All anonymised data included in the analysis.

IntroductionAlthough peripheral venous blood gas (pVBG) analysis is used in the Emergency Department (ED), its effect on clinical decision making is unknown. We assessed whether pVBG analysis combined with pulse oximetry could replace arterial blood gas (ABG) analysis to determine treatment and disposition of ED patients with respiratory complaints. In addition, we assessed agreement between venous and arterial values and pulse oximetry (SpO2).MethodWe performed a 12-week prospective observational study in ED patients with respiratory complaints. ABG and pVBG samples were drawn as simultaneously as possible, with a maximum of five minutes in between. Physicians initially determined treatment and disposition using pVBG results, after which they were shown the ABG results. Subsequent alterations in treatment and disposition were registered. We calculated pVBG and ABG mean differences (MDs) using Bland-Altman analysis and SaO2 and SpO2 MD and correlation using Passing-Bablok regression analysis and Bland-Altman analysis.ResultsIn 56/154 (36.4%) patients, the ABG results changed the preliminary treatment and disposition. Most (57.5%) changes consisted of a change in supplemental oxygen therapy. The MDs (95% CIs) between pVBG and ABG results were: pH −0.04 (−0.05 to −0.04) pH units, bicarbonate 1.57 (1.20 to 1.93) mmol/l, pCO2 0.85 (0.70 to 0.99) kPa and lactate 0.34 (0.28 to 0.40) mmol/l. We found a good correlation between the SaO2 and SpO2.ConclusionIn over one third of patients with respiratory complaints in the ED, ABG results changed treatment and/or disposition based on pVBG results. Most changes could be considered as minor. The arterial pO2 was most frequently mentioned as the reason for the changes.