Transforming the Venous Blood Gas to the Arterial Blood Gas from Concurrent Measurements

KEY POINTS: Here, we report the relationship between arterial and venous blood gas parameters (pH, pressure of carbon dioxide, and bicarbonate) on the basis of 5419 samples collected concurrently. The data show that the venous blood gas can be transformed with reasonable confidence intervals to arrive at a corresponding arterial blood gas. BACKGROUND: Arterial blood gas (ABG) analysis remains the gold standard for diagnosis and management of acid-base disorders. Obtaining arterial blood through arterial puncture, however, is cumbersome, at times difficult and carries potential risks. For these reasons, the venous blood gas (VBG), despite of its inaccuracy, is increasingly used instead of the ABG. Currently, there is no standardized accepted calculation to estimate arterial values of pressure of carbon dioxide (pCO 2 ), pH, and bicarbonate (HCO 3- ) from the venous gas panel. Here, we report the relationship between ABG and VBG parameters (pH, pCO 2 , and HCO 3- ) based on 5419 samples collected concurrently. METHODS: For pCO 2 , linear regression analysis of venous and arterial values showed a strong correlation ( r =0.93, P < 0.0001) and a formula derived from this relationship calculated the estimated arterial pCO 2 , compared with the actual measured arterial pCO 2 , with a median difference of 0.02 mm Hg and 95% limits of agreement ranging from -5.6 to +5.5 mm Hg. For pH, linear regression analysis of venous and arterial values also showed a strong correlation ( r =0.94, P < 0.0001) and a formula derived from this relationship calculated the estimated arterial pH, compared with the actual measured arterial pH, with a median difference of -0.002 pH units and 95% limits of agreement ranging from -0.05 to +0.04 pH units. To calculate an estimated arterial HCO 3- , the arterial pCO 2 and pH estimated from the respective formulas were entered into the Henderson-Hasselbalch equation. RESULTS: This resulted in an estimated arterial HCO 3- with a median difference of -0.08 mEq/L and 95% limits of agreement ranging from -2.29 to +1.96 mEq/L as compared with the reported arterial HCO 3- . The data show that the VBG can be transformed with reasonable confidence intervals to arrive at a corresponding ABG. CONCLUSIONS: The results of this transformation meet acceptable clinical expectations and are supported by the large database used to generate the formulas that estimate independently the three ABG components (pCO 2 , pH, and HCO 3- ).