Assessment of repeatability of eNose (SpiroNose) measurements in healthy and asthmatic subjects

Abstract

Rationale: Breathomics is an emerging metabolomics field with potential applications in respiratory medicine. High between-days repeatability is a prerequisite to bring electronic nose (eNose) technology to the point-of-care. Aim: To assess the between-days repeatability of exhaled breath analysis by SpiroNose in asthmatic and healthy subjects. Methods: This was a prospective observational 3 months follow-up study (33 visits) in 12 asthmatic and 12 healthy subjects. Exhaled breathprints were collected in duplicate by eNose (SpiroNose) [De Vries ERJ 2018]. Other clinical data (FEV1 and FeNO) were used for comparison. Data-analysis involved signal processing, ambient correction and principal component analysis. Repeatability was expressed as the intraclass correlation coefficient (ICC), which represents the agreement between multiple evaluations in the same set of subjects. Results: All principal components (PC 1-4), for both asthmatic and healthy subjects, showed ICC values ranging between 0.70-0.93, which can be interpreted as having ‘good to excellent repeatability’. PC1 (explained 51.4% of the total variance within the dataset) showed the highest repeatability in asthmatic (ICC=0.93) and healthy (ICC=0.90) subjects. The lowest repeatability (healthy: ICC=0.70, asthma: ICC=0.72) was obtained in PC4 (explained variance 7.8%). FEV1 and FeNO showed ICC values of 0.73 and 0.43. Conclusion: The SpiroNose showed high between-days repeatability values in healthy and asthmatic subjects, which was better compared to other routine diagnostic tests (spirometry and FeNO). These results suggest that exhaled breath analysis by SpiroNose allows for reliable patient monitoring at the point-of-care.