DOB, delta over baseline; Hp+, Helicobacter pylori-positive; Hp–, H. pylori-negative.
Reply to Jonderko et al. The reproducibility of 13CO2 measurement. 19: 142–4
Article first published online: 30 JAN 2004
Alimentary Pharmacology & Therapeutics
Volume 19, Issue 3, pages 376–378, February 2004
How to Cite
Perri, F. and (on behalf of ‘Helicobacter pylori Sige Study Group on 13co2 Measurement Standardization’) (2004), Reply to Jonderko et al. The reproducibility of 13CO2 measurement. 19: 142–4. Alimentary Pharmacology & Therapeutics, 19: 376–378. doi: 10.1111/j.0269-2813.2004.01815.x
- Issue published online: 30 JAN 2004
- Article first published online: 30 JAN 2004
Sirs, 13C breath analysis is generally performed by either isotope ratio mass spectrometry or non-dispersive infrared spectroscopy. Both methodologies are accurate provided that instruments are properly set up and quality control procedures are periodically performed in each laboratory. Our multi-centre study was performed in 22 isotope ratio mass spectrometry centres and showed unexpectedly high inter- and intra-laboratory variabilities in 13C measurements.1 The low analytical accuracy significantly affected the results of the octanoic acid breath test and the urea breath tests. As shown in Table 1, the gastric half-emptying time ranged from 88 to 149 min for the first sample measurement and from 89 to 285 min for the second. Assuming a cut-off value of 120 min for the gastric half-emptying time,2 three of the 44 (6.8%) octanoic acid breath test results (indicated in bold type) were falsely consistent with delayed gastric emptying. With regard to the urea breath tests, nine of the 132 (6.8%) results (indicated in bold type) were either falsely negative or meaningless. From a clinical point of view, this finding is relevant as inappropriate therapeutic decisions could be adopted on the basis of inaccurate results.
|Centre||First set of measurements||Second set of measurements|
|OBT t1/2 (min)||1st UBT Hp– (DOB30)||2nd UBT Hp+ (DOB30)||3rd UBT Hp+ (DOB30)||OBT t1/2 (min)||1st UBT Hp– (DOB30)||2nd UBT Hp+ (DOB30)||3rd UBT Hp+ (DOB30)|
|14||113||− 5.19||46.40||5.12||285||− 3.25||47.03||3.39|
Regrettably, for technical reasons (all instruments were temporarily out of order at the time of study), three centres equipped with non-dispersive infrared spectroscopy devices did not provide any results and were excluded from the final analysis. However, there is no reason to suppose that ‘non-responders wanted to avoid comparison of their results with others’, as several papers showed no difference in the analytical precision between isotope ratio mass spectrometry- and non-dispersive infrared spectroscopy-based devices.3–5 This was partially confirmed by the results presented by Jonderko et al.6 who showed a good repeatability (‘within-laboratory’ accuracy) of 13C breath measurements in their laboratory equipped with an infrared spectrometer. Unfortunately, no information on the reproducibility of 13C measurements can be obtained from their data. As clearly stated in our paper, the reproducibility is the sum of the repeatability plus the variability observed between different laboratories analysing the same submitted samples. Therefore, we would like to invite the Polish researchers to design a multi-centre study (using samples appropriate for non-dispersive infrared spectroscopy devices) to evaluate the ‘between-laboratory’ accuracy and reproducibility of 13C measurements obtained using non-dispersive infrared spectroscopy.
We would also like to invite Jonderko et al. to use the standardized terminology. In their letter, delta over baseline (DOB) seems to be erroneously used instead of δ13CPDB. δ13CPDB (or ‘δ value’) is calculated using the formula, [(Rsample − RPDB)/RPDB] × 1000, with R = 13C/12C, which gives the relative 13C enrichment of one sample in comparison with the international reference PDB (a limestone of North Carolina with which 13C/12C ratios are compared). The absolute 13C/12C ratio of this standard (RPDB) is 0.0112372. By means of another calculation, it is possible to convert δ13CPDB to %13C: %13C = (δ′ + 1) × RPDB × 100/[1 + RPDB × (1 + δ′)], where δ′ = δ13CPDB/1000. Conversely, DOB (also known as ‘excess δ13CO2’) is a measure of the 13C enrichment of breath after the administration of 13C substrates. DOB is calculated as an algebraic difference between two δ13CPDB values. For instance, in the urea breath test, DOB is the difference between δ13CPDB at 30 min and δ13CPDB at time zero. Using standardized terminology, Jonderko et al. should express their results in terms of a difference in δ13CPDB values (or Δδ value) for each paired breath sample and summarize them as mean Δδ values (with 95% confidence intervals) for all paired samples analysed on the same day or on different days.
In conclusion, we believe that the results obtained by the Polish researchers are very encouraging, but a larger multi-centre study adopting standardized terminology (repeatability, reproducibility, δ values, DOB, etc.) is needed.