ABSTRACT
BACKGROUND: If acid-base status and electrolytes on blood gases during cardiopulmonary resuscitation (CPR) differ between the arrest causes, this difference may aid in differentiating the arrest cause. We sought to assess the ability of blood gases during CPR to predict the arrest cause between primary cardiac arrest and asphyxial arrest. METHODS: A retrospective study was conducted on adult out-of-hospital cardiac arrest patients for whom blood gas analysis was performed during CPR on emergency department arrival. Patients were divided into two groups according to the arrest cause: a primary cardiac arrest group and an asphyxial arrest group. Acid-base status and electrolytes during CPR were compared between the two groups. RESULTS: Presumed arterial samples showed higher potassium in the asphyxial arrest group (p < 0.001). On the other hand, presumed venous samples showed higher potassium (p = 0.001) and PCO2 (p < 0.001) and lower pH (p = 0.008) and oxygen saturation (p = 0.01) in the asphyxial arrest group. Multiple logistic regression analyses revealed that arterial potassium (OR 5.207, 95% CI 1.430-18.964, p = 0.012) and venous PCO2 (OR 1.049, 95% CI 1.021-1.078, p < 0.001) were independent predictors of asphyxial arrest. Receiver operating characteristic curve analyses indicated an optimal cut-off value for arterial potassium of 6.1 mEq/L (sensitivity 100% and specificity 86.4%) and for venous PCO2 of 70.9 mmHg (sensitivity 84.6% and specificity 65.9%). CONCLUSIONS: The present study indicates that blood gases during CPR can be used to predict the arrest cause. These findings should be confirmed through further studies.