Article Text

Download PDFPDF

Change in pulmonary mechanics and the effect on breathing pattern of high flow oxygen therapy in stable hypercapnic COPD
  1. Lara Pisani1,
  2. Luca Fasano2,
  3. Nadia Corcione1,
  4. Vittoria Comellini1,
  5. Muriel Assunta Musti3,
  6. Maria Brandao4,
  7. Damiano Bottone5,
  8. Edoardo Calderini6,
  9. Paolo Navalesi7,8,
  10. Stefano Nava1
  1. 1Department of Clinical, Integrated and Experimental Medicine (DIMES), Respiratory and Critical Care Unit, S. Orsola-Malpighi Hospital, Alma Mater University, Bologna, Italy
  2. 2Sant'Orsola Malpighi Hospital, Bologna, Italy
  3. 3Department of Public Health, Local Health Authority of Bologna, Bologna, Italy
  4. 4Respiratory Department, Centro Hospitar de Tràs-os-Montes e Alto Douro, São Pedro de Vila Rea, Portugal
  5. 5Respiratory Medicine Unit, Department of Clinical and Experimental Sciences, Universita’ degli Studi di Brescia, Brescia, Italy
  6. 6Department of Anesthesia and Intensive Care, Fondazione IRCCS Cà Granda, Ospedale Maggiore Policlinico, Milano, Italy
  7. 7Department of Translational Medicine, Università del Piemonte Orientale ‘Amedeo Avogadro’, Novara, Italy
  8. 8Department of Anesthesia and Intensive Care Medicine, Sant'Andrea Hospital, Vercelli, Italy
  1. Correspondence to Professor Stefano Nava, Respiratory and Critical Care Unit, Sant'Orsola Malpighi Hospital, Via Massarenti 9, Bologna 40138, Italy; stefano.nava{at}


We studied the effects of high flow oxygen therapy (HFOT) versus non-invasive ventilation (NIV) on inspiratory effort, as assessed by measuring transdiaphragmatic pressure, breathing pattern and gas exchange. Fourteen patients with hypercapnic COPD underwent five 30-min trials: HFOT at two flow rates, both with open and closed mouth, and NIV, applied in random order. After each trial standard oxygen therapy was reinstituted for 10 min. Compared with baseline, HFOT and NIV significantly improved breathing pattern, although to different extents, and reduced inspiratory effort; however, arterial carbon dioxide oxygen tension decreased but not significantly. These results indicate a possible role for HFOT in the long-term management of patients with stable hypercapnic COPD.

Trial registration number NCT02363920.

  • Respiratory Measurement
  • Long Term Oxygen Therapy (LTOT)
  • Non invasive ventilation
  • COPD Pathology

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


High flow oxygen therapy (HFOT) is a new technique for delivering oxygen. In patients with acute respiratory failure, compared with standard oxygen therapy, HFOT was repeatedly shown to improve comfort, avoid mucosal dryness and injury, and deliver a more reliable and stable fraction of inspired oxygen (FiO2).1 In a study performed in patients with stable COPD, HFOT reduced the arterial carbon dioxide oxygen tension (PaCO2), increased end-expiratory and tidal volumes (VTs), and decreased respiratory rate.2 This makes HFOT an appealing form of treatment in patients with stable chronic hypercapnic respiratory failure (CHRF). In this setting, HFOT could be used in place of non-invasive ventilation (NIV) in the least tolerant and compliant patients, or in association with NIV, to reduce mask-related side effects.

In this randomised short-term physiological investigation we compare the physiological effects of standard oxygen therapy, NIV and HFOT in patients with stable hypercapnic COPD, as assessed at two flow rates with open and closed mouth. In addition to breathing pattern and arterial blood gases (ABGs), we measured the inspiratory effort by measuring the transdiaphragmatic pressure (Pdi).

Materials and methods

We enrolled 14 consecutive patients with COPD and CHRF. Patient characteristics and inclusion criteria are shown in the online supplementary table 1R and figure 1R. The study was approved by the Institutional Ethical Committee and written informed consent was signed by each patient.

Data were recorded during five 30-min trials applied according to a predetermined computer-generated random sequence. After each trial standard oxygen therapy through a nasal cannula was reinstituted for 10 min.

NIV and HFOT (Airvo 2; Fisher & Paykel Healthcare, Auckland, New Zealand) settings during the experimental procedure are illustrated in the online supplement. FiO2 was kept constant.

ABGs were obtained at baseline and at the end of each trial. During baseline and HFOT trials VT was obtained by integrating the flow signal. Flow and VT could not be determined during the open mouth HFOT trials. Inspiratory and expiratory breath durations were determined from Pdi tracing, as previously described.3 The patient's own respiratory rate (RRp) was also determined from the Pdi.3 The last 5 min of each trial was recorded and averaged for data analysis.

In the online repository we provide detailed information on the inclusion and exclusion criteria, study procedures and statistical analysis.


As shown in table 1, compared with baseline, breathing frequency was significantly reduced in HFOT trials with the mouth closed and with NIV. Patient's own expiratory time (TE,p) was significantly prolonged and VT higher compared with baseline for all the settings. The patient's own inspiratory time (TI,p) was no different between trials.

Table 1

Breathing pattern, inspiratory effort and lung mechanics in different settings

Pdi swing and diaphragm pressure time product (PTPdi) were reduced compared with baseline in all trials. However, the reductions observed during NIV were significantly larger, as opposed to all of the HFOT trials. Dynamic intrinsic positive end expiratory pressure (PEEPi, dyn) was significantly reduced compared with baseline in all trials.

Breathing frequency, TI,p and TE,p, did not change between the different HFOT trials with the mouth closed or open, while Pdi at HFOT 20 L/min, was statistically higher with the mouth closed compared with open.

As shown in table 2, the PaCO2 level decreased but not significantly with HFOT at 30 L/min and NIV compared with standard oxygen.

Table 2

Arterial blood gas values and comfort scores at different settings

Also shown in table 2, comfort did not vary among the different trials.


We have demonstrated that compared with low oxygen flow, HFOT and NIV both reduce the respiratory muscle load on the respiratory system, resulting in a change in breathing pattern, increasing Te and reducing respiratory rate, suggesting a change in the pressure–volume curve.

Therefore, HFOT is an appealing technique as a potential alternative to NIV because it is less of a burden, it provides a more physiological humidification and heating of the airways, and a more ‘easy to fit’ interface.

Two studies in normal4 controls and patients with COPD2 have also shown that HFOT led to a marked increase in VT that was offset by a reduction in respiratory rate.

The most likely explanation for this response seems to be related to the increase in the expiratory resistance, with a mechanism different from that of CPAP.4

The respiratory pattern elicited by HFOT resembles pursed lip breathing which is, however, associated with increased work of breathing and patients cannot maintain this pattern over a longer time period.5 In contrast, during HFOT we could demonstrate for the first time that inspiratory effort was reduced.

Several mechanisms have been advocated to explain the effect of HFOT on work of breathing, such as minimisation of inspiratory resistance,4 attenuation of the activation of cold receptors or osmoreceptors in the nasal mucosa6 and reducing the anatomical dead space in the upper airways.7 Indeed the prolonged expiratory time may also reduce the amount of PEEPi, which may increase the inspiratory load.8

Furthermore, HFOT generates a modest degree of positive pressure, unlikely to be above 5–6 cmH2O,9 which may also partially counteract the threshold load imposed by the presence of PEEPi.10

The reduction in transcutaneous CO2 in Fraser's study2 and in PaCO2, despite not being statistically significant, in our investigation, support the hypothesis that it is possible to reduce hypercapnia using HFOT. Indeed, carbon dioxide directly controls the activity of inspiratory muscles alone and therefore its reduction may lead to a decrease in diaphragmatic effort. We cannot rule out the effect of a higher PaO2/FiO2 ratio as explanation for the PaCO2 increase during baseline conditions. Moreover baseline conditions consisted of breathing oxygen through nasal cannula, and under these conditions, FiO2 cannot be controlled depending on the breathing pattern, the patient's inspiratory flow and whether patients breathe predominantly through the mouth or the nose. Therefore, the decrease in PaO2 during HFOT can be explained by a higher actual FiO2 under nasal oxygen therapy compared with HFOT.

We have further explored the physiological changes induced by mouth or nasal breathing, since it is totally unrealistic to assume that patients recruited for long-term treatment will always breathe with their mouth perfectly ‘sealed’. It has been shown9 that breathing with the mouth open negatively influences the generation of a positive pressure. Despite this, we were unable to demonstrate any ‘detrimental’ effect of this behaviour on the breathing pattern and inspiratory effort compared with breathing with the mouth closed.

The results of this study show overall similar acute physiological changes between HFOT and NIV, and support the need for further investigations to assess the effectiveness of domiciliary HFOT versus NIV in patients with stable hypercapnia. Obviously our findings could not be translated to the situation of an acute exacerbation of COPD.



  • Twitter Follow Nadia Corcione @nadia_corcione

  • Contributors LP, study design, experimental procedure, writing and data analysis. LF, study design, data analysis. NC, data analysis and experimental procedure. VC, experimental procedure. MAM, statistical analysis. MB, experimental procedure. DB, experimental procedure. EC, data analysis and writing. PN, study design and writing. SN, study design, experimental procedure, data analysis and writing.

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval Comitato Etico Indipendente dell'Azienda Ospedaliero-Universitaria di Bologna, Policlinico S. Orsola-Malpighi.

  • Provenance and peer review Not commissioned; externally peer reviewed.