Predicting the successful application of high-flow nasal oxygen cannula in patients with COVID-19 respiratory failure: a retrospective analysis.

ABSTRACT Background The right time of high-flow nasal cannulas (HFNCs) application in COVID-19 patients with acute respiratory failure remains uncertain. Research design and methods In this retrospective study, COVID-19-infected adult patients with hypoxemic respiratory failure were enrolled. Their baseline epidemiological data and respiratory failure related parameters, including the Ventilation in COVID-19 Estimation (VICE), and the ratio of oxygen saturation (ROX index), were recorded. The primary outcome measured was the 28-day mortality. Results A total of 69 patients were enrolled. Fifty-four (78%) patients who intubated and received invasive mechanical ventilatory (MV) support on day 1 were enrolled in the MV group. The remaining fifteen (22%) patients received HFNC initially (HFNC group), in which, ten (66%) patients were not intubated during hospitalization were belong to HFNC-success group and five (33%) of these patients were intubated later due to disease progression were attributed to HFNC-failure group. Compared with those in the MV group, those in the HFNC group had a lower mortality rate (6.7% vs. 40.7%, p = 0.0138). There were no differences in baseline characteristics among the two groups; however, the HFNC group had a lower VICE score (0.105 [0.049–0.269] vs. 0.260 [0.126–0.693], p = 0.0092) and higher ROX index (5.3 [5.1–10.7] vs. 4.3 [3.9–4.9], p = 0.0007) than the MV group. The ROX index was higher in the HFNC success group immediately before (p = 0.0136) and up to 12 hours of HFNC therapy than in the HFNC failure group. Conclusions Early intubation may be considered in patients with a higher VICE score or a lower ROX index. The ROX score during HFNCs use can provide an early warning sign of treatment failure. Further investigations are warranted to confirm these results. Plain Language Summary High flow nasal cannulas (HFNCs) were widely used in patients with COVID-19 infection related hypoxemic respiratory failure. However, there were concerns about its failure and related delayed intubation may be associated with a higher mortality rate. This retrospective study revealed patients with higher baseline disease severity and higher VICE scores may be treated with primary invasive mechanical ventilation. On the contrary, if their baseline VICE score is low and ROX index is high, HFNCs treatment might be safely applied initially. The trends of serial ROX index values during HFNC use could be a reliable periscope to predict the HFNC therapy outcome, therefore avoided delayed intubation.