key: cord-1036855-b8r25ze8
authors: He, Yuewen; Liu, Na; Zhuang, Xuhui; Wang, Xia; Ma, Wuhua
title: High-flow nasal cannula versus noninvasive ventilation in patients with COVID-19: a systematic review and meta-analysis
date: 2022-03-23
journal: Ther Adv Respir Dis
DOI: 10.1177/17534666221087847
sha: 94b6b4ad8b20ad49ee3ae6ccb5ba5299bf220946
doc_id: 1036855
cord_uid: b8r25ze8

BACKGROUND: During the novel coronavirus disease 2019 (COVID-19) pandemic raging around the world, the effectiveness of respiratory support treatment has dominated people’s field of vision. This study aimed to compare the effectiveness and value of high-flow nasal cannula (HFNC) with noninvasive ventilation (NIV) for COVID-19 patients. METHODS: A comprehensive systematic review via PubMed, Web of Science, Cochrane, Scopus, WHO database, China Biology Medicine Disc (SINOMED), and China National Knowledge Infrastructure (CNKI) databases was conducted, followed by meta-analysis. RevMan 5.4 was used to analyze the results and risk of bias. The primary outcome is the number of deaths at day 28. The secondary outcomes are the occurrence of invasive mechanical ventilation (IMV), the number of deaths (no time-limited), length of intensive care unit (ICU) and hospital stay, ventilator-free days, and oxygenation index [partial pressure of arterial oxygen (PaO(2))/fraction of inhaled oxygen (FiO(2))] at 24 h. RESULTS: In total, nine studies [one randomized controlled trial (RCT), seven retrospective studies, and one prospective study] totaling 1582 patients were enrolled in the meta-analysis. The results showed that the incidence of IMV, number of deaths (no time-limited), and length of ICU stay were not statistically significant in the HFNC group compared with the NIV group (ps = 0.71, 0.31, and 0.33, respectively). Whereas the HFNC group performed significant advantages in terms of the number of deaths at day 28, length of hospital stay and oxygenation index (p < 0.05). Only in the ventilator-free days did NIV show advantages over the HFNC group (p < 0.0001). CONCLUSION: For COVID-19 patients, the use of HFNC therapy is associated with the reduction of the number of deaths at day 28 and length of hospital stay, and can significantly improve oxygenation index (PaO(2)/FiO(2)) at 24 h. However, there was no favorable between the HFNC and NIV groups in the occurrence of IMV. NIV group was superior only in terms of ventilator-free days.

The outbreak of novel coronavirus disease 2019 (COVID-19) has caused untold harm and challenges to people in more than 200 countries and territories around the world. To this day, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) continues to ravage the planet, with more than 270 million people infected with COVID-19 worldwide. 1 Pneumonia caused by SARS-CoV-2 differs from community-acquired pneumonia (CAP) and hospital-acquired pneumonia (HAP)/ ventilator-associated pneumonia in that it is highly transmissible from person to person and can even infect people asymptomatically. 2, 3 Patients with severe COVID-19 can trigger acute respiratory distress syndrome (ARDS), 4, 5 which can progress to acute respiratory failure (ARF), manifested by severe hypoxemia and dyspnea. What's more, hypoxemia is associated with more rapid disease progression and higher mortality. Currently, the choice of ventilation support treatment for patients with COVID-19 is particularly important.

Traditionally, when encountering severe respiratory diseases, medical staff would first think of invasive mechanical ventilation (IMV), because years of experience have shown that it is a potential intervention to save the lives of acute patients. However, years of clinical experience have taught us that IMV is a risk factor for the development of ventilator-associated pneumonia (VAP). 6 Longterm IMV is a major risk factor for patients with hospital-acquired infections, which may worsen the patient's condition and is associated with a high mortality rate in hospital. 7 Noninvasive ventilation (NIV) can reduce intubation rate in COVID-19 patients with ARF, shorten length of stay, and reduce in-hospital mortality. NIV is the delivery of a specific flow of fresh gas to the lungs without an invasive endotracheal airway. It can both give a certain flow of oxygen alone via a nasal cannula or a face mask and can also be connected to a ventilator and used in combination with the PEEP (positive end-expiratory pressure) or CPAP (continuous positive airway pressure) mode of the ventilator to give the patient noninvasive positive pressure ventilation (NIPPV) support. 8 Although oxygen flow is limited to no more than 15 l/min with conventional NIV equipment, the high flow nasal cannula (HFNC) is a relatively new respiratory support technology that serves as an alternative to standard oxygen therapy by delivering an air/oxygen mixer and connecting to the nasal cannula via an actively heated humidifier. It provides adequate heated and humidified medical gases at a flow rate of 15-70 l/min via the nasal route, provides positive airway pressure effect, and reduces anatomic dead space, respiratory rate, and work of breathing. 9,10 Therefore, HFNC is considered to have several physiological advantages compared with other standard oxygen therapies. 11 In actual clinical practice, the physician can choose different forms of respiratory support depending on the patient's condition, such as noninvasive mechanical ventilation (NIMV), HFNC, helmet ventilation, and IMV. Individualized ventilation support treatment for COVID-19 patients is of great significance to avoid endotracheal intubation, reduce the mortality of intensive care unit (ICU) patients, and improve the prognosis of patients.

The fact that COVID-19 will continue to exist and expand its impact has been recognized by experts around the world. According to our data search, the amount of meta-analysis that compares the ventilation strategies of COVID-19 patients is very limited. Therefore, it is essential to compare the various ventilation treatment strategies suitable for patients with COVID-19. On this premise, we conducted this systematic review and meta-analysis, with the purpose of comparing HFNC with NIV and exploring which one can better reduce the occurrence of IMV and the death at 28 days of COVID-19 patients.

This systematic review and meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We followed the PRISMA checklist to complete the current meta-analysis. To identify studies comparing the efficacy of HFNC with conventional oxygen therapy in COVID-19 patients, two investigators (Y.W.H. and X.W.) systematically searched PubMed, Web of Science, Cochrane, Scopus, WHO database, China Biology Medicine Disc (SINOMED), and China National Knowledge Infrastructure (CNKI) databases for relevant articles published before 20 October 2021. The whole retrieval process is shown in Figure 1 In this process, we rated the studies by using the starring feature of EndNote X9. Two investigators marked the studies with low relevance as 'one star', the controversial research that required two investigators to re-screen them marked as 'two or three stars', and the undoubted study is 'four or five stars'. The studies with one star can be excluded at once by a single investigator, two or three stars' studies need extra evaluation by all the investigators, and the research with four or five stars can be included in this meta-analysis. We used the modified methodological index for nonrandomized studies (MINORS) score 12, 13 to evaluate the quality of nonrandomized controlled trials (NRCTs) and excluded studies with a total score of ⩽12. Finally, two researchers identified the included literature based on the full text. When the results of the two researchers diverged, the opinion of a researcher (W.H.M.) was used to reach a consensus. Figure 1 includes a screening process to illustrate the number of studies that we exclude at each stage.

For the inclusion of this systematic review and meta-analysis, studies had to meet the following criteria: randomized controlled trials (RCTs), observational cohort studies, or retrospective The risk of bias summary is shown in Figure 2 . When researchers disagree on the biased analysis of the same study, another researcher (W.H.M.) will make the decision.

Three investigators independently extracted the relevant information and data. All differences are double-checked, and W.H.M. was responsible for handling different points of view. According to the modified MINORS score, we will analyze the data included in the NRCT and complete the quality assessment (Table 1) . We extracted the following data based on the characteristics of the included studies -groups, nationals, type of hospital, ages, gender, body mass index (BMI), and comorbidities -and summarized in Table 2 .

And also, we extracted the intervention characteristics of studies in Table 3 .

The primary outcome is the number of deaths at day 28. The secondary outcomes are the occurrence of IMV, the number of deaths (no timelimited), length of ICU and hospital stay, ventilator-free days, and oxygenation index [partial pressure of arterial oxygen (PaO 2 )/fraction of inhaled oxygen (FiO 2 )] at 24 h.

RevMan 5.4 computer software was used for all data analysis in this study. For all the dichotomous data that need to be analyzed, we chose to report odds ratio (OR). And for continuous variables, the mean and standard deviation (SD) should be reported. If the median and interquartile range (IQR) are reported in the study, it can be converted into the mean and SD through formulas.

RevMan 5.4 also reported the heterogeneity of the data while producing the forest plot. For heterogeneity test p < 0.05 or I 2 > 50%, we choose random-effects model. When the heterogeneity test p > 0.05 or I 2 < 50%, the fixed-effects model is often selected. The fixed-effects model is used when the results of heterogeneity between subgroups are consistent, and the random-effects model is used when the results of heterogeneity are inconsistent. If the heterogeneity test result I 2 > 80%, we need to perform a sensitivity analysis on the data to exclude studies with significant heterogeneity. Only studies with scores ⩾12 can be included in the meta-analysis.

The protocol for this review has been published in Prospero, and the registration number is CRD42021289413. Figure 1 .

Since only one RCT can be retrieved, the inclusion criteria of this study are not limited to prospective studies. We use the MINORS to evaluate the quality of NRCT based on the researches of Vinuela et al. 23 and Slim et al. 12 Each item can get 0-2 points, and we will include studies with scores ⩾12 points into the meta-analysis. The modified MINORS score of all eligible NRCT is shown in Table 1 . For the only RCT, we assessed its risk of bias as described in the Cochrane handbook . Only performance bias is evaluated as high risk ( Figure 2 ).

In nine studies, 1582 patients were included in the meta-analysis. The investigators extracted the characteristics of the patients in the included studies, containing groups, nationals, type of hospital, ages, gender, BMI, and comorbidity (diabetes mellitus, hypertension, and coronary heart disease), in Table  2 . In the study of Jarou et al., 19 the number of patients in the NIV group was less than 30. In the study of Kabak et al. 20 and Duan et al., 21 the number of patients in both groups was less than 30. The data of four studies are reported using median (IQR). During the meta-analysis process, we converted median (IQR) to mean (SD) according to the method proposed in Wan et al. 24 6 journals.sagepub.com/home/tar 

The researchers extracted intervention characteristics of included studies; the main content are the oxygenation strategy and concomitant medications of the NIV and HFNC groups (Table 3) .

Only the study of Burnim et al. 18 did not show a specific oxygenation strategy, while the study of Kabak et al. 20 only proposed the use of nonrebreather masks and nasal cannula as a ventilation strategy for NIV. In three studies, 14,17,20 the NIV group used a mask or nasal cannula for ventilation. Bonnet et al., 15 Sayan et al., 16 (Figure 3(a) ). Using a fixed-effects model, these results reached statistical significance (p = 0.0005) and favored HFNC group. (Figure 3(b) ). Using a fixed-effects model, the result is not statistically significant (p = 0.31). Figure 5(a) ). Using a random-effects model, the result is not statistically significant (p = 0.33).

Length of hospital stay. Four studies with 308 patients (NIV group: 120; HFNC group: 188) reported the length of hospital stay. The pooled MD (95% CI) was 1.62 (0.05, 3.19) units, I 2 = 18% ( Figure 5(b) ). Using a fixed-effects model, these results reached statistical significance (p = 0.04) and favored HFNC group. (Figure 7) . Using a fixed-effects model, these results reached statistical significance (p < 0.00001) and favored HFNC group.

Three years have passed since the outbreak of COVID-19, which has shown the world how frightening and aggressive it can be. Early epidemiological studies 5,25,26 stated that 8.2% of cases suffer from rapid and progressive respiratory failure, similar to ARDS. In most severe cases, patients with COVID-19 require high probability of needing IMV, which is associated with high mortality. 4 Therefore, patients often need different degrees of noninvasive respiratory support treatment to maintain the arterial partial pressure of oxygen at a normal level. Our systematic review and meta-analysis yielded one RCT and eight high-quality NRCTs to evaluate the efficacy of different NIV strategies in the patients with COVID-19. By comparing the HFNC group with the NIV group, we found that the HFNC group was significantly better than NIV concerning the number of deaths at day 28, length of hospital stay, and oxygenation index at 24 h. The comparison between the HFNC and NIV groups was not statistically significant in terms of the incidence of IMV, number of deaths (no timelimited), and length of ICU stay. NIV group was superior only for the outcome of ventilator-free days.

How to improve patients' survival rate, reduce the occurrence of IMV, and improve the prognosis of patients are the questions that always have been discussed since the outbreak of the COVID-19. Severe COVID-19 patients usually undergo ARF. Therefore, the use of various NIMV methods to maintain the patient's FiO 2 at a high level has become the global consensus. 27, 28 NIV is an oxygen therapy modality that emerged in the 1990s of the last world. 29 Esteban et al. 30 conducted a nested comparative study in 349 ICUs in 23 countries, and found out that NIV usage rate increased from 4% in 1998 to 11% in 2004. As NIV has long been in widespread clinical use, it has become the favorable option for ventilation support treatment of patients with COVID-19. 31 However, the shortcomings of NIV were also quickly exposed. An RCT carried out by Lemiale et al. 32 found out that early use of NIV did not reduce 28-day mortality in immunocompromised patients with ARF. With conventional equipment for NIV, oxygen flow is limited to no more than 15 l/min and FiO 2 is journals.sagepub.com/home/tar 11 often unstable, which make less effective than expected. 33 This is what made HFNC therapy received considerable attention when it was introduced in the 2000s as a respiratory support treatment for patients with severe respiratory failure. By providing a constant gas flow rate of up to 70 l/min, 34 HFNC can improve the ability to flush the dead space of the upper airway better than NIV and reduce the patient's work of breathing. 35, 36 In a prospective study, Parke et al. 37 found that HFNC therapy could produce low levels of positive airway pressure, approximating the effect of PEEP. Despite the many advantages of HFNC as outlined above, there are two sides to the coin. Too much reliance on HFNC can lead to delayed intubation and increase mortality. 38 In the past 3 years, the shortage of ventilators following the outbreak of COVID-19 has led to a wider clinical use of this therapy. As you know, COVID-19 is transmitted mainly by respiratory droplets/aerosols. 39 The World Health Organization (WHO) 40 Interestingly, the NIV group has a significant advantage over the HFNC only in the aspect of ventilator-free days [mean difference = -2.26, 95% CI = (-3.18, -1.35), p < 0.00001]. There was no significant difference between HFNC and NIV in terms of the occurrence of IMV, length of ICU stay, and the number of long-term deaths, both of which resulted in a corresponding reduction in intubation rates and improved long-term prognosis.

To our knowledge, this is the first systematic review and meta-analysis to explore efficacy and value of these two types of oxygen strategies for patients with COVID-19. This study was created through an extensive search, iterative screening, and data extraction. We provide a comprehensive overview of the comparison between HFNC and NIV, with the aim of evaluating its true efficacy and benefit to patients by including thorough various assessment criteria. Meanwhile, we have taken into account the potential limitations of this meta-analysis. First, despite an extensive literature search, the limited number of relevant RCTs led to the inclusion of only one RCT and eight high-quality NRCTs. Although they passed the quality assessment, this may affect the accuracy of the results. Furthermore, COVID-19 patients in the eight included studies had different reasons for receiving ventilation therapy. Some of the studies included COVID-19 patients with ARF, and some have ARDS. Patients with different medical backgrounds receiving the same ventilation treatment may cause bias in the study outcomes. Finally, the timing of treatment as well as the pattern of treatment also vary between HFNC and NIV, which further increases heterogeneity.

For COVID-19 patients, the use of HFNC therapy is associated with the reduction of the number of deaths at day 28 and length of hospital stay, and can significantly improve oxygenation index (PaO 2 /FiO 2 ) at 24 h. However, there was no favorable option in the comparison between the HFNC and NIV groups in the occurrence of IMV. NIV group was superior only for the outcome of ventilator-free days. Large samples and high-quality clinical studies are still needed to evaluate different ventilation strategies for patients with COVID-19.

World Health Organization. COVID-19 weekly epidemiological update, edition 62. Geneva: World Health Organization

SARS-CoV-2 transmission from people without COVID-19 symptoms

A systematic review of asymptomatic infections with COVID-19

Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China

Prevention of ventilator-associated pneumonia

Clinical challenges in mechanical ventilation

Noninvasive ventilation in acute respiratory failure

High-flow nasal cannula oxygen therapy is superior to conventional oxygen therapy but not to noninvasive mechanical ventilation on intubation rate: a systematic review and meta-analysis

High-flow nasal cannula oxygen therapy in adults: physiological benefits, indication, clinical benefits, and adverse effects

The role for high flow nasal cannula as a respiratory support strategy in adults: a clinical practice guideline

Methodological index for non-randomized studies (minors): development and validation of a new instrument

The methodological quality assessment tools for preclinical and clinical studies, systematic review and meta-analysis, and clinical practice guideline: a systematic review

The value of high-flow nasal cannula oxygen therapy in treating novel coronavirus pneumonia

High flow nasal oxygen therapy to avoid invasive mechanical ventilation in SARS-CoV-2 pneumonia: a retrospective study

Impact of HFNC application on mortality and intensive care length of stay in acute respiratory failure secondary to COVID-19 pneumonia

Characteristics and outcomes of patients receiving high flow nasal cannula therapy prior to mechanical ventilation in COVID-19 respiratory failure: a prospective observational study

The effectiveness of high-flow nasal cannula in coronavirus disease 2019 pneumonia: a retrospective cohort study

Emergency department-initiated high-flow nasal cannula for COVID-19 respiratory distress

Feasibility of non-rebreather masks and nasal cannula as a substitute for high flow nasal oxygen in patients with severe COVID-19 infection

Use of highflow nasal cannula and noninvasive ventilation in patients with COVID-19: a multicenter observational study

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

Laparoscopic versus open distal gastrectomy for gastric cancer: a meta-analysis of randomized controlled trials and high-quality nonrandomized studies

Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range

Respiratory function in patients post-infection by COVID-19: a systematic review and meta-analysis

ECMO for ARDS due to COVID-19

Clinical consensus recommendations regarding non-invasive respiratory support in the adult patient with acute respiratory failure secondary to SARS-CoV-2 infection

Noninvasive ventilation

Evolution of mechanical ventilation in response to clinical research

COVID-19 pandemic and non invasive respiratory management: every Goliath needs a David. An evidence based evaluation of problems

Effect of noninvasive ventilation vs oxygen therapy on mortality among immunocompromised patients with acute respiratory failure: a randomized clinical trial

High flow nasal cannula versus conventional oxygen therapy and non-invasive ventilation in adults with acute hypoxemic respiratory failure: a systematic review

Clinical evidence on high flow oxygen therapy and active humidification in adults

Nasal high flow clears anatomical dead space in upper airway models

Nasal high flow reduces dead space

Nasal high-flow therapy delivers low level positive airway pressure

Failure of highflow nasal cannula therapy may delay intubation and increase mortality

Pathophysiology, transmission, diagnosis, and treatment of coronavirus disease 2019 (COVID-19): a review

Clinical management of severe acute respiratory infection when novel coronavirus (2019-nCoV) infection is suspected: interim guidance. Geneva: World Health Organization

High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure

Feasibility and physiological effects of prone positioning in non-intubated patients with acute respiratory failure due to COVID-19 (PRON-COVID): a prospective cohort study

Efficacy and safety of early prone positioning combined with HFNC or NIV in moderate to severe ARDS: a multicenter prospective cohort study

Assessment of PaO 2 /FiO 2 for stratification of patients with moderate and severe acute respiratory distress syndrome

The authors received no financial support for the research, authorship, and/or publication of this article.

Wuhua Ma https://orcid.org/0000-0002-4078-5100