Carbon monoxide poisonings: exploring new approaches for quantification and evaluating measurement errors from an analytical and epidemiological point of view

Abstract

Thousands of deaths and hospital admissions worldwide are reported every year due to carbon monoxide poisoning. However, current biomarkers of exposure and quantification devices employed for detection and diagnosis lack sufficient specificity and accuracy, leading to frequent errors. This results in an underestimation of the true disease burden attributable to CO exposure, making it a topic of concern for the clinical and public health communities. The aim of this doctoral research was to investigate some of the factors leading to underestimation of the health risks associated with CO exposure and evaluate their margins of error, in order to propose improvements from an analytical and epidemiologic perspective. The quantification of Total Blood CO (TBCO) via Airtight gas Syringe-Gas ChromtographyMass Spectrometry (AGS-GC-MS) was identified as more accurate, specific and less errorprone than current biomarker carboxyhaemoglobin (COHb) and proposed as alternative biomarker for CO exposure. The method was validated for clinical and forensic applications and tested on real cases in both fields. Results for the comparison between the two biomarkers in the different settings (e.g. different storage conditions in clinical and postmortem cases) confirmed TBCO as more appropriate biomarker for CO exposure and highlighted the limitations of COHb. To determine the potential impact on CO exposure assessments of TBCO under controlled conditions and, consequently, its effect on the relative risk, the sources of measurement error in CO exposure assessment studies were determined and the improvement on measurement error was calculated on one exemplary study. The resulting relative risks were increased moderately, thus getting closer to the true risk of CO exposure in the population. This affirms the importance of even small improvements in one part of the error sources being able to achieve tangible changes with important public health implications. Despite COHb being a cheaper and more established biomarker, I currently recommend implementing TBCO for challenging cases and encourage further research in this area.