Discrepancy between in vivo and in vitro comparisons of forced oscillation devices

June 3, 2020 / in Scientific articles / by Eve-Gabrielle Bissonnette

The forced oscillation technique (FOT) is an emerging clinical lung function test, with commercial devices becoming increasingly available. However comparability across existing devices has not been established. We evaluated in vivo and in vitro measurements made using three commercial devices against a custom-built device (WIMR): Resmon Pro Diary (Restech srl, Italy), tremoFlo C-100 (Thorasys Medical Systems, Canada), Jaeger Masterscope CT IOS (CareFusion, Hoechberg, Germany). Respiratory system resistance Rrs and reactance Xrs at 5 Hz were examined in twelve healthy subjects (mean age 33 ± 11 years, 7 males), and in two test standards of known resistance and reactance. Subjects performed three measurements during tidal breathing on the four devices in random order. Total, inspiratory and expiratory Rrs and Xrs were calculated and compared using one-way repeated measures ANOVA and Bonferroni post-hoc tests. Rrs did not differ between devices, with <10% deviation from predicted, except for the IOS device. With Xrs, similar values were seen between the WIMR and Resmon devices and between the tremoFlo and IOS devices. No differences were observed using test standards; deviation from theoretical value was <2% for resistance and <5% for reactance. The WIMR, tremoFlo and Resmon Pro but not IOS devices measure similar Rrs, whereas there was more disparity across devices in the estimation of Xrs parameters. The discrepancy between in vivo and in vitro measurements suggest that FOT validation procedures need to take into account the breathing pattern, either using biological controls or a breathing model.


Forced oscillation technique, Airway physiology, Pulmonary function testing, Resistance, Monitoring

Zimmermann SC, Watts JC, Bertolin A, Jetmalani K, King GG, Thamrin C. Discrepancy between in vivo and in vitro comparisons of forced oscillation devices. J Clin Monit Comput. 2018;32:509–12