By the same authors

From the same journal

Transfusion Volume for Children with Severe and LifeThreatening Anaemia

Research output: Contribution to journalArticle

Published copy (DOI)


  • Kathryn Maitland
  • Peter Olupot-Olupot
  • Sarah Kiguli
  • George chagaluka
  • Florence Aloroker
  • Robert Opoka
  • Ayub Mpoya
  • Charles Engoru
  • J Nteziyaremye
  • Macpherson mellewa
  • Pedro Rafael Saramago Goncalves
  • Elizabeth George
  • Diana M. Gibb
  • Sarah Walker


Publication details

JournalNew England Journal of Medicine
DateAccepted/In press - 17 Jun 2019
DatePublished (current) - 1 Aug 2019
Number of pages12
Pages (from-to)420-431
Original languageEnglish


Background: Severe anaemia (haemoglobin<6g/dl) is a leading cause of hospital admission and mortality in children in sub-Saharan Africa. World Health Organization recommends 20mls/kg whole blood equivalent regardless of haemoglobin, which standard calculations suggest is insufficient for deficit-correction. Methods: The TRACT factorial open-label trial randomized Ugandan and Malawian children aged 2 months to 12 years with haemoglobin <6g/dl and severity signs to immediate transfusion with 20mls/kg vs 30mls/kg whole blood equivalent. Three other randomizations investigated triggered transfusion, post-discharge micronutrients and/or cotrimoxazole. The primary endpoint was 28-day mortality. Results: 3196 eligible children (median 37 months; 2050(64%) with malaria) were randomized to 30mls/kg (n=1598) or 20mls/kg (n=1598) transfusion and followed for 180-days (134(4%) lost-to-follow-up). 1592(99.7%) 30mls/kg vs 1596(99.7%) 20mls/kg started transfusion, both a median 1.2h from randomization. Mean(standard deviation) total blood transfused was 475(385) vs 353(348) mls respectively. Haemoglobin recovery was significantly faster and greater with 30mls/kg (p<0.001). 55(3%) 30mls/kg vs 72(5%) 20mls/kg children died before 28-days (hazard ratio[HR]=0.76 (95% CI 0.54-1.08) p=0.13), and 134(8%) vs 154(10%) respectively before 180-days (HR=0.86 (0.68-1.09) p=0.21). Nevertheless, this masked significant heterogeneity (p≤0.004) at both time-points for fever (>37.5C) at screening. 30mls/kg reduced mortality in the 1943(61%) children without fever (28-day HR=0.43 (0.27,0.69) p=0.001), but increased mortality in the 1253(39%) children with fever (HR=1.91 (1.04,3.49) p=0.04). There was no evidence of differences between groups in re-admissions (p=0.38), serious adverse events (p=0.58) nor in haemoglobin recovery at 180-days (p=0.10). Conclusions: Mortality could be reduced by transfusing 30mls/kg whole blood equivalent in children presenting with severe anaemia without fever.

Bibliographical note

© 2019 Massachusetts Medical Society. All rights reserved.

Discover related content

Find related publications, people, projects, datasets and more using interactive charts.

View graph of relations