The Use of Non-Rebreather Masks on Top of High-Flow Nasal Cannula at Maximal Settings

Abstract

Background: During the COVID-19 pandemic, we frequently were able to support patients with singular, severe COVID-19 pneumonia with high-flow nasal cannula (HFNC) with settings of FIO₂ 0.90 and flows of 60 L/m. When the O₂ saturation dropped into the 80s, our respiratory therapist placed non-rebreather masks (NRB) to increase the O₂ saturations. Of note, these patients with exceptionally low O₂ saturations were surprisingly asymptomatic. In absence of data to support this practice, we collected data on our patients to document the benefit.

Methods: Consecutive patients with uncomplicated COVID-19 pneumonia requiring HFNC with settings of FIO₂ 0.90 and flows of 60 L/m and have not been previously placed on a NRB were eligible for the study. An oximetry recorder was placed on the patient and oximetries were recorded for one hour. A NRB was placed on the patient and oximetries were recorded for another hour. There was an average of 2,000 data points per hour. The oximetry data was downloaded into a database and the average oximetry before and after NRB was calculated. Respiratory rates, heart rates, and perfusion indexes were collected before, at the time, and after the placement of the NRB.

Results: For individual patients, there were 8 that had significant differences comparing before vs after. Of these 8, 6 had higher readings after, and 2 had higher readings before. Of the 3 differences that were not significant, all were slightly higher before (see table). The differences of the averages of all before and after patients was 1.79% (P = NS). The figure below shows each patient listed in order with before then after distributions. P-values were calculated using a Wilcoxon two-sample test. Only patient 10 had a dramatic difference. At trial start, his O₂ saturation was 92% and his respiratory rate was 24. At NRB placement, his O₂ saturation was 88% and at the end the O₂ saturation returned to 92% and his respiratory rate was 24. He received appropriate respiratory interventions during his low saturation episodes. The perfusion index did not explain the difference.

Conclusions: The addition of a NRB to HFNC did not make a difference from a statistical standpoint, even with the dramatic improvement with patient 10. The improvement in patient 10 was mostly due to appropriate respiratory care interventions, with the NRB having a small role. The overall increase of 1.79 percentage points in O₂ saturation is also not clinically important.

• Download figure
• Open in new tab
• Download powerpoint