Discomfort of noninvasive ventilation, high flow nasal cannula oxygen, and conventional oxygen implemented after extubation in non-COPD patients

Abstract

Background

Discomfort of noninvasive respiratory supports remain controversial. In the present study, the discomfort of high flow nasal cannula oxygen, noninvasive ventilation, and conventional oxygen therapy were evaluated in the post-extubation period.

Methods

Noninvasive ventilation alternating with high flow nasal cannula oxygen or conventional oxygen therapy were implemented for 48 h following extubation in patients free of chronic obstructive pulmonary disease. Using a 10-cm visual analog scale, the discomfort of the interface (oxygen mask, nasal prongs or facial mask), was self-evaluated by 264 patients after extubation whereas the discomfort of the respiratory support (conventional oxygen therapy or high flow nasal cannula oxygen alternating with noninvasive ventilation) was evaluated by the nurses in 306 patients. Evaluations were performed in patients at high (n = 127) and low (n = 179) risk of postextubation respiratory failure, 6, 24 and 48 h after extubation.

Results

Facial mask was the source of a significant and persisting self-evaluated discomfort. Nasal prongs and oxygen mask were less uncomfortable, with a progressive reduction of the discomfort with time. High flow nasal cannula oxygen alternating with noninvasive ventilation caused a significant and persisting nurse-reported discomfort. Conventional oxygen therapy was significantly less uncomfortable, and the degree of discomfort was not different in patients at high and low risk of extubation failure.

Conclusions

Noninvasive ventilation was associated with a significant and persisting discomfort whereas high flow nasal cannula oxygen-induced discomfort remained limited. After 48 h, discomfort resulting from high flow nasal cannula oxygen and conventional oxygen therapy was similar and negligible.

Background

High Flow Nasal Cannula Oxygen (HFNO) is generally considered as more comfortable than Non Invasive Ventilation (NIV) or conventional oxygen therapy in patients with or without respiratory distress. However, two systematic reviews concluded that HFNO was not more comfortable than NIV in patients with acute or post-extubation respiratory failure [1, 2]. In 2020 a randomised controlled trial performed in patients with chronic obstructive pulmonary disease, reported that HFNO implemented after extubation, was more comfortable than NIV, both techniques being equally efficient for preventing reintubation [3]. In the present study, the discomfort of interfaces used for HFNO, NIV and conventional oxygen therapy was self-evaluated after extubation by patients free of chronic obstructive pulmonary disease. During the same period of time, the discomfort of both respiratory supports implemented after extubation, was evaluated by the nurse: HFNO alternating with NIV and conventional oxygen therapy.

Methods

We analysed the discomfort of HFNO, NIV and conventional oxygen therapy assessed during the WIN-IN-WEAN French Chinese RCT [4] registered at chictr.org (ChiCTR1800018530). It is a secondary analysis of the WIN-IN-WEAN trial whose primary endpoint was to compare the efficiency of two noninvasive respiratory supports for preventing postextubation respiratory failure: conventional oxygen therapy versus 1 h-NIV alternating with 2 h-HFNO, implemented in high-risk patients free of chronic obstructive pulmonary disease immediately after extubation. The risk of postextubation respiratory failure was assessed by measuring lung ultrasound score (LUS) at the end of the successful spontaneous breathing trial: patients with LUS ≥ 14 were considered at high risk whereas patients with LUS < 14 were considered at low risk [5]. LUS was measured by experienced investigators who attented a specific training curriculum [6, 7]. Before extubation, patients were randomly assigned to the LUS intervention group or to the control group. In the LUS intervention group, the LUS was communicated to the attending physician. Patients with LUS ≥ 14 received prophylactic 2 h HFNO alternating with 1 h NIV during two days whereas patients with LUS < 14 received conventional oxygen. In the control group, the LUS was not communicated to the attending physician and patients received conventional oxygen. In this group some patients had a LUS ≥ 14 and others a LUS < 14.

HFNO (50–60 L.min⁻¹) was delivered using binasal symetrical silicon prongs connected to an Optiflow device (Fisher & Paykel Healthcare®, Auckland New Zealand) providing 44 mg H₂O/L at a temperature of 37 °C. NIV was delivered through an oro-nasal mask, tightly applied to the face to avoid leaks (Fig. 1a), and connected to a ventilator equipped with a MR850 humidifier (Fisher & Paykel Healthcare®, Auckland New Zealand) providing 44 mg H₂O/L at a temperature of 37 °C. NIV delivered a pressure support ranging between 5 and 10 cmH₂O to obtain a tidal volume of 7–10 ml kg⁻¹ of predicted body weight. Fraction of inspired oxygen and positive end expiratory were selected to maintain a peripheral capillary oxygen saturation ≥ 92%. Nasal mask, total face mask and helmet were never used. Conventional oxygen therapy was delivered by a non-rebreather face mask with a reservoir bag, providing non-humidified 10–15 L.min⁻¹ oxygen flow rates at ambient temperature.

Fig. 1.

Evaluation of the patient’s discomfort in the WIN-IN-WEAN study using a visual analog scale (a) Self-evaluation of the discomfort of the different interfaces (non-rebreather facial mask for conventional oxygen therapy, nasal silicon prongs for high flow nasal cannula oxygen, and oro-nasal mask for noninvasive ventilation) performed 6, 24 and 48 h after extubation. (b) Evaluation by the nurse of the discomfort of the respiratory support: either conventional oxygen (a 10–15 L.min−1 oxygen flow) or high flow nasal cannula oxygen (a 50–60 L.min−1 oxygen flow delivered using an Optiflow® device) alternating with noninvasive ventilation (pressure support ventilation with PEEP delivered through a face mask connected to a ventilator equipped with a conventional humidifier). High flow nasal cannula oxygen was delivered for 2-h periods and noninvasive ventilation for 1-h periods. Discomfort evaluation was performed 6, 24 and 48 h after extubation using the visual analog scale (right lower corners). Discomfort of conventional oxygen was evaluated in patients at high risk of post extubation respiratory failure (dark blue: lung ultrasound score ≥ 14 measured at the end of the spontaneous breathing trial before extubation), and in patients at low risk of post extubation respiratory failure (light blue: lung ultrasound score < 14 measured at the end of the spontaneous breathing trial). Box plots are used to display data: median values, interquartile 25–75, confidence intervals and extreme values. LUS = Lung Ultrasound Score; NIV = Non Invasive Ventilation; HFNO = High Flow Nasal Cannula Oxygen; p values between groups are represented

The discomfort of the interface (oxygen mask, Optiflow or NIV mask), was self-evaluated by 264 patients receiving conventional oxygen therapy after extubation and 42 patients treated by NIV alternating with HFNO. All patients were fully aware and considered able to assess their discomfort as inclusion criteria in the WIN IN WEAN study were: a Glasgow Coma Score ≥ 13, a good synchronisation with the ventilator, a calm behavior (equivalent to a Richmond Agitation Sedation Score of 0 or -1) with an interruption of any sedation for at least 24 h. Evaluations were performed by patients at high and low risk of postextubation respiratory failure 6, 24 and 48 h after extubation, using a 10-cm visual analog scale. Patients pointed the discomfort score on a white board displaying options shown in the lower part of Fig. 1a and 1b. The questions asked where: 1—« How confortable is your face mask ?» for patients receiving noninvasive ventilation or conventional oxygen 2—« How confortable are your nasal prongs delivering high flow oxygen ?» for patients receiving high flow nasal cannula oxygen. Discomfort evaluation was performed at the end of the one-hour NIV period and at the end of the two-hour HFNO period.

The discomfort of the respiratory support (conventional oxygen therapy or HFNO alternating with NIV) was evaluated by the nurses in 306 patients. NIV alternating with HFNO was evaluated at the end of the three-hour period. Evaluations were performed in patients at high (n = 127) and low risk (n = 179) of postextubation respiratory failure, 6, 24 and 48 h after extubation, using a 10-cm visual analog scale (Fig. 1b). As nurses were blinded to the allocation group, they were not aware of patients with a LUS ≥ 14 receiving conventional oxygen therapy. All patients gave written informed consent. Patients who had to be reintubated for laryngospasm or emergency surgery were excluded from the analysis.

Statistical analysis

Qualitative data are expressed as percentage (%). Quantitative data are expressed as mean ± SD or median values and interquartile [p25-p75] according to data distribution. Clinical characteristics of patients with LUS ≥ 14 and < 14 were compared using an unadjusted chi-square test or Fisher’s exact test as appropriate. Self-evaluated discomfort and discomfort evaluated by the nurse were compared between conventional oxygen therapy, HFNO and NIV using a Kruskal–Wallis test followed by Mann–Whitney test 6, 24 and 48 h after extubation. No compensation was done for missing data. A two-sided p value ≤ 0.05 was considered to indicate statistical significance (Version 15.0 StataCorp).µ.

Results

Discomfort of conventional oxygen therapy interface was self-evaluated by 179 patients at low risk and 85 patients at high-risk of extubation failure. Discomfort of NIV and HFNO interfaces were self-evaluated by 42 patients at high risk of postextubation respiratory failure. Discomfort of the respiratory support HFNO alternating with NIV was evaluated by the nurses in 86 patients at high risk of extubation failure. Discomfort of conventional oxygen therapy was evaluated by the nurses in 41 patients at high risk and 179 patients at low risk of extubation failure. Patients’ clinical characteristics are shown in Table 1. Patients at high risk of extubation failure were older, had a higher SOFA score, a lower left ventricular ejection fraction, a greater weight gain and a higher incidence of pleural effusion.

Table 1.

Clinical characteristics of the patients who performed the self-evaluation. LUS = Lung Ultrasound Score; COT = Conventional Oxygen Therapy; HFNO = High Flow Nasal Cannula Oxygen; NIV = NonInvasive Ventilation

NIV was the source of a significant and persisting self-reported discomfort. Conventional oxygen therapy and HFNO were less uncomfortable, with a progressive reduction of the discomfort with time (Fig. 1a). HFNO alternating with NIV caused a significant and persisting nurse-reported discomfort. Conventional oxygen therapy was significantly less uncomfortable, and the degree of nurse-reported discomfort was not different in patients at high and low risk of extubation failure. Conventional oxygen therapy was the least uncomfortable respiratory support whatever the risk of postextubation respiratory failure.

Discussion

This study provides evidence that HFNO and conventional oxygen therapy are less uncomfortable than NIV, when implemented after extubation in patients free of chronic obstructive pulmonary disease. The NIV discomfort is similar whatever the risk of postextubation respiratory failure.

NIV has long been considered a source of discomfort, possibly participating to extubation failure [8]. The tight-fitting face mask causes a sensation of claustrophobia, creates painful pressure on the nose and often produces nasofacial skin breakdown. To reduce discomfort, patients often remove their masks, which decreases the effectiveness of respiratory support. Drinking, eating, sputum clearance and communication also require NIV interruption.

Such factors of discomfort are not observed with HFNO. However, when oxygen flows higher than 20 L.min⁻¹ are used to optimise the respiratory assistance, patients may complain from innapropriate humidification drying out the nasal mucosa or fom excessive humidity and too high temperature at the level of the nostrils [9]. To limit the discomfort, HFNO should have a minimal hygrometry of 30 mgH₂0/ L [10]. Increasing the humidity above this threshold may either increase or decrease the discomfort. The feeling of humidification is subjective and patient-dependant [11] and the degree of humidification should be personalised for a given flow. A too high temperature of the high nasal flow is another source of discomfort in critically ill patients with acute respiratory failure [12, 13]: a temperature of 37 °C is often associated with a higher discomfort than a temperature of 31 °C whatever the flow rate, suggesting that unbalanced water retention and excessive heating of the nostrils might prevail over the advantage of maximum humidity. In healthy volunters, discomfort is mainly related to flow rates and nasal cannula diameter rather than to the temperature of the flow [14, 15]. Asymmetrical prongs could reduce the discomfort [16].

In the present study, overall HFNO was the source of limited discomfort. On day 2, discomfort resulting from HFNO and conventional oxygen therapy were similar and negligible, confirming a previous study performed in critically ill patients with pneumonia [17]. Conventional oxygen therapy was the less uncomfortable respiratory support. The risk category did not influence nurse-reported discomfort, suggesting that the underlying respiratory condition did not play a major role in the feeling of discomfort.

The study has limitations and strengths. It is a post hoc and observational analysis, derived from a randomised controlled trial not primarily designed to compare comfort between HFNO, NIV, and conventional oxygen therapy. Therefore, due to their exploratory nature, the conclusions are hypothesis-generating rather than definitive. The use of a visual analog scale may be challenging for recently extubated patients and could be subject to significant inter-individual variability. Comfort of the interface was self-evaluated by 222 patients for conventional oxygen therapy and 42 patients for NIV and HFNO, an imbalance that could reduce statistical power and the robustness of between-group comparisons. However, the prospective acquisition of data, the large number of evaluations, and the double assessment of discomfort by patients and nurses over 48 h, strengthen the validity of the main result of the study: in the post extubation period, NIV was associated with a significant discomfort whereas HFNO and conventional oxygen therapy were well tolerated in patients free of chronic obstructive pulmonary disease. How differences in comfort impact the therapeutic efficiency of non invasive respiratory support remains to be determined.

Author contributions

JJR conceived and designed the study, prepared the manuscript, and had final responsibility for the decision to submit for publication. The Biostatistics Unit of the Department of Clinical Research and Innovation (CHU Clermont-Ferrand, Clermont-Ferrand, France) supervised data collection by French and Chinese centres and performed data analysis. JJR, DM, and BP take responsibility for the integrity of the data and the accuracy of the data analysis. Data analysis was performed by JJR, DM and BP during a research meeting held in Les Tines (74,400, France) in 2024. Statistical analysis was performed by BP. JPQ, ZM, SB, JL, MA, JMC and XJ played a major role in the inclusion of patients, red, reviewed and approved the manuscript.

Funding

This work was supported by the “Programme Hospitalier de Recherche Clinique Regional 2014” of the French Ministry of Health (PHRCI-14–003).

Data availability

The datasets used and analysed for the current study are available from Bruno Pereira on reasonable request.

Declarations

Human ethics and consent to participate

The institutional review board of the coordinator center (Department of Peri-Operative Medicine, University Hospital Estaing, University of Auvergne, Clermont-Ferrand, France) reviewed and approved the study protocol before trial initiation (ID ChiCTR1800018530). All enrolled patients provided written informed consent in accordance with committee recommendations.

Competing interests

J-M. Constantin reports personal fees and non-financial support from Drager GE Healthcare, Sedana Medical, Baxter, and Amomed; personal fees from Fisher and Paykel Healthcare, Orion, Philips Medical, and Fresenius Medical Care; and non-financial support from LFB and Bird Corporation, outside of the submitted work. All other authors declare that they have no competing interests.