Abstract
Blood collection tubes containing citrate lower pH, thereby inhibiting glycolysis. When compared to other additives, they introduce an over-estimation in measured glucose. This study explored this over-estimation across a range of glucose values. Blood samples collected into lithium-heparin tubes then cooled prior to immediate refrigerated plasma separation, were used as the primary comparator.
Venous blood from individuals with and without diabetes was collected into tubes containing lithium-heparin, or fluoride, or fluoride-citrate (Terumo™ Venosafe). Plasma was separated at time intervals of zero, 2 and 24h. Preparation of the ‘time zero’ lithium-heparin and fluoride samples was optimised by processing these samples under cooled conditions. The remaining samples were prepared at room temperature. Plasma was analysed in the routine clinical laboratory using the hexokinase method.
Median plasma glucose for the 50 participants was 7.1mmol/L (range 3.1–21.5). At ‘time zero’, fluoride-citrate glucose was 0.37mmol/L (95% CI 0.26–0.48) higher than lithium-heparin glucose and 0.29mmol/L (95% CI 0.21–0.36) higher than glucose from fluoride tubes. Following delayed plasma separation at 24h, glucose loss from the lithium heparin tubes averaged 0.2mmol·L−1·hr−1. In contrast, the fluoride-citrate tubes showed minimal glucose loss over 24h.
Acid stabilises glycolysis but causes an over-estimation in glucose, across a range of plasma glucose values, when compared to blood collected into conventional tubes under cooled conditions. The magnitude of the over-estimation seen with the fluoride-citrate tubes is unlikely to be due solely to the differential glucose stabilisation rates of acid, compared to cooling.
•Terumo™ Venosafe citrate tubes ‘read’ glucose higher (0.37mmol/L) than research grade preparation with lithium heparin tubes•When plasma separation is delayed, citrate blood collection tubes stabilise glucose over a wide range of glucose values•Pre-analytical glucose loss over 24 hr best fits a model of absolute loss (around 0.2mmol.L-1.hr-1) rather than % loss