Blood transfusions, blood storage, and correlation with elevated pulmonary arterial pressures

Marilynn Cullison, Richard Mahon, Gerald McGwin, Richard McCarron, Robert Browning, Charles Auker*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


BACKGROUND: The aim of this study was to determine if transfusion with RBCs is associated with a rise in mean pulmonary artery pressure (MPAP) and whether such a rise is influenced by the duration of RBC storage. STUDY DESIGN AND METHODS: A retrospective chart review of intensive care unit patients with pulmonary artery catheters was conducted at two military medical centers. RESULTS: RBC transfusion is associated with a sustained (≥4 hours) statistically significant 2- to 3-mm Hg rise in MPAP relative to both pretransfusion levels (p < 0.05) and compared to asanguinous fluid infusions (p < 0.05). The magnitude of the rise (all infusions, RBCs, and asanguinous) correlates positively with in-hospital mortality (p < 0.01) and hospital length of stay (p < 0.01). The duration of RBC storage was not statistically correlated with the magnitude of rise in the population studied. Mean infusion volume was greater for RBC (vs. asanguinous) infusions, but volume adjustment of MPAP values did not alter the pattern or statistical significance of the results. CONCLUSIONS: Analysis of retrospectively collected data suggests that transfusion of RBC-containing fluids results in a sustained elevation of MPAP. In the patient population studied, the duration of RBC storage did not correlate with the magnitude of MPAP rise. Future prospective studies of transfusion effects should consider including assessment of MPAP and subpopulation analyses.

Original languageEnglish
Pages (from-to)1259-1266
Number of pages8
Issue number4
StatePublished - Apr 2019
Externally publishedYes


Dive into the research topics of 'Blood transfusions, blood storage, and correlation with elevated pulmonary arterial pressures'. Together they form a unique fingerprint.

Cite this