Comparing Noninvasive Pulse CO-Oximeter vs Blood Gas Analysis in Emergency Department Patients with Carbon Monoxide Poisoning

Introduction: In order to diagnose carbon monoxide (CO) poisoning, clinical suspicion, a reliable history, and the detection of high level of carboxyhemoglobin (COHb) in blood gas are required. The purpose of this study is to compare noninvasive pulse CO-oximetry versus blood gas analysis (BGA) in emergency department (ED) patients with CO poisoning. Methods: From the patients, who were 18 years of age or over, presenting with suspicion of CO poisoning and who were eligible for inclusion in the study, arterial/venous blood gas was collected and SpCO was measured by pulse CO-oximetry and recorded by the time of ED visit. Sensitivity and specificity percentages, positive and negative predictive values were determined with ROC analysis. Bland-Altman analysis was used to assess the agreement between two measurement methods. Results: The study was carried out prospectively on 213 patients in total, 133 (62%) of whom were female with a mean age of 38 ±15. The limits of agreement were -6.5 to 9.9% COHb (bias 1.7%, precision 4.1%). Cut off value in CO measurement was 23, sensitivity was 97.2%, specificity was 80.4%, NPV was 96.6%, and PPV was 83%. Concordance coefficient value (0.868) between BGA, and pulse CO-oximetry was found out to be significantly high. Conclusion: When pulse CO-oximetry is compared with BGA, it appears to diagnose CO poisoning rapidly and accurately and therefore, it may be used both in pre-hospital period and EDs as a screening test for CO poisoning