High-flow nasal cannula in COVID-19: the HiFLo-Covid randomized controlled trial


Non-invasive modalities of respiratory support are increasingly being employed in the management of acute hypoxemic respiratory failure in COVID-19, considering early reports that suggested high mortality with invasive mechanical ventilation.1 Respiratory support with high-flow nasal cannula (HFNC) has been found to be associated with reduced 90-day mortality in non-COVID-19-related acute hypoxemic respiratory failure.2

In a previous observational study from France, 379 patients with COVID-19-related acute respiratory failure were evaluated. In this study, 146 received HFNC, while 233 did not. The use of HFNC was associated with a significantly lower requirement for invasive ventilation at day 28 (56% vs. 75%; P < 0.0001); however, the 28-day mortality was not significantly different.3  In another observational cohort from South Africa, consecutive patients with COVID-19-related acute hypoxemic respiratory failure were studied. In this study, 137/293 (47%) patients were successfully weaned off HFNC therapy. A higher ROX score (the ratio of oxygen saturation/FiO2 to the respiratory rate) at 6 hours after the commencement of HFNC was associated with successful weaning. Overall, 139 patients (52%) survived to hospital discharge.4 A randomized controlled trial (RCT) from Italy compared helmet non-invasive ventilation (NIV) for a minimum duration of 48 hours followed by HFNC with high-flow oxygen alone at 60 L/min. The authors observed no significant difference in respiratory support-free days at 28 days. However, the requirement for endotracheal intubation was significantly lower with helmet NIV compared to HFNC. The median number of invasive ventilation-free days was also significantly higher in the helmet NIV group.5

Against this background, the investigators of the HiFLo-Covid RCT compared the efficacy of HFNC with conventional oxygen therapy in patients with COVID-19-related acute respiratory failure.

Design and setting 

The HiFLo-Covid randomized controlled trial was conducted across three centers in Columbia over a 5-month period between August 2020 to January 2021.6 Patients were randomized in a 1:1 ratio to high flow nasal cannula (HFNC) or conventional oxygen therapy. The trial was unblinded, but the main investigators were unaware of study group outcomes until the database was unlocked. A web-based randomization was carried out with stratification by study site.  


The study included adult patients in acute respiratory failure with suspected or confirmed SARS-CoV-2 with a P/F ratio of less than 200 mm Hg and respiratory distress. Clinical signs of respiratory distress included the use of accessory muscles and tachypnea with a respiratory rate of more than 25/min. The time interval between fulfilling criteria for respiratory failure and randomization was less than 6 hours. 


The study excluded patients with hypercapnia (PaCO2 more than 55 mm Hg), poor LV function with EF less than 40%, advanced COPD, hospitalization for COPD within the last year, peripheral demyelinating disease, advanced cirrhosis, and anatomical abnormalities that precluded the use of HFNC. Patients who were not expected to survive and pregnant patients were also excluded. 

HFNC arm

HFNC therapy was commenced at a flow of 60 l/min and FiO2 of 1.0. FiO2 was adjusted to maintain oxygen saturation of 92% or higher. The flow rate was decreased if the patient experienced discomfort. HFNC therapy was continued until criteria for weaning of therapy or intubation were met. Criteria for weaning included improvement in respiratory distress, a P/F ratio higher than 200 mm Hg, and oxygen saturation of 92% or more with less than 9 l/min of conventional oxygen therapy. If hypoxemia recurred after weaning, HFNC therapy could be recommenced unless immediate intubation was required. 

Conventional oxygen arm

In the conventional oxygen therapy arm, any low-flow device, including nasal prongs, masks (with or without reservoir), or venturi masks were used alone or in combination. The flow rate and FiO2 were adjusted to maintain an oxygen saturation of 92% or more. Oxygen therapy was continued until recovery or intubation.

Common management

Patients could be placed in the prone position for improvement of oxygenation at the discretion of the clinician. The decision for intubation was based on pre-specified criteria. Patients who were on HFNC were continued on this therapy during laryngoscopy and intubation. The use of HFNC was allowed as appropriate in both groups after extubation. Non-invasive ventilation was not used in the study, based on local recommendations. Steroids, antibiotics, and antivirals were administered as appropriate. Patients were followed up until 28 days; when the hospital discharge occurred before 28 days, follow-up was carried out through a telephonic interview. 

The authors calculated the sample size based on an intubation rate of 60% according to their previous experience. A sample size of 196 patients was calculated to demonstrate a reduction in the intubation rate by 20% in the HFNC arm (80% power, 2-sided alpha of 0.05). For a reduction in the time to recovery by 2 days, the calculated sample size was 160 patients (80% power, 2-sided alpha of 0.05). Thus, a sample size of 196 patients was chosen. However, during the course of the pandemic, patients had to be transferred from study centers to non-study centers for continued care due to logistical reasons. Continued protocolized care and follow-up were difficult in these patients, and hence, they were excluded from the final analysis. Hence, the final sample size was re-set to 220 patients to allow for transfer-outs. 

The investigators screened 652 patients; 432 were excluded for various reasons. Thus, 220 patients were randomized. In the final analysis, 99 were included in the HFNC group and 100 in the conventional oxygen therapy group. Six patients from the HFNC group and seven from the conventional oxygen therapy group were transferred to another facility. 


The study evaluated two primary outcomes. 

Co-primary outcome 1. Intubation rate within 28 days: significantly lower with HFNC: 34/99 (34.3%) vs. 51/100 (51%); (hazard ratio: 0.62; 95% CI, 0.39–0.96; P = .03) 

Number of patients who experienced clinical recovery in 28 days: 77/99 (77.8%) vs. 71 (71%)

Co-primary outcome 2. Time to clinical recovery in 28 days: Defined as a reduction by 2 or more points on a modified 7-category ordinal scale from the time of randomization. The median time to clinical recovery within 28 days was significantly lower with the use of HFNC compared to conventional oxygen therapy; 11 (IQR, 9–4) days in patients randomized to high-flow oxygen therapy vs 14 (IQR, 11–19) days in those randomized to conventional oxygen therapy (hazard ratio, 1.39; 95% CI, 1.00–1.92; P = 0.047)

Secondary outcomes 

The number of patients intubated within 7 days (31.3 vs. 50%; p = 0.03), and 14 days (34% vs. 51%; p = 0.04) were significantly lower with HFNC compared with conventional oxygen therapy. The median number of ventilation-free days at 28 days was significantly higher with HFNC [28 (19–28) vs. 24 (14–28), P = 0.01] The number of days off renal replacement therapy, and the ICU and hospital length of stay were not different between groups. Mortality at days 14 and 28 was also not different between the two groups. Suspected bacterial pneumonia occurred in 13 patients (13.1%) in the HFNC group and 17 (17.0%) in the conventional oxygen therapy group. Bacteremia occurred in seven patients (7.1%) with HFNC and 11 patients (11%) with conventional oxygen therapy. 

Tertiary outcomes 

The duration of HFNC or conventional oxygen therapy prior to intubation was not significantly different. The duration of time with the requirement of FiO2 of more than 70% was significantly higher with conventional oxygen therapy. The incidence of extrapulmonary organ involvement at 7 days not different, based on the SOFA scores. 

Subgroup analysis 

On subgroup analysis, the requirement for intubation was lower with HFNC compared to conventional oxygen therapy among patients less than 60 years old. In patients 60 years or older, the difference was non-significant. Among patients with an IL-6 level of less than 100 pg/ml, the requirement for intubation was less in the HFNC compared to the conventional oxygen group. The difference was not significant between groups in patients with IL-6 levels ≥100 pg/ml. There was no difference in intubation rates between groups among patients with a P/F ratio of 100 mm Hg or more compared to less than 100 mm Hg. There was no difference between subgroups regarding the time to clinical recovery. All subgroup analyses were exploratory and not adjusted for possible confounders. 

Post-hoc sensitivity mixed-model analysis

On post-hoc analysis using the study site as random effect, there was no difference in the primary outcome. Regardless of the study site, HFNC reduced intubation rates and increased the probability of clinical recovery compared to conventional oxygen therapy. 


The HiFLo-Covid RCT revealed a reduced requirement for intubation and less time to clinical recovery in patients with COVID-19-related acute respiratory failure in patients treated with HFNC compared to conventional oxygen therapy. Against the background of previous observational studies, this study offers more robust evidence that the use of HFNC may be an effective intervention in the alleviation of hypoxemia with improved clinical outcomes in COVID-19 pneumonia. However, there are a few limitations that need to be considered.  

The authors assumed a 20% reduction in intubation rates with HFNC; this difference may have been unrealistically large and resulted in a reduced sample size. The sample size was calculated based on an intubation rate of 60%. However, the actual intubation rate was 42.7%, thus resulting in an underpowered study. Non-invasive ventilation was not employed in this study. NIV is widely used as a support modality in acute hypoxemic respiratory failure due to COVID-19; however, NIV was not used in this study. The use of NIV may have reduced intubation rates in both groups of patients. Thirteen patients were transferred to non-study centers and were not included in the primary analysis. The study excluded a large number of patients 432/652 (66.2%) after initial screening. Considering that the study was conducted in only three centers in Columbia, it has limited external validity. Longer-term outcomes were not evaluated; the sample size was too small to assess any difference in mortality. The study was unblinded due to the practical difficulty of masking the intervention; this could have led to possible bias. 


1.         Richardson S, Hirsch JS, Narasimhan M, et al. Presenting Characteristics, Comorbidities, and Outcomes Among 5700 Patients Hospitalized With COVID-19 in the New York City Area. JAMA. 2020;323(20):2052-2059. doi:10.1001/jama.2020.6775

2.         Frat JP, Thille AW, Mercat A, et al. High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure. N Engl J Med. 2015;372(23):2185-2196. doi:10.1056/NEJMoa1503326

3.         Demoule A, Vieillard Baron A, Darmon M, et al. High-Flow Nasal Cannula in Critically III Patients with Severe COVID-19. Am J Respir Crit Care Med. 2020;202(7):1039-1042. doi:10.1164/rccm.202005-2007LE

4.         Calligaro GL, Lalla U, Audley G, et al. The utility of high-flow nasal oxygen for severe COVID-19 pneumonia in a resource-constrained setting: A multi-centre prospective observational study. EClinicalMedicine. 2020;28. doi:10.1016/j.eclinm.2020.100570

5.         Grieco DL, Menga LS, Cesarano M, et al. Effect of Helmet Noninvasive Ventilation vs High-Flow Nasal Oxygen on Days Free of Respiratory Support in Patients With COVID-19 and Moderate to Severe Hypoxemic Respiratory Failure: The HENIVOT Randomized Clinical Trial. JAMA. 2021;325(17):1731-1743. doi:10.1001/jama.2021.4682

6.         Ospina-Tascón GA, Calderón-Tapia LE, García AF, et al. Effect of High-Flow Oxygen Therapy vs Conventional Oxygen Therapy on Invasive Mechanical Ventilation and Clinical Recovery in Patients With Severe COVID-19: A Randomized Clinical Trial. JAMA. 2021;326(21):2161-2171. doi:10.1001/jama.2021.20714

Top 35 critical care blogs from feedspot: https://blog.feedspot.com/critical_care_blogs/

Leave a Reply