Carbon monoxide (CO) is one of the most important toxic gases in the atmosphere. Its high affinity for hemoglobin made carboxyhemoglobin (COHb) the most appropriate biomarker for CO poisonings. COHb is measured using spectrophotometric (UV-spectrophotometry, CO-oximetry) or gas chromatographic (GC) methods hyphenated with flame ionization or mass spectrometry detectors. However, inconsistencies in many cases were reported between measured values and reported symptoms, raising doubts over the suitability of COHb as biomarker together with the accuracy and reliability of its measurement methods. Therefore, we aimed to review the accuracy of current methods to measure CO and to determine their sources of error and their effects on the interpretation process.
MethodsA detailed search of PubMed was performed in November 2018 using relevant keywords. After exclusion criteria, 46 articles out of 191 initial hits were carefully reviewed.
ResultsWhile optical methods are highly affected by changes of blood quality due to degradation of the samples during storage, GC methods are less affected. However, measurement of COHb does not quantify free CO, which is mainly responsible for toxicity mechanisms other than hypoxia such as inhibition of hemoproteins, thus underestimating the true CO burden. Therefore, measurement of COHb is insufficiently accurate for CO poisoning diagnosis.
ConclusionsAn alternative biomarker is required, such as determining the total amount of carbon monoxide in blood. Even though further research is required, we recommend toxicologists to consider all sources of error that can alter COHb concentrations and, in more challenging cases, they should use GC-MS methods to confirm the results obtained by spectrophotometry.