Restricted Fluid therapy in early septic shock: results of the fluid in shock (FiSh) pilot trial

Inwald DP. Arch Dis Childhood 2018; Online first 07 Aug 2018. doi: 0.1136/archdischild-2018-314924

Clinical Question

• In children with septic shock, does a restricted fluid bolus volume, compared with a standard fluid bolus volume, improve outcomes?

Background

• Rapid liberal fluid bolus therapy for paediatric septic shock is standard practice in high income countries
• The FEAST trial, set in Africa, demonstrated an increased mortality with this approach
• No previous RCTs have compared restrictive vs. standard fluid boluses in a high income country

Design

• Randomised controlled trial
• 1:1 randomisaiton
• Open-label
• Pilot RCT without defined primary outcome, hence without usual power calculation
• Intention to treat analysis

Setting

• 4 regional hospitals and 9 district general hospitals in England
• Acute assessment area (e.g paediatric assessment unit or ED)
• Data collected: July 2016 – April 2017

Population

• Inclusion:
  • Age: older than 37 weeks corrected gestational age and < 16 years
  • Clinical suspicion of infection
  • Shock: age-adjusted systolic BP <5th centile and/or capillary refill time (CRT) >= 3s after 20ml/kg of bolus fluid
• Exclusion: Prior receipt of >20ml/kg bolus fluid, conditions in which bolus fluid resuscitation should be curtailed; full active resuscitation not within current goals of care
• 75 patients randomised
• Comparing baseline characteristics of intervention vs. control groups
  • Median age in months (IQR): 11 (1-35) vs. 2 (1-17)
  • Male: 62% vs. 53%
  • Median weight in kgs (IQR): 9 (5-13) vs. 5 (4-10)
  • Shock criteria
    • CRT only: 77% vs. 88%
    • Hypotension only: 5% vs. 6%
    • CRT & hypotension: 3% vs. 6%
    • Neither: 15% vs. 0%
  • Paediatric Index of Mortality 2 score: 2.1 vs. 2.0
  • Infection confirmed: 51% vs. 50%
    • Bacterial infection: 31% vs. 50%
    • Viral infection: 63% vs. 50%
    • Malaria: 6% vs. 0%

Intervention

• • Restrictive fluid boluses (n=39)
    • 10ml/kg bolus (500ml maximum)
    • 8% of patients did not receive their 1st bolus because deemed no longer required fluid bolus post randomisation

Control

• Standard fluid bolus (n=34)
  • 20ml/kg (1000ml maximum)
  • All patients received their 1st fluid bolus
  • Of their subsequent boluses, 1 patient received a fluid bolus when the shock criteria were not met, and 2 patients did not receive a bolus when the shock criteria were met

Management common to both groups

• 4 hour resuscitation period. One fluid bolus given at 15 minutes intervals
• At end each 15 minutes, if age adjusted signs of shock persisted then another bolus given
• if no signs of shock or patient developed signs of fluid overload (pulmonary oedema, new/increasing hepatomegaly) then fluid boluses were witheld
• Fluid type and other interventions at discretion of treating clinician
• After 4 hour resuscitation period treatment at discretion of treating clinician

Outcome

• Outcomes – comparing intervention vs. control group
  • Total volume study fluid, mean (SD) during
    • 1st hour: 13.5ml/kg (8) vs. 20.7ml/kg (8.3)
    • 1st 4 hours: 14.5ml/kg (11.1) vs. 25.7ml/kg (12), p<0.001
  • No significant difference in:
    • Number of fluid boluses administered
      • 0: 8% vs. 0%
      • 1: 59% vs. 74%
      • 2: 21% vs. 18%
      • 3: 8% vs. 3%
      • >3: 5% vs. 6%
    • Bolus given within 15 minutes: 81% vs. 68%
    • Mortality at 30 days: 0% vs. 0%
    • Serious adverse events: 0% vs. 0%
    • Admitted to PICU: 26% vs. 32%
    • Mechanical ventilation: 11% vs. 25%
    • Receipt of inotropes: 3% vs. 16%

Authors’ Conclusions

• A larger FiSH trial with the current design in the UK is not feasible

Strengths

• Randomised controlled trial
• Multi-centre

Weaknesses

• Pilot study, no primary outcome or power calculation
• Open-label
• Small number of patients included with significant baseline differences
• 6 patients recruited who did not meet inclusion criteria- All randomised to intervention arm
• Lack of equipoise among some clinicians identified during interviews with staff, favouring restrictive fluid boluses due to concerns over fluid overload

The Bottom Line

• Paediatric patients with septic shock in the UK who were randomized to a restrictive fluid bolus strategy received less fluids that patients randomized to a standard fluid bolus
• No significant differences were seen in any of the patient centred outcomes studied in this pilot trial
• A low level of severe illness in the population studied means that detecting differences in mortality would not be feasible with a larger trial in the UK

External Links

• [article] Restricted fluid bolus volume in early septic shock: results of the Fluids in Shock pilot trial
• [further reading] Mortality after fluid bolus in African Children with Severe infection
• [further reading] Exploring mechanisms of excess mortality with early fluid administration: insights from the FEAST trial

Metadata

Summary author: @davidslessor
Summary date: 12th September 2018
Peer-review editor: Segun Olusanya