Serum bicarbonate declines with sample size in vacutainer tubes.

Serum bicarbonate testing has been reported to be inaccurate in several clinical studies. This prompted a search for systemic errors in bicarbonate testing devices, which has been inconclusive. The effect on serum bicarbonate level measurement of partially filling plain red-top (clot) vacutainer tubes was studied in healthy volunteers. First, 10-mL vacutainers were filled with 9.5 mL, 6 mL, 3 mL, and 1 mL of identical blood. Second, pediatric (3 mL) vacutainers were filled with 3 mL, 1 mL, and 0.5 mL of identical blood. The serum bicarbonate level was measured on an automated stat analyzer. Linear regression with dummy variables for subjects showed a significant reduction in bicarbonate reading as volume in the vacutainer decreased. Consistent decrease was noted regardless of whether samples were analyzed immediately after clotting (10 minutes) or at 20, 30, and 40 minutes after withdrawal of blood. This decrease was evident regardless of whether blood was aspirated directly into a vacutainer or first aspirated into a syringe and injected into a vacutainer. Underfilling of vacutainers may account for low bicarbonate readings and thus vacutainers should be filled completely to obtain accurate measurements of serum bicarbonate levels.