholamines and vasopressin. Dopamine is explicitly not recommended due to increased arrhythmia risk compared to norepinephrine, as demonstrated in the SOAP II trial (24% versus 12% arrhythmia rate). Phenylephrine, a pure alpha-1 agonist, should generally be avoided as it may reduce cardiac output in the setting of septic cardiomyopathy.

Inotropic Support

Dobutamine is indicated for septic cardiomyopathy when persistent hypoperfusion continues despite adequate volume resuscitation and vasopressor therapy. The therapeutic goal should be normalization of lactate and ScvO2 through restoration of adequate cardiac output, rather than targeting supranormal oxygen delivery, which has not been shown to improve outcomes and may cause harm. Levosimendan, a calcium sensitizer, was specifically evaluated in the LeoPARDS trial for septic shock and demonstrated no benefit with an increased incidence of supraventricular tachyarrhythmias.

Corticosteroids

Intravenous hydrocortisone at 200 mg per day, administered either as 50 mg every 6 hours or as a continuous infusion, is recommended for patients with septic shock that remains refractory to vasopressor therapy despite adequate resuscitation. The ADRENAL trial of 2018 demonstrated no improvement in 90-day mortality with hydrocortisone but showed faster resolution of shock. In contrast, the APROCCHSS trial of 2018, which used the combination of hydrocortisone and fludrocortisone, demonstrated a significant reduction in 90-day mortality (43% versus 49%, NNT 17).

TrialYearNInterventionControlPrimary OutcomeKey Finding
ADRENAL20183,658Hydrocortisone 200 mg/day IVPlacebo90-day mortalityNo mortality difference; faster shock resolution
APROCCHSS20181,241Hydrocortisone + fludrocortisonePlacebo90-day mortalityMortality reduction (43% vs 49%, NNT 17)The cosyntropin stimulation test should not be used to guide the decision to administer corticosteroids, as its predictive value in the context of critical illness is insufficient to influence clinical decision-making. A reasonable threshold for initiating corticosteroid therapy is a norepinephrine requirement of 0.25 mcg/kg/min or greater persisting after 4-6 hours of resuscitation.

<image>Sepsis resuscitation decision tree flowchart. Entry point: "Suspected sepsis with organ dysfunction (SOFA >=2)." First tier: Hour-1 bundle (measure lactate, obtain cultures, give antibiotics, start fluids if indicated, vasopressors if hypotensive). Second tier: Reassessment at 1-3 hours -- branches based on hemodynamic response. If fluid responsive and improving: continue targeted resuscitation. If refractory hypotension: escalate vasopressors (norepinephrine -> add vasopressin -> consider angiotensin II). If refractory hypoperfusion despite MAP >=65: assess cardiac function with echo (branches to septic cardiomyopathy -> dobutamine, vs. vasoplegia -> add vasopressin/hydrocortisone). Include specific drug doses, MAP targets, and reassessment timepoints throughout.</image>

Organ Support

Respiratory Support

Oxygen therapy should target an SpO2 of 94-98%, with deliberate avoidance of hyperoxia, which has been associated with increased mortality in some observational studies. The ICU-ROX trial demonstrated that a conservative, permissive approach to oxygenation did not improve ventilator-free days, while the HOT-ICU trial showed no benefit of targeting a lower PaO2 of 60 mmHg versus 90 mmHg. When ARDS develops, lung-protective ventilation with tidal volumes of 6 mL/kg ideal body weight and plateau pressures below 30 cmH2O remains the cornerstone of ventilatory management.

Renal Support

The timing of renal replacement therapy initiation in sepsis-associated AKI should be individualized rather than protocol-driven. The STARRT-AKI trial of 2020 demonstrated no benefit of accelerated RRT initiation compared to a standard strategy based on conventional indications, and the IDEAL-ICU trial of 2018 found no difference in 90-day mortality between early and delayed RRT initiation in septic shock. Avoidance of nephrotoxic agents and maintenance of adequate renal perfusion pressure remain the primary preventive strategies.

Glucose Management

Blood glucose should be targeted between 144-180 mg/dL, reflecting the findings of the NICE-SUGAR trial of 2009, which demonstrated that tight glucose control (81-108 mg/dL) was associated with increased mortality compared to conventional management targeting glucose below 180 mg/dL. Insulin infusion should be initiated when glucose exceeds 180 mg/dL, with careful avoidance of hypoglycemia below 70 mg/dL.

Venous Thromboembolism Prophylaxis

Low-molecular-weight heparin is preferred over unfractionated heparin for VTE prophylaxis, offering better DVT prevention with comparable bleeding risk. Pharmacological prophylaxis should be administered unless contraindicated by active bleeding or severe thrombocytopenia, with mechanical prophylaxis using intermittent pneumatic compression devices employed when pharmacological prophylaxis is contraindicated. Therapeutic-dose anticoagulation is not beneficial in critically ill septic patients, as the REMAP-CAP, ATTACC, and ACTIV-4a trials demonstrated harm in this population.

Stress Ulcer Prophylaxis

Proton pump inhibitors are preferred over H2-receptor antagonists for stress ulcer prophylaxis in high-risk patients, defined as those receiving mechanical ventilation for more than 48 hours or those with coagulopathy. However, the SUP-ICU trial of 2018 demonstrated no difference in 90-day mortality or clinically important gastrointestinal bleeding between pantoprazole and placebo, suggesting that stress ulcer prophylaxis may be safely withheld in lower-risk patients, particularly those receiving enteral nutrition.

Biomarkers and Prognostication

Procalcitonin (PCT)

Procalcitonin has established a role as a biomarker for distinguishing bacterial infection from non-infectious inflammation, with levels exceeding 0.5 ng/mL demonstrating a sensitivity of 77% and specificity of 79% for bacterial infection. Its greatest clinical utility lies in guiding antibiotic de-escalation: discontinuation of antibiotics when procalcitonin falls below 0.5 ng/mL or declines by more than 80% from peak has been shown to safely reduce antibiotic exposure. Limitations include false elevations following cardiopulmonary bypass, in burns, severe trauma, and medullary thyroid cancer, and inappropriately low levels in localized infections without systemic involvement.

Lactate Kinetics

Serial lactate measurement provides more clinically useful information than a single value, as the trajectory of lactate over time reflects the adequacy of resuscitation more reliably than any absolute threshold. Lactate clearance exceeding 20% at 2 hours correlates with improved survival, while persistent lactate elevation above 4 mmol/L at 24 hours identifies a population with mortality exceeding 50%.

Emerging Biomarkers

Several novel biomarkers are under investigation for early sepsis detection and risk stratification. Presepsin (soluble CD14 subtype) shows promise as an early marker of bacterial infection, while pancreatic stone protein may outperform procalcitonin for early sepsis detection. Monocyte HLA-DR expression, when reduced below 8000 antigens per monocyte, serves as a marker of immunoparalysis and predicts the risk of nosocomial infection. While none of these biomarkers has yet been incorporated into routine clinical practice, they represent potential tools for future precision medicine approaches to sepsis management.

Key Clinical Pearls

  • Sepsis-3 eliminated SIRS criteria from the sepsis definition -- sepsis is now defined by organ dysfunction (SOFA >= 2)
  • The Hour-1 Bundle should be initiated within the first hour of sepsis recognition; delays in antibiotics increase mortality
  • Fluid resuscitation should be individualized -- reassess responsiveness before each bolus; the 30 mL/kg is a guideline, not a mandate
  • Norepinephrine is first-line vasopressor; vasopressin is the preferred second agent
  • Source control is often the definitive treatment -- antibiotics alone cannot sterilize undrained collections
  • Hydrocortisone (200 mg/day) should be considered in refractory septic shock; do not use cosyntropin stimulation testing
  • Procalcitonin kinetics can safely guide antibiotic de-escalation and reduce antibiotic exposure
  • Think beyond the initial resuscitation: transition to de-resuscitation and organ support optimization within 24-48 hours

References

  1. Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315(8):801-810.
  2. 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;47(11):1181-1247.
  3. Annane D, Renault A, Brun-Buisson C, et al. Hydrocortisone plus fludrocortisone for adults with septic shock. N Engl J Med. 2018;378(9):809-818.
  4. Self WH, Semler MW, Wanderer JP, et al. Balanced crystalloids versus saline in noncritically ill adults. N Engl J Med. 2018;378(9):819-828.
  5. Khanna A, English SW, Wang XS, et al. Angiotensin II for the treatment of vasodilatory shock. N Engl J Med. 2017;377(5):419-430.
Sepsis and Septic Shock: Surviving Sepsis Campaign — figure 1
Sepsis and Septic Shock: Surviving Sepsis Campaign — figure 2

Read this lecture as Markdown