• Lundblad Lennart K A LKA Meakins-Christie Laboratories, McGill University, Montréal, QC, Canada. lennart.lundblad@mcgill.ca. , Ruzica Miletic, Eeva Piitulainen, and Per Wollmer.
• Meakins-Christie Laboratories, McGill University, Montréal, QC, Canada. lennart.lundblad@mcgill.ca.
• Sci Rep. 2019 Aug 12; 9 (1): 11618.

AbstractImpedance, or oscillometry, measurements of the respiratory system can generate information about the function of the respiratory system not possible with traditional spirometry. There are currently several instruments on the market using different perturbations. We have compared a new respiratory oscillometry instrument, the tremoflo, with Impulse Oscillometry (IOS). Patients with a physician's diagnosis of chronic obstructive lung disease (COPD) and healthy subjects were recruited. They underwent assessment of respiratory function with oscillometry using the IOS and tremoflo devices and the resulting impedance data from the two methods were compared. The two devices were also tested against a reference respiratory phantom with variable resistances. Whereas both devices detected impairments in the patients' lung function commensurate with small airways pathology, the tremoflo appeared to be more sensitive than the IOS. We found systematic differences between the two instruments especially for reactance measurements where the area over the reactance curve (AX) was significantly lower with the IOS compared with the tremoflo (p < 0.001). Moreover, the agreement between the two devices was reduced with increasing severity of the disease as determined with a Bland-Altman test. Testing both instruments against a respiratory phantom unit confirmed that the resistance measured by the tremoflo compares closely with the known resistance of test loads, whereas the IOS' resistance correlated with a test load of 0.19, kPa.s.L-1 at higher loads it deviated significantly from the known resistance (p < 0.0028). We conclude that the absolute values measured with the two devices may not be directly comparable and suggest that differences in the calibration procedures might account for the differences.

13 keywords...

hide…