To measure a patient’s metabolic response to an administered 13C-enriched substrate, isotope selective nondispersive infrared spectrometry is used. Isotope abundance levels are relative, i. e., reported as differences between a tested sample and a reference sample. The reference 13C abundance is not known exactly. This uncertainty, uncertainty in CO2 production, and the inaccuracy of the measuring instrument contribute to the uncertainty in the results of breath tests. In this study, the particular impacts of uncertainty are estimated and expressed in a mathematical way by an uncomplicated formula illustrated by an example dealing with real-life data. It is shown that the uncertainty in the reference 13C abundance does not have severe consequences, so that the method can compete with the spectrometry methods that are able to deliver an absolute value of the 13C abundance. The inaccuracy of the measuring instrument is also manageable, though its influence is greater than in the previous case. The analysis reveals that the uncertainty in CO2 production deserves great attention because it is difficult to estimate and its influence is rather strong. The problem of determination of a proper cut-off level is outlined.

        Key words: isotope selective nondispersive infrared spectrometry, breath tests, uncertainty, accuracy.