A comparison of venous blood gas analysis with paired specimens collected in syringes and evacuated blood collection tubes

Background: Venous blood samples collected in evacuated blood collection tubes are generally not considered suitable for measurement of oxygenation status. However, whether pH, pCO₂ and HCO₃^- results from venous blood samples collected in evacuated tubes are reliable remains unclear.

Methods: Paired samples were collected from 38 healthy volunteers using two collection methods: (1) collection of blood directly into a blood gas syringe and (2) collection of blood into an evacuated tube with subsequent anaerobic transfer into a blood gas syringe. Samples were analyzed for pH, pCO₂, HCO₃^-, and pO₂.

Results: Samples collected in evacuated tubes showed significant positive mean biases of 0.03 and 7.5 mmHg for pH and pO₂, respectively, and significant negative mean biases of 5.0 mmHg and 1.2 mmol/l for pCO₂ and HCO₃^-, respectively. Collection in evacuated tubes resulted in biases exceeding the total allowable errors in 16%, 40%, 21% of samples for pH, pCO₂, HCO₃^-, respectively, when compared to samples collected in syringes directly.

Conclusions: Samples for venous blood gas analysis should be collected directly into blood gas syringes for accurate assessment of ventilation and/or acid-base status. Biases observed for samples collected in evacuated tubes are consistent with air contamination and/or vacuum effects.