Comparison of transfusion thresholds during neonatal extracorporeal membrane oxygenation

Alexandra A. Sawyer, Linda Wise, Santu Ghosh, Jatinder Bhatia, Brian K. Stansfield

Research output: Contribution to journalArticlepeer-review

18 Scopus citations


BACKGROUND: The objective of this study was to assess complications and patient outcomes associated with a lower reflexive red blood cell (RBC) transfusion threshold for neonates undergoing extracorporeal membrane oxygenation (ECMO) for hypoxic respiratory failure. STUDY DESIGN AND METHODS: A retrospective cohort study was conducted at a single tertiary neonatal intensive care unit of neonates undergoing ECMO support for refractory hypoxic respiratory failure for more than 24 hours between December 2009 and December 2014. Seventy-two neonates received ECMO support for hypoxic respiratory failure for longer than 24 hours during the study period. Patient cohorts were determined based on transfusion threshold of hematocrit (Hct) level of less than 40% (December 2009-October 2012) and Hct level of less than 35% (November 2012-December 2014). RESULTS: Patients who had a lower threshold for transfusion (Hct < 35) had a lower mean Hct (38.3% vs. 41.4%, p < 0.0001) and received less total RBC transfusion volume (10.4 mL/kg/day vs. 13.3 mL/kg/day, p = 0.002) while undergoing ECMO support. Survival off ECMO, survival to discharge, and complication rates were similar between the cohorts. CONCLUSIONS: A lower Hct threshold of 35% is associated with a reduction in RBC transfusion volume and does not appear to alter complication rates or patient outcomes for neonates receiving ECMO support for respiratory failure.

Original languageEnglish (US)
Pages (from-to)2115-2120
Number of pages6
Issue number9
StatePublished - Sep 2017

ASJC Scopus subject areas

  • Immunology and Allergy
  • Immunology
  • Hematology


Dive into the research topics of 'Comparison of transfusion thresholds during neonatal extracorporeal membrane oxygenation'. Together they form a unique fingerprint.

Cite this