Septic shock management

Contents

Definitions (Sepsis-3)

The Third International Consensus Definitions for Sepsis and Septic Shock (JAMA 2016):

Sepsis: life-threatening organ dysfunction caused by a dysregulated host response to infection.

  • Operationalised as: SOFA score increase ≥2 from baseline in the context of suspected infection

Septic shock: sepsis with circulatory and metabolic abnormalities sufficient to increase mortality.

  • Operationalised as: vasopressor requirement to maintain MAP ≥65 mmHg AND lactate >2 mmol/L despite adequate volume resuscitation
  • Hospital mortality ~40%

qSOFA (bedside screen): respiratory rate ≥22/min, altered mentation, SBP ≤100 mmHg; ≥2 criteria identifies patients at risk of poor outcome. Useful outside ICU but lower sensitivity than full SOFA.

Resuscitation Bundles

SSC Hour-1 Bundle (2018)

  1. Measure lactate; re-measure if initial >2 mmol/L
  2. Obtain blood cultures (≥2 sets) before antibiotics
  3. Administer broad-spectrum antibiotics
  4. Give 30 mL/kg crystalloid for hypotension (MAP <65) or lactate ≥4 mmol/L
  5. Apply vasopressors if hypotensive during or after fluid resuscitation to target MAP ≥65 mmHg

Resuscitation Targets

Target Threshold
MAP ≥65 mmHg (≥70–80 for chronic hypertensives — SEPSISPAM)
Lactate clearance ≥10% per 2h (>20% per 6h)
Urine output ≥0.5 mL/kg/h
ScvO₂ ≥70% (contextual)

Fluid Resuscitation

Crystalloid vs colloid

  • Balanced crystalloids preferred over 0.9% NaCl: SMART trial (Semler 2018, NEJM): balanced crystalloids reduced MAKE30 (major adverse kidney events) vs 0.9% NaCl in critically ill patients; 0.9% NaCl causes hyperchloraemic acidosis and may worsen renal vasoconstriction
  • Albumin: no mortality benefit vs crystalloid (SAFE trial); may have a role in septic shock patients requiring large volumes; 4% albumin appropriate; avoid 20% albumin for resuscitation
  • Avoid synthetic starches (HES): associated with AKI and increased mortality in sepsis (6S, CHEST trials)

Fluid responsiveness

After initial resuscitation, further fluids should only be given if the patient is fluid-responsive:

  • Passive leg raise (PLR): gold standard bedside test; raise legs to 45° from semi-recumbent; fluid-responsive if CO or pulse pressure increases ≥10%; reversible; works even in AF
  • Pulse pressure variation (PPV): >13% predicts fluid responsiveness in controlled mechanical ventilation (sinus rhythm, tidal volume ≥8 mL/kg)
  • POCUS: inferior vena cava collapsibility index, LVOT VTI response to PLR

Volume in established shock

  • CLASSIC trial (Meyhoff 2022, NEJM): restrictive vs standard fluid strategy after initial resuscitation in ICU septic shock → no difference in 90-day mortality; restrictive was non-inferior and safe. Supports reassessment-guided fluid therapy rather than continued empirical boluses.

Vasopressor Strategy

First-line: Noradrenaline

  • Potent α₁ agonist; raises MAP via SVR; first-line per SSC guidelines
  • Target MAP ≥65 mmHg (higher for chronic hypertensives)

Second-line adjuncts

Agent Dose Role
Vasopressin 0.03–0.04 units/min (fixed) V1 agonist; noradrenaline-sparing; relative vasopressin deficiency in sepsis; VASST trial: no mortality benefit but less RRT with vasopressin (VANISH)
Hydrocortisone 200 mg/day continuous infusion Vasopressor-sparing via glucocorticoid receptor-mediated upregulation of adrenoceptors; see ADRENAL, APROCCHSS
Adrenaline 0.01–1 mcg/kg/min Adds β₁ inotropic effect; useful if low CO co-exists; causes β₂-mediated lactic acidosis — do not treat lactate rise in isolation
Angiotensin II 5–80 ng/kg/min RAAS-mediated vasoconstriction; ATHOS-3 trial: increased MAP response in vasodilatory shock refractory to conventional vasopressors

MAP target

  • SEPSISPAM trial (Asfar 2014, NEJM): MAP 80–85 mmHg vs 65–70 mmHg → no 28-day mortality difference overall; chronic hypertensives had less AKI (less RRT) with higher target
  • Default: MAP ≥65 mmHg; individualise upward to 70–80 mmHg for known hypertensives

Adjunctive Therapies

Corticosteroids

  • ADRENAL trial (Venkatesh 2018, NEJM): hydrocortisone 200 mg/day vs placebo in mechanically ventilated septic shock → faster shock reversal, less blood transfusion, shorter ICU stay; no difference in 90-day mortality
  • APROCCHSS trial (Annane 2018, NEJM): hydrocortisone 200 mg/day + fludrocortisone 50 mcg/day vs placebo → reduced 90-day mortality (43% vs 49.1%); also faster shock reversal
  • SSC 2021 recommendation: consider hydrocortisone if MAP cannot be maintained with vasopressors alone
  • Mechanism: restores adrenoceptor sensitivity, anti-inflammatory; fludrocortisone adds mineralocorticoid effect (uncertain additional benefit)

Glucose control

  • NICE-SUGAR trial (2009, NEJM): intensive glucose control (4.5–6 mmol/L) vs conventional (≤10 mmol/L) → intensive control associated with increased 90-day mortality and more severe hypoglycaemia
  • Target: 6–10 mmol/L (SSC 2021 recommends <10 mmol/L)

What does NOT work

  • Activated protein C (drotrecogin alfa): withdrawn after PROWESS-SHOCK trial (2012) showed no mortality benefit
  • Routine packed red cell transfusion for Hb >70 g/L (TRICC, no benefit unless cardiac disease)
  • Tight glucose control (NICE-SUGAR)
  • Polymyxin B haemoperfusion: EUPHRATES trial neutral

Antimicrobials and Source Control

Antibiotic timing

  • Within 1 hour of septic shock recognition (SSC 2018)
  • Broad-spectrum empirical coverage; narrow on culture results (antimicrobial stewardship)
  • Duration: typically 5–7 days for uncomplicated bacteraemia; guided by clinical response and PCT (PRORATA trial: PCT-guided duration non-inferior and resulted in fewer days of antibiotics)

Source control

  • Identify and control the source within 6–12 hours
  • Drain over debride where possible
  • Remove infected prosthetic material promptly
  • Common sources: abdominal collection, biliary obstruction, infected line, necrotising fasciitis, empyema

Key Trials

Trial Year Question Finding
ARISE (Peake, NEJM 2014) 2014 Early goal-directed therapy (EGDT) vs usual care in septic shock No mortality benefit from EGDT; usual care equivalent
ProCESS (Yealy, NEJM 2014) 2014 Protocol-based EGDT vs usual care No mortality benefit; protocolised EGDT not required
PROMISE (Mouncey, NEJM 2015) 2015 EGDT vs usual care (UK multicentre) No mortality benefit from EGDT
SEPSISPAM (Asfar, NEJM 2014) 2014 MAP 65–70 vs 80–85 mmHg No mortality difference; higher target reduced RRT in hypertensives
ADRENAL (Venkatesh, NEJM 2018) 2018 Hydrocortisone 200 mg/day in septic shock Faster shock reversal; no mortality benefit
APROCCHSS (Annane, NEJM 2018) 2018 Hydrocortisone + fludrocortisone in septic shock Reduced 90-day mortality (43% vs 49.1%)
SMART (Semler, NEJM 2018) 2018 Balanced crystalloid vs 0.9% NaCl Balanced crystalloids reduced MAKE30
CLASSIC (Meyhoff, NEJM 2022) 2022 Restrictive vs standard fluid in established septic shock No mortality difference; restrictive non-inferior

Viva Questions

1. Define septic shock using Sepsis-3 criteria and explain the rationale for the definition.

Septic shock is defined as sepsis with circulatory and metabolic abnormalities sufficient to substantially increase mortality: specifically, a vasopressor requirement to maintain MAP ≥65 mmHg and a lactate >2 mmol/L despite adequate volume resuscitation. Sepsis itself is defined as life-threatening organ dysfunction — a SOFA score increase of ≥2 from baseline — in the context of suspected infection. The Sepsis-3 definitions moved away from SIRS criteria (which were non-specific) toward organ dysfunction as the defining feature, recognising that it is the dysregulated host response (not infection alone) that drives poor outcomes. The lactate threshold in septic shock identifies a subgroup with cellular metabolic dysfunction beyond haemodynamic compromise, associated with ~40% hospital mortality. The definitions provide prognostically meaningful, operationalisable criteria rather than purely pathophysiological ones.

2. A patient with septic shock requires 0.5 mcg/kg/min of noradrenaline. What are your next steps?

I would take a structured approach. First, reassess the basics: is the diagnosis correct (exclude obstructive or cardiogenic shock), is there adequate source control (undrained collection, infected line that needs removal), is volume status optimised (assess fluid responsiveness with PLR or PPV). Then consider adjuncts in parallel: add vasopressin 0.03–0.04 units/min as a noradrenaline-sparing V1 agonist with independent mechanism of action; consider hydrocortisone 200 mg/day continuous infusion given the refractory shock (restores adrenoceptor sensitivity, vasopressor-sparing — consistent with ADRENAL/APROCCHSS data). Bedside echo to exclude a cardiogenic component: if CI is low, add dobutamine or switch to adrenaline. If the diagnosis or source remains unclear, consider CT. Ensure a clear management plan and escalation ceiling discussion are documented, and revisit prognosis with the family.

3. How should you approach fluid resuscitation in septic shock and when should you stop?

Resuscitation has two phases. Initial resuscitation: in the first hour, give 30 mL/kg crystalloid (SSC) for hypotension or lactate ≥4 mmol/L, targeting MAP ≥65 mmHg and lactate clearance. Use balanced crystalloid (Ringer's lactate, Plasmalyte) rather than 0.9% NaCl to avoid hyperchloraemic acidosis and reduce MAKE events (SMART trial). After initial resuscitation, further fluid should only be given if the patient is fluid-responsive AND likely to benefit: use PLR (gold standard), PPV (if controlled MV, sinus rhythm), or POCUS to assess. Fluid responsiveness predicts CO improvement, not clinical benefit — stop if not responsive. The CLASSIC trial showed restrictive fluid strategy after initial resuscitation was non-inferior to standard care in established septic shock, supporting a reassessment-guided approach rather than continued empirical boluses. Avoid fluid overload — positive cumulative fluid balance is independently associated with worse outcomes in sepsis.

4. Summarise the corticosteroid evidence in septic shock and what you would do in practice.

Two large recent trials give conflicting signals. ADRENAL (Venkatesh 2018) found hydrocortisone 200 mg/day accelerated shock reversal and reduced blood transfusion requirements but did not reduce 90-day mortality. APROCCHSS (Annane 2018) found hydrocortisone 200 mg/day plus fludrocortisone 50 mcg/day reduced 90-day mortality (43% vs 49.1%). The difference may relate to fludrocortisone, patient selection, or chance. SSC 2021 recommends considering corticosteroids when shock is not adequately controlled with fluids and vasopressors. In practice: I would initiate hydrocortisone 200 mg/day (as a continuous infusion to avoid glucose excursions) in a patient requiring noradrenaline >0.25 mcg/kg/min despite adequate resuscitation and source control. I would also consider adding fludrocortisone 50 mcg daily given the APROCCHSS result. I would not routinely test cortisol levels (ACTH stimulation test) before starting — response testing does not guide clinical benefit in this context. Wean vasopressors first, then wean hydrocortisone over days once the patient is off vasopressors.