Mathematical Arterialization of Capillary Blood for Blood Gas Analysis in Critically Ill Patients

Abstract

Background: In clinical practice, capillary blood taken from hyperemized earlobes (CBG_E) or fingertips (CBG_F) is frequently used as substitute for arterial blood (ABG) for blood gas analysis. While there is a close agreement between ABG and CBG_E/CBG_F regarding pH and pCO₂, pO₂ is often underestimated by CBG. Recently, a software tool (v-TAC^®; Roche Diagnostics, Risch-Rotkreuz, Switzerland) has been developed to calculate ABG values based on a peripheral venous blood gas analysis supplemented with peripheral oxygen saturation. Objective: Here we investigate whether v-TAC can also be used to calculate ABG values from capillary blood samples. Methods: Patients (n = 85) with an indwelling arterial line were included in the study. A reference ABG sample (ABG₁) was obtained, followed by CBG_E, CBG_F, and finally a second ABG (ABG₂). Results of CBG_E/CBG_F before and after mathematical arterialization by v-TAC (aCBG_E/aCBG_F) were compared to ABG₁. Results: After mathematical arterialization by v-TAC, the mean bias in pO₂ between ABG₁ and CBG_E went down from 5.24 mm Hg (95% limit of agreement [95% LoA]: −14.19 to 24.67) to 0.18 mm Hg (95% LoA: −11.84 to 12.20) and was in a similar range as the mean bias between ABG₁ and ABG₂ (0.39 mm Hg [95% LoA: −13.46 to 14.24]). Differences in pH and pCO₂ between arterial and capillary samples were small before and after mathematical arterialization. Very similar results were obtained when using fingertip instead of earlobe capillary blood. Conclusion: In summary, v-TAC can be used for mathematical arterialization of capillary blood samples for blood gas analysis resulting in increased diagnostic accuracy for pO₂.