Measuring nitrous oxide after death requires more than finding a mass-spectrometric peak shared with carbon dioxide. Li and colleagues developed a blood method that removed interfering carbon dioxide and then applied it to two suspected fatal poisonings. The measured heart-blood concentrations and positive tissue findings supported the investigations, but did not establish a universal lethal range or quantitative validation for every specimen.
Why the identification problem mattered
Nitrous oxide and carbon dioxide share the same nominal molecular mass. In this study, the selected chromatographic column did not completely separate them, even at relatively low carbon dioxide concentrations. A signal at mass-to-charge ratio 44 therefore could not, by itself, resolve the analytical problem.
The investigators used sodium hydroxide treatment to remove carbon dioxide and headspace gas chromatography–mass spectrometry, or HS-GC–MS, to analyze the remaining gas. Identification compared retention time with a reference within 1% and evaluated ions at m/z 44, 30 and 28 and their relative abundances. Quantification used m/z 30 for nitrous oxide and n-pentane as the internal standard.
These were defined identification criteria and a separate quantitative calculation. A reference match, an instrument response and an accurately measured concentration are related but distinct parts of the evidence.
What blood performance was demonstrated?
The study evaluated eight blank human blood samples and eight blood samples with internal standard. It reported a blood calibration range of 0.02–0.5 mL nitrous oxide per mL of blood, with R² above 0.995. The adopted detection limit was 0.005 mL/mL and the quantitation limit was 0.02 mL/mL.
The detection limit was practical rather than the lowest visible signal: lower levels created unacceptable operational variation during manual gas dilution. The quantitation limit also reflected the practical precision of the gas syringes. Neither limit is a biological threshold for impairment or poisoning.
Accuracy and precision were tested at 0.06, 0.2 and 0.4 mL/mL, using six replicates per level over four nonconsecutive days. Reported bias ranged from −7.6% to +4.9%; intraday relative standard deviations ranged from 4.3% to 10.4%, and interday values from 7.0% to 8.8%.
Collection was part of the analytical procedure
During the two autopsies, blood, tissues and urine were placed directly into prepared headspace bottles and immediately sealed. Lung and stomach gas was collected using a gas-tight syringe and transferred to gas collection bags. Specimens were returned promptly to the laboratory and tested immediately.
The paper did not provide a uniform death-to-autopsy interval, a controlled delayed-storage experiment, or a paired antemortem comparison. Its successful application therefore does not validate arbitrary transport delays, repeatedly opened containers or reconstruction of the concentration during life.
What the two cases contributed
Heart blood contained 0.034 mL/mL in the first case and 0.14 mL/mL in the second. These were cardiac specimens, not femoral blood. The investigators also reported nitrous oxide in other biological specimens, including tissues and collected gases.
The method’s quantitative validation was expressly limited to human blood. Positive findings in lung, kidney, brain or other specimens did not become validated tissue concentrations merely because the same instrument detected them. The authors identified extension to other matrices as future work.
Two fatal cases can provide reference observations, but cannot establish the concentration separating survival from death. Detection does not independently quantify oxygen deprivation, establish the administered dose or prove an earlier period of impairment. Cause-of-death assessment still requires the complete investigation.
The method had acknowledged limitations
The authors noted possible loss of nitrous oxide through dissolution in the sodium hydroxide solution. They also acknowledged lower sensitivity than a previously published method reporting in mL/L. Units must be aligned: 0.005 mL/mL equals 5 mL/L, not 0.005 mL/L.
When reviewing a result, retain the blood site, sealing and transfer history, calibration units, identity criteria and validated matrix. The same need to examine analyte-specific scope appears in validation of a targeted 82-drug blood assay. Instrument capability alone does not establish every downstream interpretation.
Primary source
Li Z, Li Z, Qiang H, Xie W, Su M, Xiang P, Shi Y. Quantitative determination of nitrous oxide in human blood by HS-GC–MS: forensic application of two fatal poisoning cases. Forensic Science International. 2024;360:112067. doi:10.1016/j.forsciint.2024.112067.
Leave a Reply