Carbon monoxide (CO) is one of the leading causes of domestic deaths. Exposure occurs in- and outdoors, in occupational or accidental (e.g. fires) settings, or by smoking. Current biomarker COHb has poor correlation with symptoms of CO poisoning, which are unspecific. Thus, this study aims to evaluate alternative methods for estimating CO, in order to reduce the error in diagnosis of CO poisoning.Methods
We explore a new technique based on gas-chromatography coupled to mass-spectrometry (GC-MS) that measures the total CO amount in blood (TBCO). CO is released from a blood sample (100 µL) by mixing with sulphuric acid in an airtight vial and heating. It is then analysed by GC-MS. This method is then compared to an optical method (CO-oximetry). An assessment of the measurement errors deriving from an analytical perspective is performed for both approaches through monitoring of analytical factors, such as storage conditions (temperature, time, volume, freeze- and thawing cycles, tube reopening, initial concentration levels) and instrumental errors, with analysis of ~2500 bovine blood samples. Results are compared and assessed using descriptive statistics and regression models.Results
A novel approach for CO quantification in small amounts of blood was developed and validated for clinical and postmortem range. The total amount of CO present in blood at analysis time is released and analysed through airtight gas syringe (AGS)-GC-MS.Preliminary results of the monitoring study show important statistical influences by all parameters analysed but ‘tube re-opening’ (p-values>0.05). No linear correlation is found between COHb and TBCO, with resulting backcalculated COHb from TBCO concentrations generally higher than COHb measured optically. The approach was applied to real poisoning cases, with resulting COHb between 45%–75%, as opposed to COHb backcalculated from TBCO, which ranges between 83%–275%, suggesting the presence of a part of CO dissolved in blood and not bound to HbConclusion
CO levels based on optically measured COHb may lead to underestimation of CO exposure. CO not linked to Hb might be of pathophysiological relevance. The use of TBCO as more reliable biomarker of CO exposure in the clinical field requires further testing. Furthermore, the new approach presents lower cost and invasiveness. Storage guidelines can be derived for use in clinical and forensic domains.