Introduction & Objectives In patients with COPD, an arterial blood gas (ABG) is considered the gold standard method of directly measuring serum pH, pO2, pO2 and calculating bicarbonate (HCO3) levels. These values allow the assessment of a patient’s acid base status and adequacy of ventilation and oxygenation. The aim of this study was to identify if a venous blood gas can accurately reflect an arterial blood gas in determining the ventilatory function of patients presenting with acute exacerbations of COPD.
Methods This prospective observational study was conducted in a Scottish urban ED. All consecutive patients presenting with acute exacerbations of COPD were eligible. An ABG was taken from each patient, as deemed necessary by the treating physician, along with a venous gas at the time of venepuncture. Pearson's correlation coefficient and Bland Altman analysis methods were used to identify correlation and agreement between the arterial and venous data sets.
Results 68 paired samples were obtained over a two month period. Correlation was strong (r = 0.953) between arterial and venous pH. Bland Altman analysis showed an average difference (bias) of 0.017, with 95% limits of agreement (LOA’s) of -0.052 to 0.087. Arterial and venous values for HCO3 were also strongly correlated (r = 0.914). The agreement was -0.834mmol/l with 95% LOA’s of -4.82 to 3.15 mmol/l. Despite arterial and venous pCO2 strongly correlating (r = 0.973), the agreement bias was 4.66mmHg with 95% LOA’s of -4.94 to 14.26 mmHg. Arterial hypercarbia, defined as pCO2 > 45 mmHg, was present in 31 patients (46%). All cases of arterial hypercarbia were detected on venous blood gas sampling using a pCO2 screening cut-off of 45mmHg. This was found to be 100% sensitive (95% CI 89–100%) and 86% specific (95% CI 71–95%).
Conclusion There is a strong correlation between arterial and venous pH, pCO2 and HCO3. Agreement between pH and HCO3 is acceptable enough to substitute a venous blood gas value for an arterial blood gas value. However, venous pCO2 does not agree with arterial pCO2, therefore cannot be substituted. A venous pCO2 screening cut-off of 45 mmHg has 100% sensitivity in detecting arterial hypercarbia. Had a venous blood gas been performed initially, 47% of ABG’s could be avoided in these patients.