The impact of different sample matrices in delayed measurement of glucose.

Objectives: This study was designed to compare glucose values in serum or lithium-heparin samples immediately centrifuged with those paired specimens collected in tubes containing Naf-KOx and centrifuged and analyzed 2.5h after collection.

Methods: Three blood samples were drawn from 20 volunteers. Blood samples collected in tubes with and clot activator and gel separator but without anticoagulant (SST) as well as those collected in tubes containing Lithium-Heparin and gel separator were centrifuged within 30min and analyzed 2h thereafter. Blood samples drawn in tubes containing the glycolysis inhibitor NaF-KOx were centrifuged after 2.5h and then analyzed.

Results: The glucose median value was 4.72mmol/L in SST tubes, 4.67mmol/L in lithium-heparin and 4.44mmol/L in NaF-KOx tubes. The difference between SST and lithium-heparin tubes was not statistically or clinically significant, whereas that between SST and Naf-KOx tubes was both analytically and clinically meaningful, exceeding the current quality specifications for glucose measurement.

Conclusions: The rapid centrifugation of blood collected in serum or lithium-heparin tubes with gel separator is seemingly more reliable for delayed measurement of glucose compared to the use of blood tubes containing NaF-KOx.