O060 A non-invasive method to quantify work of breathing during sleep

Abstract Introduction Achieved versus attempted inspiratory work of breathing (WOB) can be determined from achieved versus attempted inspiratory pressure-volume curves. These are desirable metrics to quantify breathing effort and wasted effort in obstructive sleep apnoea (OSA) and other respiratory disorders. However, WOB requires invasive measurements not practical in routine clinical practice. This study is investigating the feasibility of non-contact mattress sensors and physiology-based analytical methods to non-invasively estimate WOB breath-by-breath. Methods This ongoing study involves people with and without OSA undergoing conventional in-laboratory polysomnography along with direct inspiratory airflow, mask, oesophageal (Poes) and epiglottic (Pepi) pressures via a mask and pressure catheters. A piezoelectric sensor beneath the participant’s chest simultaneously records breathing motion. A novel nonlinear three-compartment model of respiratory mechanics is used for direct breath-by-breath measures of breathing frequency and achieved and attempted WOB from intrusive signals, and for indirect estimates from the mattress sensor. Results Data from the first participant included over 6000 breaths. Respiratory rate estimated from the mattress sensor was 13.5 breaths/min [95% CI: 12.9–14.1] versus 12.6 breaths/min [95% CI: 12.0–13.2] from Poes. Peak inspiratory effort from the mattress signal occurred 0.36 seconds earlier than from Poes, and 0.03 seconds earlier than from Pepi. The correlation between mattress-derived and direct Poes-derived WOB was r = 0.82 [95% CI: 0.85–0.88], and for attempted WOB was r = 0.80 [95% CI: 0.84–0.87]. Conclusions These preliminary findings support the potential for non-contact sensors to continuously estimate WOB in OSA and other forms of sleep-disordered breathing.