Introduction Faecal Volatile Organic Compounds (VOCs) analysis is an emerging diagnostic tool for gastrointestinal conditions because of its sampling ease and non-invasive characteristics. Solid phase micro-extraction (SPME) is often used with gas chromatography–mass spectrometry for the analysis of VOCs; however no procedure has been standardised for an application in faecal analysis with the potential for on–site utilisation. Several aspects of the sampling preparation applied to neonatal faecal samples were studied to improve the robustness of the analytical process for paediatric studies.

Methods Thirty-three faecal neonate samples of weight 50–700 mg were analysed. The results produced by SPME coatings CAR/PDMS and DVB/CAR/PDMS were compared (n = 5/variable), as were vial volumes of 2 and 10 ml (n = 4/variable; N=3) and the addition of 0.5 and 1 ml (n = 3/variable) of a saturated NaCl solution prior to GC-MS analysis. In addition, the influence of leaving the samples for 14 h at 1°C instead of -20°C was studied (n = 3/condition). Finally, the reproducibility of the method was tested by looking at the internal variation of 10 sets of triplicate; furthermore, 3 sets were reanalysed 4 times in order to characterise the repeatability across GC-MS runs. Independent samples t-test, one–way ANOVA and Tukey’s HSD test were performed to test differences between data classes. Final p-values were adjusted by Bonferroni and p-values < 0.05 were considered significant.

Results Twenty (±2) VOCs were identified using samples of 100 mg, while 28 (±1) and 25 (±2) VOCs were identified in samples of 450 respectively 700 mg. There were no significant differences in VOCs intensities between samples of 450 and 700 mg as between 50 and 100 mg. However, all VOCs intensities were significantly higher in samples of 450 and 700 mg when compared to 100 mg. No differences were observed between the two SPME coatings, and the addition of salt did not improve results quantitatively or qualitatively. Keeping samples at 1°C instead of -20°C and/or varying the volume of the vial did not influence the results significantly. Finally, for 7 sets of triplicates out of 10, more than 90% of the ion intensities varied less than 30% but multiple runs of GC-MS resulted in significant changes in intensity of 40% or more of the VOCs identified.

Conclusion Several parameters were studied to optimise a method for paediatric faecal VOCs analysis and a robust method has now been developed and detailed. Samples of 50–100mg may be studied. This weight gives reproducible results, but samples should not be reanalysed as headspace VOCs are altered by the procedure. Other changes to sample processing had little impact on the results. A chilled auto sampling rack may be safely used.

Disclosure of Interest None Declared.