The use of high-flow oxygen therapy delivered via Airvo™ in the acute setting over a six-month period: A clinical perspective

Introduction

Oxygen delivery via nasal cannula was first implemented in the early 1940s in order to direct flow into the nose. The advantages of this were soon recognised, enabling the patient to eat, drink and speak, as well as limiting the feelings of claustrophobia often felt with a tight-fitting face mask. However, limitations were also recognised in terms of nasal discomfort, dryness and the provision of only lower levels of inspired oxygen to the patient’s minute ventilation (Ward 2016).

It is widely accepted that caution should be exercised when using nasal cannula with oxygen flow rates over 5–6L/minute due to the risk of mucosal dryness and damage. Additionally, in current clinical practice when using nasal cannula with low flow rates, this is not used in conjunction with humidification (Nishimura 2015).

In adults, HFOT was initially recognised using flows of up to 40L/minute, showing that a higher fraction of inspired oxygen (FiO₂) could be achieved with higher flows compared to the same flows with other interfaces due to the nasopharynx and oropharynx acting as internal anatomic reservoirs that increase the volume of inhaled oxygen as well as a wash out of dead space in the nasal passage. With higher flow rates there appears to be even greater washout of anatomical dead space (Ward 2016).

Over ten years ago HFOT devices were introduced with heated humidification systems allowing higher flow rates and oxygen concentrations, whilst minimising the effects of mucosal dryness and damage, enhancing patient comfort and tolerance (Ward 2016).

Landmark studies also report the ability for higher flows to generate distending pressures similar to those achieved with continuous positive airway pressure (CPAP) (Groves 2007; Park et al. 2009). In clinical practice CPAP has the effect of ‘splinting open’ alveoli, improving ventilation/perfusion matching and subsequent oxygenation whilst also decreasing work of breathing and can stabilise the chest wall in the presence of chest wall trauma. However, traditional CPAP is administered using either a tight-fitting face, nasal mask or the CPAP hood, and positive end expiratory pressure (PEEP) is achieved by placing a PEEP valve within the circuit. These interfaces can be uncomfortable for the patient and can impact on patient tolerance of the device and subsequent compliance. Care is needed to maintain skin integrity, particularly at the bridge of the nose, and often the patient can only tolerate wearing the mask for short periods of time, decreasing its overall benefit.

HFOT devices may generate clinically significant positive pressure depending on flow rate. Both Groves (2007) and Park et al. (2009, 2011a) found a positive correlation between high-flow rates and generation of PEEP. Groves (2007) report a significant positive end expiratory pressure of 3.2–5.2cm H₂O with flows of 40L/minute. This is also strongly influenced by whether the patient’s mouth is open or closed. Park et al. (2009b) found that with flows of 35L/minute, a PEEP of 2.7cm H₂O could be generated. It is accepted that an increased flow rate subsequently increases generated PEEP, with PEEP being further increased during HFOT when the patient’s mouth is closed. A possible explanation of this could be due to the higher resistance to expiration as the air is forced out through the nose. However, anecdotal evidence from clinical practice indicates that most patients find it difficult to nose breathe in times of respiratory distress, therefore the benefits of PEEP are reduced or negated.

Patient selection is varied ranging from application post-extubation, to supporting those patients with postoperative pulmonary complications (PPC) and for acute respiratory failure, to more long-term conditions such as exacerbations of chronic obstructive pulmonary disease and idiopathic pulmonary fibrosis (Braunlich et al. 2012; Maggiore et al. 2014). It also has demonstratable benefit with those patients that are not appropriate for invasive mechanical ventilation, including those in respiratory distress receiving active treatment and also in palliation (Peters et al. 2012).

Prior to commencing this work, in the author’s place of work, only three Airvo™ HFOT devices were in current use, one being in paediatrics. Within the department CPAP was the overriding modality of choice. Patients that required higher levels of oxygen were also managed with cold water humidification systems, the evidence for these being negligible (Ward 2016). In recognition of the growing body of literature to support the use of HFOT in a range of clinical areas, and as part of a service improvement initiative, the use of HFOT in the acute setting in a small tertiary hospital in New Zealand was implemented.

The intended aim was to have the option to use HFOT in patients who developed pulmonary complications such as atelectasis, retained secretions and subsequent increase in work of breathing and to report on the observed clinical outcomes in using it with particular emphasis on the clinical indications, patient suitability, machine settings and clinical outcomes. These findings could then inform the delivery of care for subsequent patients within the author’s place of work.

Method

Over a six-month period from October 2015 to April 2016, where clinical reasoning indicated HFOT was appropriate it was delivered using the Airvo™ 2 device manufactured by Fisher and Paykel Healthcare. This was used with patients on critical care and in the ward setting and delivery interface was by either nasal cannula, face mask or tracheostomy mask. In addition to the HFOT devices already available in the department, four Airvo™ 2 HFOT machines (plus consumables) were donated to the physiotherapy department by Fisher and Paykel Healthcare for the duration of the six months. Training was provided to physiotherapists by experienced staff working in the hospital to develop competence with clinical reasoning in use and set-up of HFOT machines.

A document on clinical application was made available for reference on the hospital intranet (Appendix 1). The service initiative had a prospective design and included referrals within the six-month period who presented with:

1. Oxygen saturation <94% on 4L nasal prongs and/or

2. Had signs of pulmonary complications (atelectasis, retained secretions, infection and chest x-ray changes).

Where clinical reasoning indicated that HFOT was required, initiation of HFOT via Airvo™ took place regardless of where the patient was situated in the hospital.

When HFOT was initiated, a proforma was completed by the clinician to ascertain the date it was commenced, patient diagnosis, clinical indications, respiratory rate, oxygen percentage and mode of delivery prior to the commencement of HFOT, initial settings and highest settings throughout treatment, oxygen saturations and respiratory rate whilst on HFOT, number of days the patient received HFOT and clinical outcomes (Appendix 2). HFOT was discontinued when either target saturations were reached on Airvo™ settings of 30L flow rate, FiO₂ 0.3; and/or the original clinical pathology was resolved, or active treatment was ceased, and end-of-life care was established.

Verbal informed consent was obtained from each patient in order to be assessed and treated by a physiotherapist which covered all aspects of treatment including HFOT. Where this was not possible the patient was treated in their best interests.

Consultants, charge nurses and the wider multi-disciplinary team were made fully aware of the nature and purpose of the initiative. Consultant agreement for the device to be used on patients under their care was obtained on an individual basis at the time of clinical assessment. Oxygen therapy was medically prescribed and charted along with target saturations for each patient. As this was a service improvement initiative, ethical approval was not required.

The six-month pilot was supported by Fisher and Paykel, where the equipment and consumables were donated for use. There was no payment to the author or department for this work to be completed and the author had no potential competing interest to disclose.

Results

Anonymised prospective data from 74 patients was obtained in the six-month period from October 2015 to April 2016. Another 13 patients fitted the criteria but were unable to receive HFOT due to availability of machines. These patients received alternative intervention.

Results indicate that HFOT was used extensively throughout the hospital with most wards utilising the device at some point, the majority of machines used on the surgical wards including the high dependency unit (Figure 1).

Pneumonia was the most common diagnosis. This was recorded as either their primary diagnosis or secondary to their reason for admission (Figure 2).

Clinical indicators for HFOT demonstrates that most patients were placed on the device for either consolidation, tenacious secretions, increasing oxygen requirements, increased work of breathing or a combination of the four (Figure 3).

Figure 4 highlights those patients that transitioned between CPAP and HFOT. Of the patients receiving CPAP prior to Airvo™, seven were non-compliant with the tighter fitting mask and were swapped to HFOT via nasal prongs. Three patients fluctuated between CPAP and HFOT, whilst two patients were weaned from CPAP to HFOT as a treatment progression.

The average fraction of inspired oxygen that patients received whilst on HFOT to achieve target saturations was FiO₂ 0.35. Mean starting flow rates were 41L/minute. There was a mean percentage improvement in patient’s saturations of 2.7% whilst receiving high flow.

Outcomes included that 75% of patients were successfully weaned from HFOT. There was 14% of patients that died either on HFOT or following escalation to CPAP or invasive ventilation (Figure 5). The patients in this group were all receiving flows of at least 45L/minute, with four patients receiving 50L of flow, FiO₂ 0.5 at the point of which a respiratory review by a physiotherapist was called for. Mean length of time spent on high flow was 4.37 days, however data was skewed by three patients who received it for 13, 18, and 24 days.

Clinical outcomes

Figure 5. Frequency of level of respiratory support prior to and following high flow oxygen therapy.

Not enough data was collected to compare alterations in work of breathing or respiratory rate before and during the use of HFOT. This was due to poor completion of the proforma. It was not planned for any qualitative data to be collected on patient comfort and compliance whilst on high-flow, although six patients were unable to tolerate high-flow via Airvo™ and were taken off.

As this work was carried out from October 2015 to April 2016, in a tertiary hospital in New Zealand, this was predominantly the summer months. However, apart from the month of October there was an overall steady use of HFOT. Patient numbers dipped slightly in February but peaked again in March with a sudden decline in the last month. The reasons for this were not recorded as part of the work.

Discussion

HFOT devices can be used in a range of clinical settings including post cardiac surgery, (Corley et al. 2011), in the emergency department (Lenglet et al. 2012), immediately post extubation (Maggiore et al. 2014), in the prevention of recurrent hospital admissions in patients with chronic obstructive pulmonary disease (Rae et al. 2010), and for palliation in the community setting, to name but a few. In this service initiative, clinical practice reflected the evidence base and it showed that HFOT use was widespread through clinical locations in the hospital, with all specialties utilising the device at some point within the six-month period. On Intensive Care, only 20% of patients used the Airvo™. This could be due to the fact that high flow can also be delivered through a ventilator. As this work only recorded information on high flow with the Airvo™ data for these patients was not available for analysis.

In the authors place of work, patients whose SpO₂ was ≤92% on 4L via nasal prongs and were starting to display other signs of pulmonary complications could be initiated onto HFOT. The author is aware that in many centres in the United Kingdom, patients may be initially escalated from nasal prongs to either a venturi system or higher levels of inspired oxygen via heated humidification in order to support increasing respiratory demands. Only after this stage is it commonplace for the patient to progress to HFOT. This may be partly protocol driven but may also be due to the availability of HFOT devices.

In this initiative, patients could be established onto HFOT on any ward within the hospital providing there had been the necessary training. In many UK centres protocols dictate that HFOT only be initiated and continued on designated wards that are able to support the acutely unwell respiratory patient on this level of support such as an Intensive Care Unit, High Dependency Unit, coronary care or the emergency department. The author argues that this New Zealand centre adopted a more proactive approach by initiating patients onto HFOT much earlier in their acute illness, and on any ward.

Patients required HFOT to help with either consolidation/tenacious secretions, increased work of breathing, increasing oxygen requirements, atelectasis or a combination of the four (Figure 3). This is supported by current research advocating HFOT in acute respiratory failure and for postoperative pulmonary complications due to the ability to heat and humidify gases, the option of higher flow rates to reduce dead space, and promote slower deeper breathing, improve dyspnoea, and the effect of PEEP enabling improvements in lung volumes and subsequent oxygenation (Roca et al. 2010; Ward 2010; Corley et al. 2011). It is also of interest that patients appeared to fit into one of two categories: those initially on lower levels of inspired oxygen, for instance those on one to 4l required HFOT primarily for humidification purposes, where in contrast those on higher levels of oxygen either through a face mask, venturi or non rebreathe mask required minimum flow rates of 45L/minute to match their inspiratory demand in the acutely unwell stage and to improve oxygenation and ventilation/perfusion matching.

Sztrymf et al. (2011) identified significant reductions in respiratory rates and increases in partial pressure of oxygen after 15 minutes. Dyspnoea scores decreased after 30 minutes in patients with acute respiratory failure who were switched from conventional oxygen therapy to HFOT. This work also demonstrated a mean improvement in patient oxygen saturations by 2.7% whilst receiving HFOT. In the author’s own clinical experience, respiratory variables in the first hour can often indicate if HFOT will be of any benefit. Unfortunately, due to poor completion of the proforma reductions in WOB and respiratory rate were not captured. This would have been valuable information to gather and the author recommends this for future studies. Patients at ward level did not have arterial lines in situ therefore the taking of arterial blood gases was not performed frequently enough to obtain suitable comparable data on improvements in PaO₂ post application of HFOT.

In this initiative seven patients (almost half) receiving facial CPAP were non-compliant and were switched to HFOT. Three patients alternated between CPAP and high flow (Figure 4). Although no current literature could be found on why patients would prefer to alternate between CPAP and HFOT, the author suggests this was to give the patient a break from the tight-fitting mask without compromising respiratory status. Two patients were transitioned from CPAP to high flow as a ‘step down’ as part of the weaning process. It is surmised that these patients required much higher and more accurate levels of PEEP than could be generated with high flow in the initial stages of their illness.

Studies indicate generation of PEEP with HFOT especially with higher flow rates and with the mouth closed. Stephan et al. (2015) in comparing HFOT with conventional non-invasive ventilation (NIV) via face mask noted that with 50l of flow and FiO₂ 0.5, HFOT was not inferior to NIV in patients with respiratory failure following cardiac surgery. Fratt et al. (2015) found that patients with PaO₂/FiO₂≤200mmHg receiving HFOT had a significantly lower 28 day intubation rate, as well as a significantly lower 90 day mortality rate compared to the those receiving NIV. Parke et al. (2011a) demonstrates that with 50l of flow, 3cm PEEP can be generated with the mouth closed. Current literature suggests higher flow rates to achieve successful outcomes. However, in clinical practice, some patients are unable to tolerate such pressures through the nasal passages and find it difficult to keep the mouth closed continuously therefore allowing some loss of PEEP. Patients who are in respiratory distress or asleep will tend to mouth breathe. Maggiore et al. (2014) demonstrated a decrease in respiratory rate and number of desaturation episodes as well as a reduced need for NIV or reintubation in people post extubation using starting flows of 50L/minute. Stephan et al. (2015) also advocates the use of HFOT to prevent intubation and to support the patient’s respiratory system in the critical phase post extubation. In this study starting flow rates of 50L/minute were used. In contrast, in this work patients were able to tolerate mean starting flow rates of 41L/minute, with the average fraction of inspired oxygen required to obtain target saturations 0.35. In clinical practice higher flow rates may be needed to achieve favourable outcomes.

In the author’s own clinical experience, devices are often set up at ward level in anticipation of pulmonary complications and to prevent escalation to a higher dependency unit. Patients may also be established on high flow devices at high dependency level in preparation for transition to ward-based care and to prevent readmission. High flow is also used in conjunction with CPAP in order to give the patient a break from the tighter mask. However, further research is needed to determine whether high-flow devices at ward level can prevent admissions to critical care.

Over this six-month period, anecdotally, patients preferred the comfort of HFOT over the CPAP mask, but a common complaint was that it was ‘too hot’. Unfortunately, subjective data on patient comfort whilst on HFOT was not included in the proforma. The author suggests that this was a limitation and inclusion of this information would be beneficial for future studies. Hasani et al. (2008) suggests that providing high-flow oxygen therapy at 37°C (optimum humidity) at flows of 20–25L/minute for a minimum of three hours a day significantly improves mucociliary clearance. Whilst this study was carried out in patients with bronchiectasis these findings may also apply for patients over this six-month period that had difficulty with retained secretions. Patients who were unable to tolerate the higher temperatures were reduced to 31°C as the author felt it more clinically beneficial to keep them on high-flow therapy at lower temperatures than not on at all.

It is apparent from both the available research and this clinical perspective that HFOT can be regarded as a viable treatment option for those patients who demonstrate signs of respiratory distress and who have pulmonary complications. This work has also shown that patients receiving HFOT via Airvo™ can be managed safely in a number of ward settings without the need to transfer to a high dependency unit for initiation of high flow treatment. However, the maximum amount of PEEP that can be generated with high-flow oxygen is approximately three cm H₂O. Those patients who require higher levels or need more accurate measurement of PEEP, demonstrate signs of cardiogenic pulmonary oedema or have stable chest wall trauma may still warrant conventional CPAP. HFOT should not be seen as a replacement but an alternative treatment option.

The study was carried out in a New Zealand hospital from October 2015 to April 2016, predominantly over summer months. However, apart from the month of October there was a steady use of high flow. The number of patients initiated on HFOT dipped in March with a sudden decline in April. It would be interesting to repeat or extend the study into the winter months to include seasonal variation. The author surmises that the same trends would be noticed but the demand would be greater. Thirteen patients were unable to use HFOT via Airvo™ due to lack of machines. This is interesting in itself as it demonstrates a high demand for high-flow and a clinical need for the device.

The author noted no adverse effects from the application of HFOT during the six-month period. It was safely set up and utilised by ward staff and the study has highlighted that it can be safely applied on the ward providing there is adequate staff training and regular assessment of competencies. This has implications on a service when deciding where the clinically unwell patient could be safely managed and cared for. This study provides interesting debate and food for thought as only 30 patients were cared for in a critical care setting. The remaining group (44 patients) were safely cared for on the ward (Figure 1).

Whilst the policy of a large number of hospitals is to provide CPAP under close observation for instance in a high dependency unit or coronary care facility, from this initiative it appears safe to administer HFOT via an Airvo™ in the ward setting which in turn may enable more flexible care planning and place of delivery for the acutely unwell respiratory compromised patient.

Limitations

It is acknowledged that this pilot study was carried out over a short period of time with a small sample size (n = 74). It was also carried out over the summer months with a decline in the use of HFOT in some months. Continuing into winter would potentially allow for more patients to be included. An increase in the use of such devices may be demonstrated particularly when there may be spikes in data for those patients diagnosed with pneumonias or exacerbations of chronic lung conditions in the colder months.

Data to compare alterations in work of breathing and respiratory rate before and during the use of HFOT was not collected due to poor completion of the proforma. Arterial blood gas analysis was not carried out routinely to obtain significant data on changes in PaO₂. The author recommends that this be included in future work. Subjective data regarding patient comfort whilst receiving high flow was also not recorded. This was a limitation of the proforma and is recommended for future work.

Recommendations

This pilot may enable clinicians still solely using CPAP via face, nasal mask or hood in this patient group to consider use of HFOT via Airvo™ as another treatment option in their area of practice. This may be of particular benefit for those patients who are unable to tolerate a tighter fitting mask/hood and therefore may become non-compliant with treatment. Over the six-month period, HFOT was delivered safely in the ward setting with no adverse effects. The author suggests the successful delivery of HFOT at ward level may have clinical implications with regards to intensive care admissions and discharges if these patients can be safely managed on the ward.

For future work, it would be beneficial to collect some qualitative data on patient comfort and compliance with the machine as well as some information on how confident clinicians feel in setting up and weaning patients from the device.

Conclusion

Over a six-month period in a small tertiary hospital in New Zealand, the availability of HFOT devices was increased. Within this timeframe, 74 patients used HFOT. Results indicate that for those patients who have secretion retention, atelectasis, increased work of breathing and increasing oxygen requirements regardless of initial diagnosis, HFOT via the Airvo™ is a viable treatment option. It can be particularly beneficial for those patients who are unable to tolerate a tighter fitting mask and who do not depend on high levels of accurately measured PEEP. With suitable staff training for HFOT in place, this work has shown that it can be used safely in the ward setting and has utilisation across the hospital on a variety of wards in the acute setting.

Key points

• Airvo™ HFOT is a viable treatment option as part of ward-based care which has positive clinical implications in terms of managing the patient at ward level.

• HFOT has been found to be clinically beneficial to patients demonstrating signs of pulmonary complications such as atelectasis, sputum retention, increased work of breathing and increased oxygen requirements.

• HFOT may be more easily tolerated than a tighter fitting mask and should be viewed as an alternative to CPAP not a replacement.

Acknowledgements

The author wishes to acknowledge physiotherapy colleagues at Dunedin Hospital New Zealand for assisting with the review, to Anna Higgs for her contribution and to Ange Price for her support, advice and experience. Acknowledgements are also made to Fisher and Paykel for providing the equipment needed to carry out the six-month review, with special mention to Tina Saltmarsh.

Disclosure statement

No potential competing interest was reported by the author.

Corley, A., Caruana, L. R., Barnett, A. G., Tronstad, O., & Fraser, J. F. (2011). Oxygen delivery through high-flow nasal cannulae increase end-expiratory lung volume and reduce respiratory rate in post-cardiac surgical patients. British Journal of Anaesthesia, 107(6), 998–1004. https://doi.org/10.1093/bja/aer265.

Bräunlich, J., Beyer, D., Mai, D., Hammerschmidt, S., Seyfarth, H. J., & Wirtz, H. (2013). Effects of nasal high flow on ventilation in volunteers, COPD and idiopathic pulmonary fibrosis patients. Respiration, 85(4), 319–325. https://doi.org/10.1159/000342027.

Frat, J. P., Thille, A. W., Mercat, A., Girault, C., Ragot, S., Perbet, S., Prat, G., Boulain, T., Morawiec, E., Cottereau, A., Devaquet, J., Nseir, S., Razazi, K., Mira, J. P., Argaud, L., Chakarian, J. C., Ricard, J. D., Wittebole, X., Chevalier, S., Herbland, A., REVA Network (2015). High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure. The New England Journal of Medicine, 372(23), 2185–2196. https://doi.org/10.1056/NEJMoa1503326.

Groves, N., & Tobin, A. (2007). High flow nasal oxygen generates positive airway pressure in adult volunteers. Australian Critical Care, 20(4), 126–131. https://doi.org/10.1016/j.aucc.2007.08.001.

Hasani, A., Chapman, T. H., McCool, D., Smith, R. E., Dilworth, J. P., & Agnew, J. E. (2008). Domiciliary humidification improves lung mucociliary clearance in patients with bronchiectasis. Chronic Respiratory Disease, 5(2), 81–86. https://doi.org/10.1177/1479972307087190.

Lenglet, H., Sztrymf, B., Leroy, C., Brun, P., Dreyfuss, D., & Ricard, J. D. (2012). Humidified high flow nasal oxygen during respiratory failure in the emergency department: Feasibility and efficacy. Respiratory Care, 57(11), 1873–1878. https://doi.org/10.4187/respcare.01575.

Maggiore, S. M., Idone, F. A., Vaschetto, R., Festa, R., Cataldo, A., Antonicelli, F., Montini, L., De Gaetano, A., Navalesi, P., & Antonelli, M. (2014). Nasal high-flow versus Venturi mask oxygen therapy after extubation. Effects on oxygenation, comfort, and clinical outcome. American Journal of Respiratory and Critical Care Medicine, 190(3), 282–288. https://doi.org/10.1164/rccm.201402-0364OC.

Nishimura M. (2015). High-flow nasal cannula oxygen therapy in adults. Journal of Intensive Care, 3(1), 15. https://doi.org/10.1186/s40560-015-0084-5.

Parke, R., McGuinness, S., & Eccleston, M. (2009). Nasal high-flow therapy delivers low level positive airway pressure. British Journal of Anaesthesia, 103(6), 886–890. https://doi.org/10.1093/bja/aep280.

Parke, R. L., Eccleston, M. L., & McGuinness, S. P. (2011). The effects of flow on airway pressure during nasal high-flow oxygen therapy. Respiratory Care, 56(8), 1151–1155. https://doi.org/10.4187/respcare.01106.

Parke, R. L., McGuinness, S. P., & Eccleston, M. L. (2011). A preliminary randomised controlled trial to assess effectiveness of nasal high-flow oxygen in intensive care patients. Respiratory Care, 56(3), 265–270. https://doi.org/10.4187/respcare.00801.

Peters, S. G., Holets, S. R., & Gay, P. C. (2013). High-flow nasal cannula therapy in do-not-intubate patients with hypoxemic respiratory distress. Respiratory Care, 58(4), 597–600. https://doi.org/10.4187/respcare.01887.

Rea, H., McAuley, S., Jayaram, L., Garrett, J., Hockey, H., Storey, L., O'Donnell, G., Haru, L., Payton, M., & O'Donnell, K. (2010). The clinical utility of long-term humidification therapy in chronic airway disease. Respiratory Medicine, 104(4), 525–533. https://doi.org/10.1016/j.rmed.2009.12.016.

Roca, O., Riera, J., Torres, F., & Masclans, J. R. (2010). High-flow oxygen therapy in acute respiratory failure. Respiratory Care, 55(4), 408–413.

Stéphan, F., Barrucand, B., Petit, P., Rézaiguia-Delclaux, S., Médard, A., Delannoy, B., Cosserant, B., Flicoteaux, G., Imbert, A., Pilorge, C., Bérard, L., & BiPOP Study Group (2015). High-flow nasal oxygen vs noninvasive positive airway pressure in hypoxemic patients after cardiothoracic surgery: A randomised clinical trial. JAMA, 313(23), 2331–2339. https://doi.org/10.1001/jama.2015.5213.

Sztrymf, B., Messika, J., Bertrand, F., Hurel, D., Leon, R., Dreyfuss, D., & Ricard, J. D. (2011). Beneficial effects of humidified high flow nasal oxygen in critical care patients: A prospective pilot study. Intensive Care Medicine, 37(11), 1780–1786. https://doi.org/10.1007/s00134-011-2354-6.

Ward J. J. (2013). High-flow oxygen administration by nasal cannula for adult and perinatal patients. Respiratory Care, 58(1), 98–122. https://doi.org/10.4187/respcare.01941.