RESUMO
A comparative evaluation of two methods used for carboxyhemoglobin (COHb) determination in postmortem whole blood was performed: carbon monoxide (CO)-oximetry measuring at 128 wavelengths and headspace gas chromatography with flame ionization detection (HS-GC--FID) where CO was determined after catalytic reduction of CO to CH4 and Fe was determined by atom absorption spectrophotometry (AAS, 248.3 nm). An aliquot of 100 µL whole blood was loaded into the CO-oximetry module. In the HS-GC--FID analysis, to 1.0 mL of whole blood, 3.0 mL of saponin solution was added, mixed and centrifuged. To 20 mL HS vials, 400 µL of the supernatant was added and the vials were immediately sealed. One milliliter of potassium hexacyanoferrat (III) solution was added through the HS septum and mixed. The samples were incubated at 70°C for 5 min. CO was separated using He as carrier gas and a CP-Molsieve 5 Å PLOT capillary column. Fe was determined using 400 µL of the saponin supernatant diluted to 10 mL by water. During a period of â¼3 years, 124 postmortem whole blood samples were analyzed. Bland-Altman method comparison showed satisfactory agreement and no significant bias between the methods for the whole saturation range (5 to 85% COHb). Five samples, all with %COHb >40, showed deviations of more than 10% COHb in absolute terms. One sample, in the lower COHb range <10%, was false negative on the CO-oximetry method. The between-assay accuracy, reported as bias, at 60% COHb was -0.8% and -9.0%, and precision, reported as relative standard deviation, was 1.6% and 7.7%, for the CO-oximetry and HS-GC--FID-AAS methods, respectively. Both methods obtained satisfactory results in proficiency testing rounds, with z-scores <±2 (n = 11). This study showed that the CO-oximetry method based on the 128-wavelength principle and the HS-GC--FID-AAS method are comparable and satisfactory for %COHb determination in postmortem whole blood.
