Venous vs arterial blood gases in the assessment of patients presenting with an exacerbation of chronic obstructive pulmonary disease

Abstract

ObjectiveThe purpose of this study was to investigate the clinical correlation between arterial and venous blood gas (VBG) values in patients presenting to the emergency department (ED) with acute exacerbation of chronic obstructive pulmonary disease. MethodsA prospective study of patients with chronic obstructive pulmonary disease presenting to the ED with acute ventilatory compromise was done. Patients were included if their attending physician considered arterial blood gas sampling important in their initial assessment. Data from arterial and venous samples were compared using Spearman correlation and bias plot (Bland-Altman) methods. ResultsNinety-four patients were enrolled in the study. Eighty-nine patients had complete data sets for analysis. Arterial hypercarbia was present in 30 patients (33.7%; range, 51-140.19 mm Hg). All cases of arterial hypercarbia were detected using VBG sampling when a screening cutoff of 45 mm Hg was applied (sensitivity, 100%; 95% confidence interval, 88.7%-100% and specificity, 34%; 95% confidence interval, 23.1%-46.6%). Bias plot revealed moderate agreement between arterial and venous Pco2 with an average difference of 8.6 mm Hg and 95% limits of agreement of −7.84 to 25.05 mm Hg. For pH, mean difference between each group was 0.039 (range, −0.12 to 0.03). Linear regression analysis for pH demonstrated very close equivalence with a regression coefficient of 0.955, and Spearman correlation showed significant correlation of 0.826 (P = .001). ConclusionVenous pH and HCO3 values show excellent correlation with arterial values. Using a previously validated screening cutoff of 45 mm Hg, venous CO2 has 100% sensitivity in detecting arterial hypercarbia. There is insufficient agreement between venous and arterial CO2 for VBG to replace arterial blood gas in determining the degree of hypercarbia.