Automated blood-mixing devices still fail to mix at low bleeding rates

The study evaluated mixing performance for various commercial blood-mixing devices at different flow rates.

A glycerol solution with a density equal to blood was pumped into a blood bag (mounted on a mixer) that contained 70 ml of Toluidin Blue in citrate-phosphate-dextrose (CPD). After 500 ml of glycerol solution had been pumped into the blood bag, the bag was emptied in fractions and the absorbance at 640 nm of each fraction (expressed as percentage of mixture absorbance) was plotted against the fraction number.

At a flow rate of 30 ml/min, the curves were very sigmoid, with high initial values that decreased during collection of the fractions. The absorbance values of the fractions collected later (after approximately 250-300 ml) were 10-50% (i.e. percentage of absorbance for complete mixing) for the different mixers, indicating that the CPD solution was incompletely mixed with the incoming solution, resulting in a lower-than-expected 'anticoagulant' content for the fractions collected later. At a flow rate of 50 ml/min the mixing was improved, but only at 75 ml/min did all mixers show relatively good mixing. Manual mixing, by kneading the bag three times/min, gave at all rates an almost ideal mixing curve, 105% initially to 90% in the final fractions.

At relatively high bleeding rates, all mixers performed well, although complete mixing was only obtained with manual kneading. However, most blood-mixing devices still fail to mix efficiently at normal and low bleeding rates. Although the minimal degree of necessary mixing is unknown, further optimization of mixing devices seems warranted.