Physiological Effects of High-Flow Nasal Oxygen Compared With Conventional Oxygen Therapy: A Randomized Trial in...

To compare the physiologic effects of high-flow nasal oxygen (HFNO) and conventional oxygen therapy (COT) on respiratory effort and ventilation distribution in the first 24 hours postextubation.

Two ICUs in The Netherlands ( Clinicaltrials.gov : NCT05652699).

Adult patients invasively ventilated for greater than or equal to 48 hours were included after meeting weaning readiness criteria. The main exclusion criteria were do-not-reintubate orders and contraindications for esophageal pressure (P ES ) catheter insertion.

Randomization between HFNO and COT for 24 hours postextubation.

P ES measurements and electrical impedance tomography (EIT) were performed before extubation and up to 24 hours postextubation. The primary endpoint was the change in ΔP ES at 24 hours postextubation as compared with baseline. The secondary endpoints were simplified P ES pressure-time-product (sPTP ES ) and EIT-derived ventilation parameters postextubation. After 42 of 54 targeted patients were included, inclusion was stopped because of slow recruitment. The 24-hour study protocol was completed without crossover in 20 of 22 patients (91%) receiving HFNO and in 11 of 20 patients (55%) receiving COT. At 24 hours postextubation, the change in ∆P ES from baseline was 1.5 (0.0; 5.1) and -1.3 cm H 2 O (-2.8; 2.2) in the COT and HFNO groups, respectively ( p = 0.11). The change in sPTP ES /min was 66 (-7; 111) and 0 cm H 2 O·s/min (-47; 43) in the COT and HFNO groups, respectively ( p = 0.04). Baseline ∆P ES and sPTP ES correlated inversely with ∆P ES and sPTP ES postextubation (∆P ESr = -0.6, p < 0.01, and sPTP ES /min r = -0.6, p < 0.01) in the HFNO group but not the COT group ( p > 0.05). There were no relevant differences in the EIT-derived ventilation parameters.

There was no difference between HFNO and COT on respiratory effort and ventilation distribution in the first 24 hours postextubation. Measures of respiratory effort were numerically but not always statistically lower in the HFNO group compared with the COT group.