Paediatric sepsis

Contents

How Paediatric Sepsis Differs from Adults

Children are not small adults. Several fundamental differences affect the approach to paediatric sepsis:

Physiological Differences

  • Greater physiological reserve: children compensate well; heart rate increases dramatically with sepsis while blood pressure is maintained until late — hypotension is a pre-terminal sign of circulatory failure in children, not an early one
  • Higher baseline heart rate and respiratory rate: age-specific norms must be used; adult thresholds are inappropriate
  • Smaller absolute blood volumes: 70–80 mL/kg in children (vs ~70 mL/kg in adults), but proportionally similar — fluid prescriptions must be weight-based and carefully calculated
  • Glucose homeostasis: immature glycogen stores, higher metabolic rates, and proportionally larger brains make children at higher risk of hypoglycaemia during severe illness; monitor glucose frequently and avoid prolonged fasting

Immunological Differences

  • Immature humoral immunity in neonates and young infants: dependent on maternal antibodies (IgG transferred transplacentally; IgA from breast milk); susceptible to encapsulated organisms (Streptococcus pneumoniae, Neisseria meningitidis, Haemophilus influenzae)
  • Splenic function immature in early life — post-splenectomy-equivalent risk for encapsulated organisms in asplenic or hyposplenic states
  • Neonates: complement deficiency, reduced neutrophil function, impaired opsonisation

Airway and Respiratory Differences

  • Proportionally larger head and tongue → more easily obstructed airway; neutral position (not neck extension) for airway opening in infants
  • Higher respiratory rate and shorter expiratory time
  • Cricoid is the narrowest part of the airway in young children (vs glottis in adults); uncuffed ETTs traditionally used in young children (though cuffed ETTs increasingly used down to neonates with appropriate sizing)

Age-Specific Normal Values

Sepsis is defined by deviation from age-specific norms — vital signs and laboratory parameters vary significantly with age.

Age Heart rate (normal) SBP lower normal Respiratory rate
<1 month (neonate) 100–160 60 mmHg 30–60
1–12 months 100–160 70 mmHg 30–50
1–5 years 80–140 80 mmHg 20–40
5–12 years 70–120 90 mmHg 15–30
>12 years 60–100 100 mmHg 12–25

SBP lower limit (5th percentile, approximate): 70 + (2 × age in years) for ages 1–10 years

Persistent tachycardia is a more sensitive early indicator of shock than hypotension.

Organisms by Age

Age group Common organisms
Neonate (0–28 days) Group B Streptococcus (GBS), E. coli, Listeria monocytogenes, Staphylococcus aureus, Herpes simplex virus (HSV — consider always in septic-appearing neonate)
Infant (1–3 months) GBS (declining), S. pneumoniae, N. meningitidis, Hib (rare — vaccinated), Gram-negative enteric organisms
Child (3 months–5 years) S. pneumoniae, N. meningitidis, S. aureus, group A Streptococcus
Older child / adolescent Similar to adults: S. pneumoniae, N. meningitidis, S. aureus, S. pyogenes
Immunocompromised (any age) Gram-negatives (Pseudomonas, E. coli), MRSA, Candida, Aspergillus, viral (CMV, EBV, RSV)

Neonatal HSV

Neonatal herpes simplex encephalitis/sepsis is life-threatening and may be indistinguishable from bacterial sepsis in the first 2 weeks of life. All septic-appearing neonates should receive aciclovir empirically (20 mg/kg 8-hourly) until HSV is excluded by PCR, particularly if: vesicular rash, maternal history of herpes, seizures, hepatitis, or respiratory failure.

Meningococcal Septicaemia

  • Neisseria meningitidis causes a fulminant, haemorrhagic septicaemia with characteristic non-blanching purpuric rash
  • Can progress to multi-organ failure and adrenal haemorrhage (Waterhouse-Friderichsen syndrome) within hours
  • Pre-hospital benzylpenicillin G (300 mg <1 year, 600 mg 1–9 years, 1200 mg ≥10 years IM/IV) is appropriate if meningococcal disease is suspected and IV access available; do not delay treatment for transfer

Clinical Recognition

Paediatric Sepsis Definitions

The Third International Consensus (Sepsis-3) definitions are validated in adults; the Phoenix Sepsis Criteria (Schlapbach et al, JAMA 2024) represent the most recent validated paediatric-specific definition, based on organ dysfunction scoring in children.

Historically, paediatric sepsis has been defined as SIRS plus suspected infection, and septic shock as cardiovascular failure (requiring fluid/vasopressors) — recognising that the Sepsis-3 framework was not validated in children.

Clinical Signs of Septic Shock in Children

Two clinical phenotypes:

"Warm shock" (hyperdynamic / distributive):

  • Vasodilation: flushed, warm, bounding pulses
  • Wide pulse pressure
  • Fast CRT (<1 second — "flash")
  • Tachycardia, tachypnoea
  • More common in older children and adolescents

"Cold shock" (vasoconstricted):

  • Vasoconstriction: mottled, cool peripheries, poor CRT (>2 sec)
  • Narrow pulse pressure, reduced pulse volume
  • More common in young infants
  • Higher lactate, poorer prognosis

Key Clinical Findings

  • Prolonged capillary refill time (>2 seconds): sensitive early marker in children
  • Mottled skin (marbling pattern): microcirculatory failure
  • Altered mental status, irritability, lethargy
  • Oliguria (<1 mL/kg/h under 12 months; <0.5 mL/kg/h older children)
  • Non-blanching purpuric rash: meningococcal disease until proven otherwise — do not wait for investigations

Management

Immediate Priorities (PALS/APLS framework)

  1. A, B, C in sequence
  2. High-flow oxygen, IV/IO access, blood cultures, BM/blood glucose
  3. Broad-spectrum antibiotics within 1 hour: empirical age-appropriate regimen (see below)
  4. IV/IO fluid bolus: 10 mL/kg 0.9% NaCl or Plasmalyte over 10–20 min; reassess; repeat if required (cautiously — see FEAST)
  5. Glucose: correct hypoglycaemia (2 mL/kg 10% dextrose IV)
  6. Vasopressors if fluid unresponsive
  7. ICU transfer

Antibiotics

Age Empirical regimen
Neonate (<1 month) Penicillin/amoxicillin + gentamicin + consider aciclovir
Infant (1–3 months) Cefotaxime or ceftriaxone + consider ampicillin for Listeria
Child (>3 months) Cefotaxime or ceftriaxone
Septic shock (any age) Broad-spectrum: meropenem or pip-tazo ± vancomycin if MRSA risk
Meningococcal disease Ceftriaxone 80 mg/kg/day (max 4 g)

Vasopressors

  • Adrenaline: first-line for cold shock (↑ HR and contractility + α-mediated vasoconstriction)
  • Noradrenaline: first-line for warm shock (predominantly α₁ vasoconstriction)
  • Dopamine: historically used; more arrhythmogenic; largely replaced by adrenaline in paediatric shock
  • Fluid-refractory septic shock: central access and vasopressor infusion essential; ICU admission

Corticosteroids

  • No mortality benefit from corticosteroids in paediatric septic shock (unlike adults where ADRENAL/APROCCHSS suggest vasopressor-sparing benefit)
  • Consider if: adrenal crisis suspected (meningococcal with adrenal haemorrhage, chronic steroid use, congenital adrenal hyperplasia)
  • Hydrocortisone 1–2 mg/kg/dose IV for suspected adrenal insufficiency

Dexamethasone in Bacterial Meningitis

Dexamethasone 0.15 mg/kg IV 6-hourly for 4 days reduces neurological sequelae (particularly deafness) in bacterial meningitis, especially Haemophilus influenzae type b. Less evidence for Pneumococcal meningitis; give if bacterial meningitis confirmed before or concurrently with antibiotics.

Blood Transfusion Threshold

TRIPICU trial (Lacroix 2007, NEJM): restrictive transfusion strategy (trigger Hb <70 g/L) equivalent to liberal strategy (Hb <95 g/L) in stable, non-cyanotic paediatric ICU patients with haemodynamic stability. Hb threshold of 70 g/L is safe and now standard for stable paediatric ICU patients.

FEAST Trial — Fluid Resuscitation in Children

FEAST (Fluid Expansion As Supportive Therapy) trial (Maitland, NEJM 2011) remains one of the most controversial and paradigm-shifting trials in paediatric critical care.

Design

  • Setting: sub-Saharan Africa (Kenya, Tanzania, Uganda) — resource-limited settings; no ICU
  • Population: 3141 children aged 2 months–12 years with severe febrile illness and shock features
  • Intervention: immediate bolus of 20–40 mL/kg 0.9% NaCl or albumin vs no bolus (maintenance fluids only)
  • Primary outcome: 48-hour mortality

Result

Bolus fluid administration increased 48-hour mortality by ~3.3% absolute (10.5% vs 7.3%) compared to no bolus. Both albumin and saline boluses were harmful.

Interpretation

The FEAST trial caused worldwide debate. The harm from boluses in this setting has been attributed to:

  • Volume overload in children without ICU monitoring or ventilatory support (no PEEP, no rapid diuresis)
  • Pulmonary oedema and cerebral oedema in context of severe malaria (commonest diagnosis) — fluid shifts in a disease with intrinsic vasogenic oedema
  • Rapid intracardiac volume change causing arrhythmia

Applicability to High-Resource Settings

FEAST results should NOT be directly extrapolated to UK/European ICU practice. Key differences:

  • FEAST patients had no ICU, no mechanical ventilation, no vasopressors available
  • Many had severe malaria — cerebral oedema pathophysiology is different from bacterial sepsis
  • In resource-rich settings, fluid boluses are given with close monitoring, rapid vasopressor escalation, and mechanical ventilation available if pulmonary oedema develops
  • UK/RCPCH and ESPNIC guidelines continue to recommend cautious IV boluses (5–10 mL/kg) in paediatric septic shock with reassessment, guided by haemodynamic response
  • Avoid >40 mL/kg without vasopressor initiation and ICU-level monitoring

Specific Conditions

Meningococcal Septicaemia

  • Non-blanching purpuric rash = meningococcal until proven otherwise
  • Airway, high-flow O₂, IV/IO access
  • Ceftriaxone 80 mg/kg IV immediately (pre-hospital benzylpenicillin if ceftriaxone unavailable)
  • Aggressive resuscitation: fluid boluses, early vasopressors (adrenaline)
  • Adrenal support: hydrocortisone if haemodynamically unstable (Waterhouse-Friderichsen)
  • DIC management: FFP, platelets, cryoprecipitate for haemostasis
  • Notifiable disease: public health notification; close contact prophylaxis (rifampicin or ciprofloxacin)

Neonatal Sepsis

  • Subtle presentation: temperature instability (hypothermia as well as fever), poor feeding, respiratory distress, apnoeas, jaundice, mottled skin
  • Full septic screen: blood cultures, LP (if stable), urine culture
  • Empirical treatment: benzylpenicillin (or amoxicillin) + gentamicin ± aciclovir
  • Gentamicin: therapeutic drug monitoring essential (extended interval dosing in neonates; trough <1 mg/L, peak 8–12 mg/L)

Viva Questions

1. How does paediatric septic shock differ from adult septic shock in terms of presentation and management?

Children compensate physiologically much better than adults — the cardiovascular system maintains blood pressure via tachycardia and vasoconstriction until very late in the disease course. Hypotension is a pre-terminal sign in children, not an early indicator. This means clinical recognition depends heavily on tachycardia, altered perfusion (prolonged CRT >2s, mottled skin, cool peripheries), altered mental status, and oliguria — all requiring age-specific norms. Two phenotypes exist: warm shock (hyperdynamic, vasodilatory — treat with noradrenaline) and cold shock (vasoconstricted — treat with adrenaline). Fluid resuscitation is given in smaller boluses (10 mL/kg) with frequent reassessment, informed by the FEAST trial showing harm from large rapid boluses in resource-limited settings — though this cannot be directly extrapolated to monitored ICU settings. Empirical antibiotics must cover age-appropriate organisms (GBS and HSV in neonates; meningococcus and pneumococcus in infants). Corticosteroids have no proven mortality benefit in paediatric sepsis (unlike some adult data). Transfusion threshold of Hb 70 g/L is safe in haemodynamically stable paediatric ICU patients (TRIPICU trial).

2. Describe the FEAST trial and discuss the implications for fluid management in paediatric sepsis.

The FEAST trial (Maitland 2011) randomised over 3000 children in sub-Saharan Africa with severe febrile illness and shock features to bolus saline, bolus albumin, or no bolus (maintenance fluids only). Unexpectedly, both bolus groups had significantly higher 48-hour mortality (~10.5% vs 7.3%). The proposed mechanisms include pulmonary and cerebral oedema in patients without ICU monitoring, mechanical ventilation, or vasopressors — and the specific pathophysiology of severe malaria (the dominant diagnosis), in which brain oedema is a prominent complication. FEAST caused major debate and challenged the universal teaching that bolus fluids improve outcomes in all shock states. However, FEAST should not be directly applied to UK practice: the resource-limited setting without ICU or ventilatory support is fundamentally different from a monitored paediatric ICU where pulmonary oedema can be detected and treated rapidly. Current guidance in high-income settings retains cautious bolus fluid therapy (10 mL/kg, repeated with close assessment) in paediatric septic shock, with early escalation to vasopressors rather than continued fluid boluses beyond 40 mL/kg.

3. A 4-month-old infant presents with fever, poor feeding, tachycardia, prolonged CRT, and mottled skin. What are your immediate priorities?

This infant has clinical signs of septic shock. Immediate priorities: (1) Airway and breathing: high-flow oxygen; assess for respiratory failure — may need early CPAP or intubation if deteriorating; (2) Circulation and access: IO or IV access immediately (EZ-IO if veins not accessible); blood cultures, BM, blood gas; (3) Antibiotics within 1 hour: age 4 months — empirical ceftriaxone 80 mg/kg IV or cefotaxime; add amoxicillin to cover Listeria if not yet 3 months; add aciclovir if HSV suspected (vesicles, hepatitis, maternal history); (4) Fluid: 10 mL/kg Plasmalyte/0.9% NaCl bolus over 10–20 minutes; reassess; repeat cautiously; (5) Glucose: check BM; treat hypoglycaemia with 2 mL/kg 10% dextrose; (6) Vasopressors: if fluid-unresponsive: adrenaline infusion (cold shock — cool mottled peripheries); (7) ICU transfer: this infant needs ICU-level monitoring. Simultaneously consider specific diagnoses: non-blanching rash → meningococcal disease → immediate ceftriaxone; bulging fontanelle → meningitis; notify paediatric registrar/consultant immediately. Don't miss the rare but critical causes: septicaemia, HSV encephalitis, inborn errors of metabolism.