Thresholds for Transfusion Practice During ECMO Support. A Systematic Review and Network Meta-Analysis.

Abstract

Purpose: The benefits of adopting restrictive transfusion strategies, as opposed to a liberal approach, during extracorporeal membrane oxygenation (ECMO) remain unclear. This network meta-analysis aims to determine whether a restrictive transfusion strategy is more effective than liberal thresholds in minimizing red blood cell (RBC) transfusions and improving secondary outcomes.

Methods: A comprehensive literature search (PROSPERO-CRD42025637794) across Medline, Embase, and Scopus was conducted. All studies enrolling adults on ECMO, receiving RBC transfusions according to either restrictive or 'targeted' thresholds, compared to liberal thresholds, were included. The primary outcome was the number of RBC transfusions during ECMO. Secondary outcomes included the use of fresh frozen plasma (FFP) and platelet units, ECMO duration, and survival.

Results: Five retrospective observational studies (1339 patients) met the inclusion criteria. RBC transfusions were lower when a transfusion threshold of 7 g/dL was used, compared to a liberal threshold (mean difference (MD) -5.75, 95% confidence interval (CI) -10.90 to -0.59, p = 0.029), while not with thresholds of 8 or 9 g/dL. Both FFP and platelet transfusions were reduced at thresholds of 7 g/dL and 9 g/dL, compared to liberal transfusion strategies. A shorter ECMO duration was observed only at a threshold of 9 g/dL (MD -1.06, 95% CI -2.11 to -0.01, p = 0.048). Finally, a restrictive threshold of 7 g/dL improved 28-day survival.

Conclusions: A restrictive transfusion strategy reduces the number of blood products administered to ECMO patients and, secondarily, ECMO duration and mortality at 28 days. However, our findings may not be generalizable to patients with severe thrombocytopenia, bleeding disorders, or underlying cardiac conditions who may be potentially benefiting from higher transfusion thresholds.