Assessing PaCO2 in acute respiratory disease: accuracy of a transcutaneous carbon dioxide device

Background: Pulse oximetry non-invasively assesses the arterial oxygen saturation of patients with acute respiratory disease; however, measurement of the arterial partial pressure of carbon dioxide (PaCO2) requires an arterial blood gas. The transcutaneous partial pressure of carbon dioxide (PtCO2) has been used in other settings with variable accuracy. We investigated the accuracy of a PtCO2 device in the assessment of PaCO2 in patients with asthma and suspected pneumonia attending the emergency department. Methods: Patients with severe asthma (FEV1 <50% predicted) or suspected pneumonia (fever, cough and respiratory rate >18/min) were enrolled. Subjects were excluded if they had a history of chronic obstructive pulmonary disease or other conditions associated with respiratory failure. Arterial blood gases were taken at the discretion of the investigator according to clinical need, and paired with a simultaneous reading from the PtCO2 probe. Results: Twenty-five patients were studied with one set of data excluded because of poor PtCO2 signal quality. The remaining 24 paired samples comprised 12 asthma and 12 pneumonia patients. The range of PaCO2 was 19-64 mmHg with a median of 36.5 mmHg. Bland-Altman analysis showed a mean (SD) PaCO2 - PtCO2 difference of -0.13 (1.9) mmHg with limits of agreement of plus or minus 3.8 mmHg (-3.9 to +3.7). Conclusion: A PtCO2 device was accurate in the assessment of PaCO2 in patients with acute severe asthma and suspected pneumonia when compared with an arterial blood gas. These bedside monitors have the potential to improve patient care by non-invasively monitoring patients with acute respiratory disease at risk of hypercapnia.