Optimal flow settings during high-flow nasal cannula (HFNC) therapy are unknown. We investigated the optimal flow settings during HFNC therapy based on breathing pattern and tidal inspiratory flows in patients with acute hypoxemic respiratory failure (AHRF).

We conducted a prospective clinical study in adult hypoxemic patients treated by HFNC with a fraction of inspired oxygen (F _I O ₂ ) ≥ 0.4. Patient’s peak tidal inspiratory flow (PTIF) was measured and HFNC flows were set to match individual PTIF and then increased by 10 L/min every 5–10 min up to 60 L/min. F _I O ₂ was titrated to maintain pulse oximetry (SpO ₂ ) of 90–97%. SpO ₂ /F _I O ₂ , respiratory rate (RR), ROX index [(SpO ₂ /F _I O ₂ )/RR], and patient comfort were recorded after 5–10 min on each setting. We also conducted an in vitro study to explore the relationship between the HFNC flows and the tracheal F _I O ₂ , peak inspiratory and expiratory pressures.

Forty-nine patients aged 58.0 (SD 14.1) years were enrolled. At enrollment, HFNC flow was set at 45 (38, 50) L/min, with an F _I O ₂ at 0.62 (0.16) to obtain an SpO ₂ /F _I O ₂ of 160 (40). Mean PTIF was 34 (9) L/min. An increase in HFNC flows up to two times of the individual patient’s PTIF, incrementally improved oxygenation but the ROX index plateaued with HFNC flows of 1.34–1.67 times the individual PTIF. In the in vitro study, when the HFNC flow was set higher than PTIF, tracheal peak inspiratory and expiratory pressures increased as HFNC flow increased but the F _I O ₂ did not change.

Mean PTIF values in most patients with AHRF were between 30 and 40 L/min. We observed improvement in oxygenation with HFNC flows set above patient PTIF. Thus, a pragmatic approach to set optimal flows in patients with AHRF would be to initiate HFNC flow at 40 L/min and titrate the flow based on improvement in ROX index and patient tolerance.

Trial registration : ClinicalTrials.gov (NCT03738345). Registered on November 13th, 2018. NCT03738345?term=NCT03738345&draw=2&rank=1