High-flow nasal cannulas (HFNCs) have become common devices for patients with respiratory failure who are treated in general wards. Few reports have been published on in-hospital mortality associated with the ratio of oxygen saturation (ROX) index, measured by pulse oximetry/fraction of inspired oxygen to respiratory rate, in patients treated with HFNCs. We aimed to examine in-hospital mortality and associated factors in patients who initiated HFNC use in a general ward. Sixty patients who initiated HFNC use in general wards at Kobe University Hospital between December 2016 and October 2020 were retrospectively enrolled. We assessed in-hospital mortality, comorbidities, and ROX index. The in-hospital mortality was 48.3%, and ROX index values were significantly lower in patients who died than in those who did not (at HFNC oxygen therapy initiation; 6.93 [2.73-18.5] vs. 9.01 [4.62-18.1], p = 0.00861). Although the difference was not statistically significant, the change in ROX index values between HFNC initiation and 12 hours after initiation tended to be greater in the patients who died in the hospital (0.732 [-2.84-3.5] vs. -0.35[-4.3-2.6], p = 0.0536). Lower ROX index values may be associated with the in-hospital death of patients who are treated with HFNCs in general wards.
Keywords: High flow nasal cannula; In-hospital mortality; Quality of life; ROX index.