University of Michigan
Department of Pediatrics

Evidence-Based Pediatrics Web Site

Erythropoietin Does Not Reduce Early Transfusion (<10 DOL) in ELBW Infants



  • Does erythropoietin decrease need for early transfusion in extremely low birth weight (ELBW) infants?

Clinical Bottom Lines

  1. There was not a statistically significant difference between erythropoietin treated and control groups in need for transfusion (p=.08).
  2. Erythropoietin treated infants were, however, more likely to have hematocrit >30 than were control infants (p=<0.01)
  3. Early EPO treatment (DOL#3) did not result in an increased reticulocyte count until ~14 days of life.  Thus Epo is more important in decreasing the number of late transfusions (after 10d of age).1

Summary of Key Evidence

  • A prospective, randomized, control trial put ELBW infants into 3 groups on DOL#3: early EPO group included 74 infants that received EPO/Fe starting 1st week of life to 9 weeks of life, late EPO group included 74 infants that received EPO/Fe starting on 4th week of life to 10 weeks of life, and control group consisting of 71 infants that received iron.1
  • The intervention groups were given 750IU/kg/wk in 3 divided doses and iron--started on 3mg/kg/day and adjusted according to transferrin saturation.
  • Success was defined as no transfusion and a maintained Hct of >30%. Rates were as follows: early treatment group – 13%, late treatment group – 11%, control group – 4%. Significant difference between early treatment group and control (P=.03).
  • Crude incidence of transfusion for whole study period was: 74% Early, 82% Late, and 84% control.
  • Whereas there was a statistically significant difference in “success rate” between the early and control groups, there was not a statistically significant difference in need for transfusion between these groups (p=.08).1

Additional Comments

  • Blood transfusion puts infants at risk for electrolyte abnormalities, it adversely affects fluid status and can expose a relatively immunocompromised premature infant to infectious agents.2
  • Average blood volume of neonate is 90cc/kg. Study done at UCSF NICU found average of 38.9ml of blood drawn for lab studies in first week of life. Thus it may be more realistic to limit the number of transfusions/donors rather than eliminate transfusions. This can be achieved in several ways: 1) establishing transfusion criteria; 2) decreasing blood gas sampling by using micro lab technique and in-line blood gas sampling; and 3) using autotransfusion from placental blood.
  • In addition to RBC production, EPO has many beneficial affects in the brain: Decrease in glutamate toxicity: increases production of neuronal anti-apoptotic factors, increases glutathione peroxidase activity, decreases nitric oxide mediated injury, and may regulate cerebral blood flow after hypoxic injury.3


    1. Maier RF et al Early treatment with erythropoietin beta ameliorates anemia and reduces transfusion requirements in infants with birth weights below 1000 g. J Pediatr 2002;141:8-15.
    2. Galel SA, Fontaine MJ. Hazards of neonatal transfusion. Neoreviews. 2006. e69-73.
    3. Juul S. Recombinant erythropoietin as a neuroprotective treatment: in vitro and in vivo models. Clin Perinatol 2004;31:129-42.

CAT Author: Rob Rock, MD

CAT Appraisers: Robert Schumacher, MD

Date appraised: June 21, 2006

Last updated July 9, 2007
Department of Pediatrics and Communicable Diseases
© 1998-2002 University of Michigan Health System