Background: Studies examining one-year mortality respecting component blood transfusion are sparse. We hypothesize that component blood product transfusions are negatively associated with 90-day and 1-year survival for all patients requiring veno-arterial (VA) or veno-venous (VV) ECMO.
Study design and methods: This was an IRB-approved retrospective cohort analysis of 676 consecutive patients requiring ECMO at the University of Pittsburgh between 2005 and 2016. Patients were analysed both as an entire cohort and as two subsets with respect to ECMO modality (VA vs. VV). Additional data collected and analysed included patient characteristics, laboratory values and blood product transfusion.
Results: Multivariable analysis revealed that platelet transfusion was associated with 90-day mortality (OR: 1·05, P = 0·037) and one-year mortality for the entire cohort (OR = 1·05, P = 0·046,). Platelet transfusion volume was also associated with mortality in the VA-ECMO subset of patients at both 90 days (OR = 1·08, P = 0·03) and one year (OR: 1·11, P = 0·014). Age, peak International Normalized Raton ECMO, nadir haemoglobin (on ECMO) and final haemoglobin (after ECMO) were significantly associated with mortality for patients requiring VA-ECMO. For VV-ECMO patients, age, INR and peak creatinine on ECMO were associated with mortality. No individual component blood product was associated with one-year mortality for patients requiring VV-ECMO.
Conclusion: Platelet transfusion was associated with increased 90-day and 1-year mortality for patients requiring VA-ECMO.
Keywords: blood component production; platelet transfusion; transfusion - surgery; transfusion medicine; transfusion strategy.
© 2020 International Society of Blood Transfusion.