Effects of sample handling on total carbon dioxide concentrations in canine and feline serum and blood

Objective - To determine whether underfilling blood collection tubes leads to in vitro reduction in serum measured total CO₂ concentration ([TCO₂]_m) in canine and feline blood samples sufficient to create the impression of metabolic acidosis (pseudometabolic acidosis) or high anion gap. Sample Population - Blood samples from healthy client-owned animals (16 dogs, 17 cats). Procedure - Venous blood samples were collected in random order for determination of serum [TCO₂] and blood gas tensions. Blood gas analysis was performed on iced, capped blood samples. In dogs, serum [TCO₂] was measured in 1-, 3-, and 10-ml samples in 10-ml type-B tubes and in a 3-ml sample in 3-ml type-A tubes. In cats, serum [TCO₂] was determined in 1-, 2-, and 3-ml samples in 3-ml type-A tubes and in a 3-ml sample in 10-ml type-B tubes. Results - For dogs, serum [TCO₂] in full-tube, 10-ml samples was a mean ± SD, 2.0 ± 1.1 mmol/L greater than that in 3-ml samples and 3.7 ± 1.3 mmol/L greater than the value in 1-ml samples; both differences were significant at P < 0.0001. The serum [TCO₂] in full 3-ml samples was lower by 0.4 ± 0.6 mmol/L than the value in full-tube 10-ml samples (P=0.019). For cats, serum [TCO₂] in full-tube, 3-ml samples was 0.5 ± 0.6 mmol/L greater than that in 2-ml samples (P = 0.004) and was 1.5 ± 0.8 mmol/L greater than the value in 1-ml samples (P < 0.0001). Serum [TCO₂] in 3-ml samples of feline blood in 10-ml tubes was 0.8 ± 0.8 mmol/L lower than that in samples from full 3-ml tubes (P = 0.0007). In dogs and cats, [TCO₂] in fully filled collection tubes was approximately 6 mmol/L higher when calculated from blood gas analysis data than when chemically determined in serum. Conclusions and Clinical Relevance - Underfilling blood collection tubes results in a false decrease in serum [TCO₂], which can contribute in part to descrepancies between blood gas and chemical analyses as estimates of plasma bicarbonate concentration. This, and other in vitro effects of sample handling and collection, may result in a false assessment of metabolic acidosis in dogs and cats.