A recently published study in the Lancet continues to reinforce the view that less is more for blood transfusions.
Jairath and colleagues tested whether a restrictive approach (in which patients were transfused when hemoglobin concentration fell below 80 g/L) versus a liberal approach (in which transfusion was initiated when the hemoglobin concentration fell below 100 g/L). The liberal approach was not superior to the restrictive approach, suggesting we should tolerate lower hemoglobin levels in patients before starting blood transfusions.
As far as I know, most studies along these lines have shown either improved outcomes or no difference when restrictive transfusion thresholds are compared to liberal ones. These data support a less interventionist perspective on blood transfusion.