Performance of low-flow oxygen delivery systems with mouth breathing and altered respiratory mechanics: a bench study.

<bold>BACKGROUND:</bold> Scientific gaps exist regarding the efficacy and optimal use of low-flow oxygen delivery systems in various clinical scenarios. The aim of this study was to compare the FiO₂ measured by different low-flow systems in breathing conditions with opened or closed mouth, and in a normal, restrictive or obstructive respiratory pattern. <bold>Methods:</bold> An active artificial lung (ASL 5000) was connected to a three-dimensional printed head sequentially equipped with nasal canula (NC), simple facemask (FM), non-rebreathing facemask (NRM), NC + surgical facemask (NC_Sfm) or NC + double-trunk mask (NC_Dtm). Oxygen flow supplied NC, NC_Sfm and NC_Dtm from 1 to 10L/min, FM from 5 to 10L/min and NRM at 10L/min. The experiments were carried out with three lung mechanics (normal, obstructive and restrictive) with and without mouth closed. The FiO₂ sampled in the artificial lung was measured. <bold>Results:</bold> From 1 to 4 L/min of O₂ flow, the mean ± SD FiO₂ was 29 ± 6; 39 ± 9 and 48 ± 2% with NC, NC_Sfm and NC_Dtm, respectively (p<0.01). From 5 to 9 L/min of O2, the mean ± SD FiO₂ was 45 ± 11; 59 ± 13; 69 ± 13 and 41 ± 8% with NC, NC_Sfm and NC_Dtm and FM, respectively (p<0.01). At 10 L/min, the NC_Dtm significantly outperformed all other systems (FiO₂ = 69 ± 13%) and was followed by the NRM (59 ± 9%). The results remained consistent regardless of the mouth condition or the set breathing pattern. <bold>Conclusions:</bold> For a given oxygen flow rate, FiO₂ values delivered by NC_Dtm was higher than those obtained with other oxygen delivery systems. This statement remains true regardless of the condition studied.