Effects of Red Blood Cell Transfusions on Morbidity and Mortality in Non-Septic Critically Ill Patients; A Propensity Score Matched Study

Abstract

BackgroundRBC-transfusions can be lifesaving, but are also associated with harm. To further examine any effect of red blood cell (RBC)-transfusions given to critically ill patients that were not exposed to the risks of anemia or sepsis, we designed this retrospective propensity score matched study. The aim was to compare mortality and morbidity in non-septic critically ill patients that were given low-grade RBC-transfusions at hemoglobin level > 70 g/L with patients without RBC-transfusions any of the first 5 days in intensive care.MethodsAdult patients admitted to a general 9-bed intensive care unit between 2007-2018 at a tertiary university hospital, were eligible for inclusion. Patients that received > 2 units RBC-transfusion during the first five days after admission, with pre-transfusion hemoglobin level <70 g/L or with severe sepsis or septic shock were excluded. Outcomes were 28-, 90- and 180-day mortality, highest acute kidney injury network (AKIN) score, days alive and free of organ support the first 28 days and highest sequential organ failure assessment score (SOFA-max).ResultsIn total 9491 admissions were recorded during the study period. Propensity score matching at 1:1 ratio resulted in two well matched group with 682 unique patients in each. Median pre-transfusion hemoglobin was 98 g/L (interquartile range 91-106 g/L). Mortality at the measured time points were higher in the RBC-group with an absolute risk increase for death at 180 days of 4.8% [95% confidence interval 2.5 to 7.2%], (p<0.001). Low grade RBC-transfusion was also associated with renal, circulatory and respiratory failure as well as higher SOFA-max score. ConclusionLow-grade leukoreduced RBC-transfusions given to non-septic critically ill patients without significant anemia, was strongly associated with increased mortality, increased kidney-, circulatory- and respiratory- failure as well as with higher SOFA-max score. These findings further strengthen the evidence supporting a restrictive use of RBC-transfusions in critically ill patients.