Which sample tube should be used for routine glucose determination

Abstract Background Glucose is one of the most frequently requested analytes in clinical laboratory. Blood glucose analysis is affected from in vitro glycolysis. In order to determine the most suitable blood collection tube for this purpose we have compared different tubes: sodium fluoride, lithium heparin, sodium fluoride/citrate buffer containing tubes and serum with clot activator tube for the measurement of glucose when the tube has been kept at room temperature (RT) for up to 4h. Methods Venous blood was collected from 49 healthy volunteers into Sarstedt S-Monovettes for glucose analysis. Reference plasma glucose was determined in a lithium heparin tube and immediately placed in an ice/water slurry. Within 10min it was centrifuged at 4°C and plasma was separated from the blood cells. Samples have been preserved at RT for 1, 2 and 4h after drawing. Glucose has been determined using a hexokinase method. Results Glucose levels tested in a serum with clot activator tube, in lithium heparin and in sodium fluoride/sodium EDTA tubes when compared with lithium-heparin reference plasma did not meet the desirable bias for glucose (±1.8%) when kept at RT for up to 4h. GlucoEXACT tubes, when corrected by the Sarsted recommended factor of 1.16, showed a mean (95% CI) bias of +0.96% (0.45–1.47) at 1h, +1.40% (0.88–1.93) at 2h and +0.95% (0.44–1.46) at 4h, reaching the analytical goal for the desirable bias. Conclusions Samples collected into GlucoEXACT tubes containing sodium fluoride/citrate buffer liquid mixture are equivalent to those collected in reference plasma tubes avoiding glycolysis completely and within a 4h delay in plasma separation.