Intended for healthcare professionals

Letters WHO guidelines on fluid resuscitation in children

Authors’ reply to Southall

BMJ 2014; 348 doi: (Published 19 February 2014) Cite this as: BMJ 2014;348:g1619
  1. Sarah Kiguli, consultant paediatrician1,
  2. Samuel O Akech, research fellow2,
  3. George Mtove, research fellow3,
  4. Robert O Opoka, consultant paediatrician1,
  5. Charles Engoru, consultant paediatrician4,
  6. Peter Olupot-Olupot, consultant paediatrician5,
  7. Richard Nyeko, consultant paediatrician6,
  8. Jennifer Evans, consultant paediatrician7,
  9. Jane Crawley, consultant paediatrician8,
  10. Natalie Prevatt, research fellow9,
  11. Hugh Reyburn, senior lecturer in clinical epidemiology10,
  12. Michael Levin, consultant paediatrician9,
  13. Elizabeth C George, statistician11,
  14. Annabelle South, policy, communications and research impact coordinator11,
  15. Abdel G Babiker, statistician11,
  16. Diana M Gibb, professor of epidemiology11,
  17. Kathryn Maitland, professor of tropical paediatric infectious disease9
  1. 1Department of Paediatrics, Mulago Hospital, Makerere University, Kampala, Uganda
  2. 2Kilifi Clinical Trials Facility, KEMRI-Wellcome Trust Research Programme, Kilifi, Kenya
  3. 3Department of Paediatrics, Joint Malaria Programme, Teule Hospital, Muheza, Tanzania
  4. 4Department of Paediatrics, Soroti Regional Referral Hospital, Soroti, Uganda
  5. 5Department of Paediatrics, Mbale Regional Referral Hospital Pallisa Road, Mbale, Uganda
  6. 6Department of Paediatrics, St Mary’s Hospital, Lacor, Uganda
  7. 7Department of Paediatrics University Hospital of Wales Heath Park, Cardiff, UK
  8. 8Nuffield Department of Medicine, University of Oxford, Oxford, UK
  9. 9Wellcome Trust Centre for Clinical Tropical Medicine, Department of Paediatrics, Faculty of Medicine, St Mary’s Campus, London W2 1PG, UK
  10. 10Joint Malaria Programme, Moshi, Tanzania
  11. 11Medical Research Council Clinical Trials Unit, UCL, London, UK
  1. k.maitland{at}

Southall made several points about our recent article.1 2

He suggests that “lethal hyperchloraemia” secondary to use of normal saline in FEAST (for boluses or maintenance) resulted in excess mortality. However, he did not comment on the key finding of the trial—that the increased 48 hour mortality was identical in both normal saline bolus (10.6%) and albumin bolus (10.6%) arms compared with the no bolus control group (7.3%).3 Harm was shown for every age group, in every condition, at each of the six hospitals, regardless of degree of acidosis,4 and for all definitions of shock.5 Despite differences in physical properties of albumin and saline, the timing of excess mortality after administration was identical (Kaplan-Meier mortality curves),3 making his hypothesis improbable.

With regard to maintenance fluids, use in the trial was pragmatic, given either as part of a quinine infusion (World Health Organization recommends 5% dextrose) or the local “standard of care” maintenance fluid, and stopped as soon as the child could retain oral fluids. The median volume of maintenance fluid received over 48 hours was 50 mL/kg (interquartile range 35.3-60) alongside quinine infusion or 51.8 mL/kg (26.1-83.2) as maintenance. The chief maintenance fluid was 5% dextrose (Uganda), 5% dextrose/0.9% saline (Teule), or 5% dextrose/0.45% saline (Kilifi). Data on fluid volumes are already in the public domain, published alongside the New England Journal of Medicine manuscript (table 3a supplementary files).3 Details of the diagnoses of the deaths by arm are listed in supplementary files (table 2b: Endpoint review committee adjudication of deaths).

Finally, Southall indicates that all children in the FEAST trial were exposed to harm (whether from normal saline or hypotonic 5% dextrose), even though these are used widely across Africa. However, global mortality in FEAST (all three arms) was substantially better than predicted by previous studies, including the quinine arm of the AQUAMAT malaria trial (10.9%)6 versus FEAST malaria subgroup bolus arms (9.2%) and control (5.7%).5

Most children in the bolus arms received only 20 mL/kg, lower than is currently recommended by WHO and other international guidelines. Even this relatively conservative volume of fluid boluses resulted in harm, which is why we are calling for WHO to reverse its liberal bolus guidance for febrile sick African children with shock not caused by fluid loss. Current WHO guidance is costly, difficult to implement (most African hospitals cannot accurately measure fluid boluses), and unsupported by evidence.


Cite this as: BMJ 2014;348:g1619


  • Competing interests: None declared.


Log in

Log in through your institution


* For online subscription