FEAST; a review

FEAST: fluid expansion as supportive therapy


What was the point?

Sepsis is a common presentation in children, more so in poorer countries. Initial resus fluid boluses are standard in the treatment of this condition in richer countries but (according to introduction) less common in poorer countries. There is no definitive answer on the amount of resus fluids that are appropriate


How was it done?

  • Open 2 stratum, 3 armed randomised open randomised tria

Stratum A- without severe hypotension

Arm 1-  20mls 0.9 % saline/kg over 1 hour

Arm 2- 20mls albumin/kg over 1 hour

Arm 3- No bolus (control )

Stratum B- Severe hypotension

Same arms but with 40mls/kg bolus

No control group (due to ethical concerns)


Further 20mls/kg given in all groups (except control) if impaired perfusion persisted.


Where and on whom was it done?

  • 6 East African hospitals

Inclusion criteria

  • Children aged 2 months- 12 years
  • Severe febrile ilness
  • Impaired consciousness and/or respiratory distress
  • And impaired perfusion (cap refill >3 seconds, weak pulse or severe tachycardia)

Exclusion criteria

  • Severe malnutrion
  • Non-infective shock
  • Gastroenteritis

What were they testing and what did they compare it to?

Primary End point; 48 hour mortality

Secondary End points; 4 week mortality, neurological squelae, episodes of hypotensive shock, adverse end points e.g pulmonary odema


The numbers that matter

Lower risk of death in control group

Absolute risk difference 3.3 percentile points(10.5% risk in combined bolus group vs 7.3% in no bolus group)


Author’s Conclusions

This study does not support rountine use of fluid boluses in severely ill febrile children in the African setting.

Good things

Hard end point (death)

Power analsysis for sample size (adjusted due to lower than expected morality)

Intention to treat analsysis used

Very little crossover

Broad inclusion criteria increase real world applicability


Less good things

Very sick cohort (judging by patient characteristics) in very different environment to what I work in.

Incidence of malaria, with a very different pathophysiology to bacterial sepsis, could affect his analysis


Implications and the road forward

Demonstrates the high quality research that can be undertaken in poorer countries and suggest some of the dogma surrounding fluid resus may be misplaced, especially considering the sub-group did not identify any groups that may be of benefit. However, due to this cohort of patients being treated very differential in the environments which I work I cannot personal change my clinical practice.

On a positive note, it does suggest the local standard of care (no initial fluid bolus) is better than importing the rich world treatment mobile. Even in this extreme setting less medicine appears to be better


If you don’t believe me…








Leave a Reply