Electrochemical measurement of carbon monoxide in breath: Interference by hydrogen

The purpose of this study was to determine the concentration of carbon monoxide (CO) in blood (COHb) and breath to demonstrate that breath hydrogen (H₂) can be a significant interferant. For this purpose, we measured blood COHb with CO-oximetry and breath CO with an electrochemical analyzer. In addition, the samples were analyzed by gas chromatography (GC). The concentration of CO in breath, collected with a Priestley tube after a 20 s breath hold, from healthy, nonsmoking adult males (n = 20) and females (n = 10) had a mean ± SD (range) of 2.6 ± 0.4 ppm (2.0–3.9), respectively, when measured by GC. However, these same samples when measured with an electrochemical (EC) analyzer showed elevated CO values of 4.7 ± 2.9 ppm (2.6–17.6). The concentration of H₂, a prominent trace gas in breath known to interfere with EC analyzers, correlated strongly with the observed EC analyzer response [EC (ppm CO) = 0.336 H₂ (ppm) + 1.93, r² = 0.98]. The EC analyzer was linear for H₂ concentrations up to 40 ppm, with a sensitivity of 0.035 V ppm⁻¹. The analyzer sensitivity to CO was 0.10 V ppm ⁻¹. Blood from this population showed COHb concentrations of 0.56 ± 0.11% (0.40–0.97), as measured by GC, but elevated values were found when measured by CO-oximeter (Ciba Corning Diagnostics Corp., Models 2500 and 270), 1.3 ± 0.2% (1.1–1.6) and 1.0 ± 0.3% (0.1–1.6), respectively. When breath CO was compared to blood COHb, only measurements by GC significantly correlated [COHb% = 0.241 CO(ppm) — 0.076, r² = 0.78]. We conclude that, relative to quantitative analysis by GC, (1) EC analyzers are susceptible to H₂ interference that cause falsely elevated CO measurements, and (2) CO-oximeters overestimate COHb concentrations in the range typical for healthy nonsmokers.