Agreement between arterial and venous pH and pCO2 in patients undergoing non-invasive ventilation in the emergency department

Background Blood gas analysis is an important part of the assessment of ventilatory function in patients with respiratory distress. Traditionally, analysis of arterial blood has been used, but recently, there has been a move towards venous blood gas analysis for selected conditions. Arteriovenous agreement for pH and pCO2 in adult patients undergoing non-invasive ventilation (NIV) has not been explored to date. The aim of this study was to address this question. Methods This was a prospective study of adult patients undergoing NIV for acute respiratory compromise in an ED. When arterial blood gas analysis was required for clinical purposes, staff also drew a venous sample as close to simultaneously as possible. Data collected included demographics, clinical diagnosis and results of blood gas analyses. Primary outcome of interest was arteriovenous agreement for pH and pCO2. Bland-Altman bias plot analysis was used. Results Eighty-nine sample pairs in 53 patients were studied. Median age was 74, and 64% of patients were male. The principal diagnoses were chronic obstructive pulmonary disease (43%) and acute pulmonary oedema (40%). Mean difference for pH (a-v) was 0.04 pH units (95% limits of agreement -0.02, 0.11). Mean difference for pCO2 (a-v) was -8.02mmHg (95% limits of agreement -22.63, 6.58) Conclusion For adult patients undergoing NIV in an ED, arteriovenous agreement for pH was close with narrow limits of agreement making venous pH clinically interchangeable with arterial pH. Agreement for pCO2 was poor with unacceptably wide limits of agreement.