ABSTRACT
Background
Patients with exertional hypoxaemia (EH) at hospital discharge face rapid deconditioning risks. However, early pulmonary rehabilitation (PR) is limited by exercise intolerance and access barriers. Optimal PR design for this group remains unclear. High‐flow nasal oxygen (HFO) may alleviate EH, but its role in home‐based PR is unexplored. We aimed to evaluate an early, intensive, home‐based PR program and the integration of HFO during exercise training for EH patients post‐hospitalisation.
Methods
In this pilot feasibility trial, EH patients across respiratory conditions were randomised to PR with HFO (HFO‐PR) or PR alone (PR‐only). Both groups commenced a three‐week home‐based PR program (two supervised, three self‐directed sessions weekly) within 1 week of discharge. Feasibility was assessed using quantitative and qualitative measures.
Results
Fifty‐two of 56 patients completed follow‐up at 4 weeks. Overall PR program acceptance and uptake were 67% and 62%, respectively. Program completion was lower in the HFO‐PR group (60%) compared to the PR‐only group (89%), primarily due to reduced adherence to self‐directed exercise sessions. Among COPD patients, the 1‐min sit‐to‐stand test at 4 weeks improved by 4.6 repetitions in the HFO group and 5.3 in the PR‐only group (adjusted difference in mean change −0.8, 95% CI −4.7 to 3.1). Qualitative findings revealed three themes: Transformations following the intervention, Receiving care and education at home, and Experiencing HFO system. Participants reported challenges with HFO during unsupervised exercise.
Conclusion
Early home‐based PR for post‐hospitalisation EH patients is feasible and safe. However, the addition of HFO was associated with lower adherence, highlighting implementation barriers to address before further evaluation.
Trial Registration
Keywords: exertional hypoxaemia, high flow nasal oxygen, pulmonary rehabilitation
This pilot randomised feasibility trial evaluated the feasibility and acceptability of early, intensive home‐based pulmonary rehabilitation (PR) and high‐flow nasal oxygen (HFO) during exercise in patients with exertional hypoxaemia post‐hospitalisation. Early home‐based PR was feasible and safe. However, further HFO device optimisation is required to overcome implementation barriers identified before broader evaluation.
1. Introduction
Pulmonary rehabilitation (PR) is a well‐established intervention that improves exercise capacity [1], alleviates symptoms, and enhances overall quality of life across a range of respiratory conditions [2]. However, despite its demonstrated efficacy, adherence to PR programs remains poor [3, 4, 5]. Among the barriers identified, poor exercise capacity and dyspnoea are commonly reported deterrents to exercise training [6, 7], a core component of PR.
Exertional hypoxaemia (EH), which affects a substantial proportion of patients referred for PR [8, 9], may contribute to exercise limitation during training. Commonly defined as a ≥ 4% decline in oxygen saturation during physical exertion [10], EH has been associated with reduced exercise capacity and greater physiological impairment in patients with chronic respiratory disease [11]. Although a direct association between EH and PR adherence has not yet been conclusively established, lower exercise capacity has been linked to an increased likelihood of PR discontinuation [12].
Patients with EH following hospital discharge may also represent a higher‐risk population for future respiratory decompensation as EH has been associated with increased exacerbation risk in COPD [11]. These patients may therefore particularly benefit from rehabilitation interventions. However, delivering PR to this group presents unique challenges and the optimal design of PR programs tailored for patients with EH remains uncertain. Addressing this gap requires consideration of factors such as timing of PR initiation, delivery setting and strategies to mitigate exercise limitations during training.
Early initiation of PR following hospitalisation may be especially important for patients with EH as it may counteract the rapid functional decline seen in the first month post‐discharge [13]. However, exercise intolerance and practical barriers can limit early participation. Travel to outpatient PR programs may be challenging due to reduced functional capacity and logistical constraints, while inpatient programs are resource‐intensive and limited in availability. Home‐based PR may therefore represent a more feasible alternative.
A key challenge remains the physiological limitation imposed by EH during exercise. Strategies such as supplemental oxygen [14] or non‐invasive ventilation [15] have shown mixed results in improving exercise performance. Physiological studies suggest that high flow nasal oxygen (HFO) may reduce airway resistance [16, 17], provide more consistent oxygen delivery, improve gas exchange via carbon dioxide washout, and normalise breathing patterns [18]. By improving respiratory mechanics, reducing dynamic hyperinflation, improving oxygenation, and lowering the work of breathing, HFO may enhance exercise tolerance and training capacity, with the potential to improve PR adherence and outcomes. However, its role in a home‐based PR setting has not been studied.
To holistically address these challenges, we propose a “fast in, fast out” early intervention strategy: an intensive 3‐week, home‐based, semi‐supervised PR program using HFO initiated within 1 week of hospital discharge. Our research question was whether this early, home‐based PR model incorporating HFO is feasible, acceptable, and safe for patients with EH, and whether it confers early signals of improved functional recovery compared with PR only. We postulate that integrating HFO into early home‐based PR may improve post‐hospitalisation outcomes in EH patients, a hypothesis to be formally tested in a future definitive trial. Therefore, the aim of this study is to evaluate an early, intensive, home‐based PR program and the integration of HFO during exercise training for EH patients post‐hospitalisation.
2. Methods
2.1. Study Design, Subject Recruitment and Randomisation
We conducted a parallel‐group randomised feasibility pilot trial at Changi General Hospital, a 1000‐bed teaching hospital in Singapore between March 2023 and February 2025. Patients hospitalised for: (a) COPD exacerbation (b) COVID‐19 pneumonia (c) bronchiectasis exacerbation (d) interstitial lung disease (ILD) exacerbation were recruited prior to discharge. Eligible patients were identified through the inpatient physiotherapy service and written informed consent was obtained.
Eligibility criteria included the presence of EH, defined by ≥ 4% drop in oxygen saturation (SpO2) during the 1‐min sit‐to‐stand test (1mSTST), age ≥ 21 years and above, a diagnosis of acute COPD exacerbation requiring additional therapy such as the use of short acting bronchodilators and/or treatment course of corticosteroids OR COVID‐19 pneumonia OR bronchiectasis exacerbation OR ILD exacerbation, and fitness for exercise as determined by physician and physiotherapist. Fitness for exercise was determined by routine clinical assessment and standard PR safety considerations such as haemodynamic stability, absence of uncontrolled cardiac or medical conditions, ability to follow instructions, and capacity to safely perform low to moderate intensity functional exercise. Exclusion criteria included uncontrolled severe medical conditions or physical conditions precluding exercise participation.
Patients were randomised 1:1 to either PR with HFO (HFO‐PR) or PR without HFO (PR‐only). Stratification was applied for two groups: Group 1 (COPD, primary focus) and Group 2 (exploratory: COVID‐19, bronchiectasis, or ILD). The exploratory group was initially limited to patients with COVID‐19 pneumonia but was expanded early in the study to include bronchiectasis or ILD as COVID‐19 hospitalisations declined. Separate randomisation sequences with random permuted blocks of 2, 4, or 6 were generated for each stratum. Allocation concealment was ensured by sequentially numbered, opaque, sealed envelopes opened by the researcher immediately before assignment. Participants commenced the home‐based PR program within 1 week of hospital discharge.
Due to the nature of the interventions, neither patients nor physiotherapists delivering the home PR program could be blinded. As a small team conducted the program, the 1mSTST assessor may have been aware of group allocation; however, the test's standardised, objective nature reduced the impact of any measurement bias. Data analysts remained blinded to group allocation. The study is reported in accordance with the CONSORT extension for pilot and feasibility trials [19]. The study was approved by the SingHealth Centralised Institutional Review Board Ref: 2022/2605 (21 November 2022) and underwent trial registration (ClinicalTrials. gov: NCT05752370).
2.2. Study Interventions
2.2.1. Early Home‐Based, 3‐Week Hybrid Pulmonary Rehabilitation Program
All patients began a home PR program within 1 week of hospital discharge. Following the American Thoracic Society/European Respiratory Society guidelines [20], the program incorporated breathing exercises, aerobic interval training, resistance training, flexibility training, and patient education.
2.2.2. Exercise Schedule
The 3‐week home program consisted of five exercise sessions a week: two 60‐min supervised sessions by a home visiting respiratory physiotherapist and three self‐directed sessions. One of the self‐directed sessions replicated the full routine, while the other two replaced resistance training with extended aerobic exercise, consistent with guidelines to limit resistance work to three times weekly. Exercise instructions were also provided via a pictorial log.
2.2.3. Intervention Arm
Patients in the HFO arm received high flow nasal oxygen administered using the Optiflow nasal cannula with the myAIRVO2 system (Fisher & Paykel, New Zealand) which delivered warmed (34°C), humidified air/oxygen mixture at flow rates up to 60 L/min. Supplemental oxygen was provided via a portable oxygen concentrator (1‐5 L/min). Flow/oxygen settings were titrated during exercise following protocolised Table 1, with the aim of maintaining oxygen saturation (SpO2) ≥ 88%.
TABLE 1.
HFO titration table.
| Steps (upward titration) | Flow rate of AIRVO2 (L/min) | Concentration of oxygen from portable oxygen concentrator (L/min) |
|---|---|---|
| 1 | 30 | 0 |
| 2 | 40 | 0 |
| 3 | 50 | 0 |
| 4 | 60 a | 0 |
| 5 | 60 | 1 |
| 6 | 60 | 2 |
| 7 | 60 | 3 |
| 8 | 60 | 4 |
| 9 | 60 | 5 |
The maximal AIRVO2 flow rate may be lower if patient is unable to tolerate 60 L/min. AIRVO2 flow rate will be kept at patient's highest tolerable flow rate and further stepwise increments will be made only to the oxygen concentration settings on the portable oxygen tank.
2.2.4. Control Arm
Patients in the control arm commenced exercise on room air or normal flow nasal oxygen, depending on their baseline oxygen requirements. Patients on long‐term oxygen therapy (LTOT) started at their usual prescribed flow (e.g., 2 L/min). During exercise, SpO2 was maintained ≥ 88%. For those on room‐air, exercise intensity was reduced if they were unable to maintain the target SpO2.
Details about the exercises, exercise progression, myAirVO2 device or oxygen settings and titrations are provided in the Supporting Information. Patients in both arms received weekly phone call reminders about the self‐directed exercises, with adherence documented via exercise diaries (Figure S1).
2.3. Study Outcome Assessments
We evaluated the feasibility of the early home PR program and the application of HFO using a combination of quantitative and qualitative measures.
2.3.1. Feasibility and Acceptability Outcomes
2.3.1.1. PR Program Uptake and Completion
We recorded PR program acceptance, uptake, and completion rates, and will discuss them in relation to findings reported in the existing literature. PR program acceptance rate was defined as the percentage of eligible patients who accepted participation in the PR program. Program uptake rate was defined as the percentage of eligible patients who completed at least 1 PR session. PR program completion was defined as attending at least 70% of the PR sessions [21] (supervised and self‐directed).
2.3.1.2. Patient Acceptability
Acceptability of both thehome PR program and HFO was assessed with an audio‐recorded, semi‐structured telephone interview in the participant's preferred language (English, Chinese, or Malay). The interviews explored their perceptions of HFO and early home PR as a form of care continuity, the impact on daily activities, barriers and facilitators to implementation, and views on future home‐based interventions. Details regarding qualitative methods, including study conduct, rigour, trustworthiness, and research team reflexivity, are provided in the Supporting Information.
2.3.1.3. Program Fidelity and Implementation Feasibility
Physiotherapists who delivered the PR program completed semi‐structured questionnaires that assessed program adoption, implementation challenges, and adaptations made during delivery.
2.3.2. Quantitative Outcomes
The following outcomes were measured to obtain preliminary estimates of potential intervention effects:
Change in 1‐min Sit‐to‐Stand Test (1mSTST) score (This is the primary clinical outcome measure selected to inform a future definitive trial, while feasibility outcomes were the primary objective of this pilot study)
Change in 30‐s Sit‐to‐Stand Test (30sSTST) score
Change in the proportion of patients with EH
Change in Modified Medical Research Council (mMRC) dyspnoea scale score
Change in forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC)
Change in Hospital Anxiety and Depression Scale (HADS) score
Change in EQ‐5D‐5L health‐related quality‐of‐life score
The change was measured from baseline to 4 weeks, and 12 weeks post‐discharge. The 4‐ and 12‐week assessments were conducted in the participants' homes. The Sit‐to‐Stand tests were performed on room air or the participant's usual baseline oxygen‐flow prescription at all timepoints (see Supporting Information for protocol details).
2.4. Sample Size Estimation and Data Analysis
Using stepped rules of thumb for external pilot trials, 20 participants per arm are recommended when planning a future 80% powered definitive trial to detect a small standardised difference [22]. We therefore aimed to recruit 56 participants with allowance for a 25% attrition rate.
2.4.1. Quantitative Analyses
Quantitative analyses followed intention‐to‐treat principles–all randomised participants were analysed in their original assigned groups. Descriptive statistics included percentages for categorical data, mean ± SD for normal, and median (IQR) for non‐normal distributions assessed via Q–Q plots. Analysis was stratified by baseline diagnosis (COPD vs. non‐COPD), consistent with the stratified randomisation.
Between‐group differences in the primary quantitative outcome (change in 1mSTST repetitions) were analysed using a multivariable linear mixed‐effects model with time (4, 12 weeks), group (PR‐only, HFO‐PR), and their interaction as fixed effects, and a random intercept for patient. Baseline 1mSTST repetition was included a priori as a covariate to improve statistical efficiency [23] and reduce regression to the mean. Model‐adjusted between‐group differences in mean change from baseline were reported with 95% CIs.
Secondary outcomes were exploratory. Within‐group changes from baseline to weeks 4 and 12 were reported as mean ± SD. Between‐group differences in mean change at each timepoint and 95% CIs were estimated using Welch's t‐test for clinical interpretability.
Statistical tests were two‐sided with a 0.05 significance level. All analyses were conducted using R version 4.5.3 (R Core Team, 2026) and Stata 18 (College Station, TX: StataCorp LLC).
Qualitative data were analysed using Braun and Clarke's thematic analysis approach [24]. Three researchers (NH, YM, JWF) independently reviewed transcripts to gain an overall understanding of the participants' narratives, followed by iterative readings to explore perceptions, physiological changes, and daily adjustments related to home exercises and HFO use. Verbatim responses were coded and organised into subthemes and overarching themes, which were collaboratively reviewed and finalised by the research team.
3. Results
3.1. Study Participants
Fifty‐six eligible patients were randomised to either PR‐only (n = 29) or HFO‐PR (n = 27). Fifty‐two patients completed study follow‐up at 4 weeks, with two dropouts in each arm. In the PR‐only group, 27 (93%) and 25 (86%) completed 1mSTST assessment at 4‐ and 12‐weeks post‐discharge, respectively. In the HFO‐PR group, 25 (93%) completed assessments at both timepoints, respectively (Figure 1).
FIGURE 1.
Study flow diagram. 1mSTST: 1 min sit‐to‐stand test. The 1mSTST is presented in this figure, as it is the primary quantitative outcome of interest. *1 participant was unable to perform the 1mSTST test due to his physical condition although he completed the rest of the secondary outcome assessments.
Participants were predominantly male (82%), with a mean age of 72 ± 8 years. Most patients had a diagnosis of COPD exacerbation (n = 48, 86%), followed by bronchiectasis exacerbation (n = 6, 11%), and one each with ILD exacerbation and Covid‐19 pneumonia. Baseline characteristics were well‐balanced between the intervention groups (Table 2).
TABLE 2.
Baseline patient demographic and clinical characteristics, by randomly assigned group.
| Characteristic | PR‐only (n = 29) | HFO‐PR (n = 27) |
|---|---|---|
| Age, mean ± SD | 72 ± 8 | 71 ± 9 |
| Female, n (%) | 4 (14) | 6 (22) |
| Diagnosis, n (%) | ||
| COPD | 25 (86) | 23 (85) |
| Bronchiectasis | 3 (10) | 3 (11) |
| COVID‐19 | 1 (3) | 0 (0) |
| ILD | 0 (0) | 1 (4) |
| BMI, median (IQR) | 20 (17.2–25.1) | 22.5 (18.8–26.2) |
| Smoking, n (%) | ||
| Active smoker | 8 (28) | 8 (30) |
| Ex‐smoker | 19 (66) | 18 (67) |
| Non‐smoker | 2 (7) | 1 (4) |
| Comorbidity, n (%) | ||
| IHD | 11 (38) | 10 (37) |
| Other CVD | 11 (38) | 8 (30) |
| Cerebrovascular | 5 (17) | 2 (7) |
| DM | 6 (21) | 9 (33) |
| Other endocrine disease | 26 (90) | 23 (85) |
| Anxiety | 1 (3) | 1 (4) |
| Depression | 1 (3) | 0 (0) |
| Other psychiatric comorbidity | 1 (3) | 0 (0) |
| Use of long‐term O2 supplementation, n (%) | 8 (28) | 8 (30) |
| Severity of EH, mean ± SD | ||
| SpO2 desaturation during 1mSTST in pp | 6.6 ± 2.6 | 6.1 ± 2.6 |
| Post‐1mSTST SpO2 in % | 87.8 ± 3.8 | 87.0 ± 3.7 |
Abbreviations: 1mSTST, 1‐min sit‐to‐stand test; BMI, body mass index; COPD, chronic obstructive pulmonary disease; COVID‐19, coronavirus disease 2019; CVD, cardiovascular disease; DM, diabetes mellitus; EH, exertional hypoxaemia; HFO‐PR, high flow nasal oxygen with pulmonary rehabilitation; IHD, ischemic heart disease; ILD, interstitial lung disease; IQR, interquartile range; O2, oxygen; PR‐only, pulmonary rehabilitation without high flow nasal oxygen; SD, standard deviation; SpO2, peripheral oxygen saturation.
3.2. PR Program Acceptance, Uptake and Completion
Of the 84 eligible patients approached, 56 agreed to study participation and 52 commenced at least 1 PR session. Therefore, program acceptance and uptake rates were 67% and 62%, respectively. Of the 28 eligible patients who rejected participation upfront, 3 cited difficulties with commitment to the PR schedule while the rest cited a lack of interest.
Program completion was lower in the HFO‐PR group (15/25, 60%) compared to the PR‐only group (24/27, 89%). For self‐directed sessions, completion was 56% (14/25) in the HFO‐PR group (at least 70% of the 9 intended sessions), and 85% (23/27) in the PR‐only group.
We also observed participants who exceeded the prescribed number of exercise days. In the HFO‐PR group, 48% (12/25) and in the PR‐only group, 59% (16/27) of participants completed more exercise days than scheduled. This additional activity was driven primarily by increased engagement in self‐directed exercise sessions.
3.3. Qualitative Findings on Patient Acceptability
All 52 patients who completed participation in the study at 4 weeks were interviewed. Three overarching themes emerged from the interviews on early home PR: (1) Transformations following the home‐based program, (2) Receiving care and education at home, and (3) Experiencing the High Flow Nasal Oxygen (HFO) system (Table 3). The full list of subthemes and quotations is reported in the Supporting Information (Table S1).
TABLE 3.
Qualitative themes and subthemes of patient experiences in receiving early home pulmonary rehabilitation.
| Themes | Subthemes | Illustrative quotes |
|---|---|---|
| 1. Transformations following the home‐based program | 1.1 Improving physical conditions |
“I am improving in my breathing…Now I do a lot of walking.” (KMRA 036) “I even see muscle “grows” on my leg. The exercises, such as climbing the stairs and pulling of string have helped me in strengthening my muscles.” (KMRA 002) |
| 1.2 Enhancing psychosocial aspects and confidence |
“I went out by myself today…I previously didn't dare to go out.” (KMRA050) “It helps us, and it boosts our morale knowing that somebody comes and is interested in us (and in how we are)living our lives…I should say so, because we need companionship and someone to visit us and, you know, always spend 5 min talking.” (KMRA 031) |
|
| 1.3 Behavioural modifications |
“I reduce the cigarette.” (KMRA 048) “Now I have changed, I'm exercising every day. I did not do that last time.” (KMRA048) |
|
| 2. Receiving care and patient education at home | 2.1 Feeling convenience and overcoming distance limitation |
“I think it's very good because as you know, people with my condition (like) COPD would mainly be elderly people…They have to go to hospital or go on their own, (or) outside to get these exercises done. It might be too much for them, for elderly people you know. This home therapy exercises, I think it's very good for us…The problem of travelling up‐down, up‐down.” (KMRA 005) “I think that it is a very good reach‐out because most people (if) left to themselves…I don't think they'll go forward to look for it. So, you can come to them, I'm sure they'll accept it.” (KMRA 048) |
| 2.2 Learning to exercise and setting goals |
“In my opinion it's very good for the patient, you see. The patient will know how to exercise and know what (are) the things they need to do.” (KMRA 029) “I want to be healthy, if possible be independent, can walk, can meet friends, can work.” (KMRA 007) |
|
| 2.3 Moving towards independence and preference |
“No. I don't want it anymore…I would like to do (exercise) on my own”. (KMRA 006) “No time.” (KMRA 008) |
|
| Experiencing the High Flow Nasal Oxygen (HFO) system | 3.1 Breathing and living with the device |
“It's just the breathing technique, that I find is important. I only realized it after using the device because I have never used such device before” (KMRA 024) “It helped me greatly…allowed me to understand the importance of breathing while exercising.” (KMRA 024) “Difficult to breathe in and out…But just have to get used to it.” (KMRA 021) “I felt it's more difficult for me to breathe while putting it on… I'm not used to putting things in my nose.” (KMRA 055) “I'm going to tell you that I'm probably doing exercises but (if I) don't rely on HFO (it) will be better…if I don't put on that oxygen thing from the HFO. Oh, I'll do (the exercises) by myself (which is) better.” (KMRA 055) |
| 3.2 Operating the device and performing daily activities |
“It's a very simple machine, you see. For the small one, what do you call it, the high flow, just throw the old water, put the new water in, then you see the signal, then you start exercise. Then after that you (switch it) off. Very simple and it's compact. Yeah, it's compact, so it (does) not take so much space.” (KMRA026) “Very troublesome. Nobody would want to use it, so complicated to use and the sound is noisy too (HFO device).” (KMRA 023) “…exercising using this device limits the exercises I can do. This device can't help in walking long distances. It is limited to certain exercise (regimes)…I can't be using the device when climbing stairs too…The current device is difficult to utilize.” (KMRA 024) |
Overall, participants viewed early home PR as a positive and transformative experience, with benefits spanning across physical, psychological, and functional domains. They described improved endurance, energy, confidence, and reduced anxiety, alongside enhanced motivation and social connection during home visits. Some adopted healthier habits, such as regular exercise and reduced smoking, indicating potential for long‐term behavioural change. The convenience of avoiding hospital travel and the relevance of home‐based education were highly valued, though some preferred autonomous exercise or found time constraints limited participation.
Experiences with HFO varied. Facilitators included interest in breathing techniques and willingness to tolerate initial discomfort; barriers included inability to acclimatise, perceived lack of benefit, or preference for ‘natural’ breathing. Some found HFO simple and manageable with family support, while others noted operational complexity and mobility restrictions. These differences highlight that HFO acceptability depends on individual preferences, capabilities, and goals.
3.4. Program Fidelity and Implementation Feasibility
Physiotherapists' feedback echoed the participants' perspectives, noting challenges with water chamber handling especially among older adults. Space limitations and poor device mobility also restricted home exercise options, indicating the value of environmental assessment. Varying levels of patient tolerance to the HFO system settings were observed, underscoring the need for more flexible protocols. Despite these challenges, physiotherapists appreciated the structured format of a protocol for ensuring fidelity and monitoring but emphasised the need for greater flexibility in tailoring exercise prescriptions.
3.5. Quantitative Outcomes
3.5.1. Primary Outcome of 1‐Min Sit‐To‐Stand Change Score
Among COPD participants, the mean number of 1mSTST repetitions at baseline was 16 ± 6 for HFO‐PR and 14 ± 6 for PR‐only. At week 4, the baseline‐adjusted mean change was 5 (95% CI 2–7) for HFO‐PR and 5 (95% CI 3–8) for PR‐only. No significant between‐group differences were found at week 4 (B −1 repetition, 95% CI −5 to 3; p = 0.700) or week 12 (B 0 repetitions, 95% CI −4 to 4; p = 0.860). Table 4 shows the results stratified by COPD diagnosis.
TABLE 4.
Main outcome of interest for the PR‐only and HFO‐PR group.
| Stratification factor and group | No. of repetitions in 1mSTST | Adjusted between‐group difference (95% CI) in mean change from baseline, HFO‐PR vs. PR‐only | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Baseline | Change at week 4 | Change at week 12 | ||||||||
| n | Raw mean ± SD | n | Raw mean ± SD | Estimated marginal mean (95% CI) | n | Raw mean ± SD | Estimated marginal mean (95% CI) | Week 4 | Week 12 | |
| COPD | ||||||||||
| PR‐only | 23 | 14.4 ± 5.8 | 23 | 5.7 ± 5.2 | 5.3 (2.7 to 8.0) | 22 | 5.4 ± 5.5 | 4.6 (1.9 to 7.4) | — | — |
| HFO‐PR | 21 | 16.0 ± 6.0 | 21 | 4.1 ± 7.9 | 4.6 (1.8 to 7.4) | 21 | 3.9 ± 9.0 | 4.3 (1.5 to 7.1) | −0.8 (−4.7 to 3.1) | −0.3 (−4.3 to 3.6) |
| Non‐COPD | ||||||||||
| PR‐only | 4 | 15.3 ± 2.1 | 4 | 3.3 ± 12.6 | 1.1 (−9.7 to 11.8) | 3 | 5.7 ± 7.6 | 0.1 (−11.0 to 11.3) | — | — |
| HFO‐PR | 4 | 20.0 ± 7.1 | 4 | −0.5 ± 7.9 | 1.7 (−9.1 to 12.5) | 4 | −4.0 ± 3.7 | −1.8 (−12.6 to 9.0) | 0.6 (−15.5 to 16.7) | −2.0 (−18.3 to 14.4) |
Abbreviations: 1mSTST, 1‐min sit‐to‐stand test; CI, confidence interval; COPD, chronic obstructive pulmonary disease; HFO‐PR, high flow nasal oxygen with pulmonary rehabilitation; PR‐only, pulmonary rehabilitation without high flow nasal oxygen; SD, standard deviation.
3.5.2. Secondary Outcomes
Quantitative findings for EH, 30sSTST, mMRC, FEV1, FVC, HADS, and EQ‐5D scores are presented in Table 5 and Table S2 (Supporting Information), respectively. As a pilot study, the trial was not powered to draw definitive conclusions regarding effectiveness.
TABLE 5.
Secondary outcomes for the PR‐only and HFO‐PR groups among patients with baseline COPD.
| Secondary outcome | PR‐only | HFO‐PR | Between‐group difference (95% CI) in change from baseline, HFO‐PR vs. PR‐only | |||||
|---|---|---|---|---|---|---|---|---|
| Baseline (n = 23) | Raw change at week 4 (n = 23) | Raw change at week 12 (n = 22) | Baseline (n = 21) | Raw change at week 4 (n = 21) | Raw change at week 12 (n = 21) | Week 4 | Week 12 | |
| EH prevalence | 23 (100) | −30.4 | −50.0 | 21 (100) | −47.6 | −38.1 | −17.2 (−47.2 to 12.8) | 11.9 (−19.2 to 43.0) |
| 30sSTST | ||||||||
| No. of repetitions | 9.0 ± 3.5 | 2.1 ± 3.5 | 2.0 ± 3.7 | 9.1 ± 3.3 | 1.7 ± 3.6 | 2.4 ± 4.2 | −0.4 (−2.6 to 1.7) | 0.3 (−2.1 to 2.8) |
| Respiratory function | ||||||||
| FEV1 in L | 0.79 ± 0.33 | 0.06 ± 0.20 | 0.08 ± 0.25 | 0.84 ± 0.34 | 0.11 ± 0.24 | 0.06 ± 0.29 | 0.05 (−0.09 to 0.18) | −0.02 (−0.19 to 0.14) |
| FVC in L | 1.58 ± 0.77 | 0.12 ± 1.27 | −0.06 ± 0.82 | 1.56 ± 0.65 | 0.15 ± 0.57 | 0.13 ± 0.54 | 0.03 (−0.56 to 0.63) | 0.19 (−0.24 to 0.62) |
| mMRC dyspnoea score | 2.6 ± 1.2 | −0.2 ± 1.3 | −0.4 ± 1.4 | 2.4 ± 1.1 | −0.6 ± 1.6 | −0.1 ± 1.5 | −0.4 (−1.3 to 0.5) | 0.3 (−0.6 to 1.2) |
| HADS | ||||||||
| Total score (depression) | 4.6 ± 3.2 | 4.1 ± 5.7 | 3.8 ± 5.5 | 2.3 ± 2.0 | 0.9 ± 3.3 | 1.4 ± 3.5 | −3.3 (−6.1 to −0.5) | −2.4 (−5.3 to 0.4) |
| Total score (anxiety) | 5.4 ± 4.3 | 1.1 ± 4.8 | −0.5 ± 5.1 | 2.9 ± 2.3 | −0.3 ± 3.0 | −0.5 ± 2.9 | −1.4 (−3.8 to 1.1) | 0.0 (−2.5 to 2.6) |
| EQ‐5D‐5L | ||||||||
| Index score | 0.36 ± 0.40 | −0.05 ± 0.41 | 0.09 ± 0.30 | 0.65 ± 0.26 | −0.00 ± 0.35 | −0.04 ± 0.47 | 0.05 (−0.19 to 0.28) | −0.13 (−0.38 to 0.12) |
| EQ VAS (Your own health state today) | 62.8 ± 19.4 | −1.1 ± 33.0 | 1.5 ± 29.1 | 70.1 ± 17.5 | 4.0 ± 15.2 | −0.3 ± 18.4 | 5.1 (−10.5 to 20.7) | −1.8 (−16.8 to 13.2) |
Note: Data are reported as mean ± standard deviation. 1mSTST EH prevalence was reported as number and percent at baseline, and change in percentage points at week 4 and 12.
Abbreviations: 1mSTST, 1‐min sit‐to‐stand test; 30sSTST, 30‐s sit‐to‐stand test; EH, exertional hypoxaemia; EQ‐5D‐5L, EuroQol 5‐dimension 5‐level; FEV1, forced expiratory volume in one second; FVC, forced vital capacity; HADS, Hospital Anxiety and Depression Scale; HFO‐PR, high flow nasal oxygen with pulmonary rehabilitation; mMRC, Modified Medical Research Council; PR‐only, pulmonary rehabilitation without high flow nasal oxygen; VAS, Visual Analogue Scale.
3.6. Adverse Events
A total of 11 serious adverse events were reported across all participants during their 3‐week intervention periods. Of these, seven were non‐respiratory causes (e.g., intestinal obstruction), three were COPD exacerbations, and one was related to a COVID‐19 infection. None of the reported events were attributed to the high‐flow oxygen (HFO) intervention or the PR program.
3.7. Patient Crossover
One LTOT (2 L/min) participant crossed‐over from the HFO‐PR to the PR‐only arm due to an inability to maintain SpO2 > 85% during exercise, despite maximum HFO protocol settings (flow rate of 60 L/min with 5 L/min oxygen) or lowest protocol flow rate setting of 30 L/min with 5 L/min oxygen. In contrast, the patient maintained SpO2 levels of 92%–95% at rest on LTOT alone and tolerated exercise without requiring HFO. Therefore, a protocol exception was granted to allow crossover to the PR‐only arm.
4. Discussion
Our early, post‐hospitalisation, three‐week intensive home‐based PR program demonstrated an overall uptake rate of 62% across the study population. Qualitative findings indicated that the home PR program was generally well received. In the PR‐only arm, which was intended to assess the feasibility of the core rehabilitation program independent of adjunctive therapies, the completion rate was 89%, with several participants completing more exercise sessions than scheduled.
Reported PR uptake varies widely depending on settings and definitions [4, 25, 26, 27], ranging from 1.9% in population‐level analyses [25] to approximately 80% in selected outpatient cohorts [26]. In a UK audit, 67% (60/90) of patients commenced PR after COPD‐related hospitalisation, of whom 72% completed the program [4]. In a survey of 155 facilities in Canada offering PR, the majority reported program completion rates of more than 60% [28]. Taken together, our overall home‐based PR program uptake and completion rate (PR‐only arm) compare favourably with those reported for conventional outpatient PR programs, supporting the feasibility of our home‐based model.
Our study specifically addressed post‐hospitalisation rehabilitation strategies for EH patients, a clinically challenging and traditionally underserved population. Our findings demonstrate that very early, home‐based PR is not only feasible but also safe, contrary to common provider concerns about early initiation. Notably, no patients expressed concerns about the early timing of PR initiation post‐discharge. The hybrid model of supervised and self‐directed sessions offers a pragmatic, and scalable approach for this population.
We also evaluated the integration of HFO into home‐based PR to mitigate physiological constraints imposed by EH during exercise. While no safety concerns were identified with HFO use, completion rates were lower in the HFO‐PR arm (60%) compared with the PR‐only arm (89%), with poorer adherence to self‐directed sessions. Qualitative findings from participant interviews and physiotherapists' feedback identified several usability challenges with the HFO device, including difficulty refilling the water chamber, limited portability, and low user confidence during unsupervised use. These issues persisted despite additional training and the provision of visual support materials. The limited mobility of the device also constrained the types of home exercises participants could perform, reducing exercise variety and potentially diminishing engagement. Future HFO designs with improved portability, battery capacity, and simplified refilling mechanisms could enhance usability and facilitate better adherence.
Beyond technical issues, qualitative findings suggest that certain psychological characteristics may influence HFO acceptance. These include interest in breathing mechanics, comfort with technology, and willingness to tolerate initial discomfort. These observations are exploratory and require further study, but identifying such factors could help refine patient selection and optimise outcomes in future HFO‐integrated PR programs. Future trials could incorporate hands‐on device orientation and assess patient comfort with technology prior to enrolment.
One LTOT‐dependent COPD patient crossed over from the HFO‐PR to the PR‐only arm after failing to maintain SpO2 ≥ 85% despite setting adjustments. This likely reflects greater disease severity and higher respiratory demand that could not be met by the current HFO device and oxygen delivery limits. Further FiO2 increases were restricted by the concentrator's 5 L/min maximum output, and the lack of real‐time FiO2 display on myAIRVO2 impeded precise titration. We recommend incorporating FiO2 monitoring and more compatible oxygen systems in future home HFO designs. Until such improvements occur, HFO use in LTOT‐dependent patients should be prescribed with caution in home PR programs.
Quantitative outcomes, including the 1mSTST and secondary measures, did not demonstrate a clear between‐group benefit of HFO. However, as this pilot study was not powered for efficacy, these findings remain exploratory and are intended to inform future trial design.
This study has several strengths. Despite being a pilot, the relatively large sample of 56 patients allowed clinical staff to gain substantial experience in delivering the early home intervention. This, in turn, enabled a thorough evaluation of its feasibility and the practical nuances of implementing HFO‐supported exercise training in the home PR setting. Our study also had a well‐defined HFO/oxygen titration protocol which enhanced replicability. The combined use of quantitative and qualitative measures enabled a comprehensive assessment of the interventions and patient experiences, facilitating future holistic refinements.
Limitations included reliance on paper‐based exercise diaries, which may introduce recall bias. Although digital tools were considered, paper formats were deemed to be more practical for our predominantly elderly, less tech‐savvy participants. In this pilot setting, physiotherapists were restricted to a pre‐defined set of exercises to minimise practice variability. Future studies could consider more flexible, intensity‐driven exercise prescriptions, which may better differentiate outcomes between intervention arms.
Oxygen titration during exercise differed between study arms. The HFO‐PR group followed a protocolised titration strategy with adjustment of flow and FiO2, whereas participants not on LTOT in the PR‐only group did not receive supplemental oxygen for exercise‐induced desaturation, reflecting usual home‐based care. This pragmatic design introduces a potential imbalance in physiological support that may favour the intervention arm. However, as this pilot study was not designed or powered to assess comparative efficacy and was primarily intended to evaluate feasibility, acceptability, and for implementation learning, this limitation should be interpreted within that context. In this setting, comparison with usual care provides a pragmatic and relevant benchmark. Future definitive trials should consider more symmetric oxygen strategies or stratified protocols to minimise this bias.
Blinding of participants and providers was not feasible during exercise sessions due to the nature of the HFO device. As a small team conducted this pilot, the 1mSTST assessor may have been aware of group allocation. Interviews conducted in Mandarin or Malay for non‐English speakers may have lost some nuance in translation, though bilingual interviewers were employed to ensure accuracy. Lastly, the study cohort was predominantly male and comprised mostly COPD patients, underscoring the need for future studies in more diverse populations and disease groups.
In conclusion, an early, home‐based PR program was feasible and well tolerated in patients with EH following hospitalisation. However, the adjunct use of HFO in its current form was not feasible. Future work should prioritise user‐centred device redesign before further evaluation of HFO's role and potential benefits in home‐based PR. Once usability issues are addressed, research should also explore more individualised exercise prescriptions, as well as identify clinical and behavioural factors predictive of adherence and response to home PR and HFO. Such patient phenotyping may help target interventions to those most likely to benefit, thereby maximising both clinical impact and cost‐effectiveness.
Author Contributions
Yingjuan Mok: conceptualization, methodology, project administration, writing – original draft, formal analysis, supervision. Hang Siang Wong: conceptualization, methodology, writing – review and editing. Norasyikin Hassan: conceptualization, methodology, investigation, formal analysis, writing – original draft. Edwin Chan Shih‐Yen: conceptualization, methodology, writing – review and editing. Shi Hua Tan: methodology, investigation, writing – review and editing. Amanda Soh: methodology, investigation, writing – review and editing. Keith Keat Huat Wong: conceptualization, methodology, writing – review and editing. Poh Choo Tan: methodology, writing – review and editing. Li Ching Soh: investigation, writing – review and editing. Piyanee Klainin‐Yobas: formal analysis, writing – original draft. Jing Wen Foong: conceptualization, methodology, investigation, formal analysis, writing – original draft. Xuan Han Koh: methodology, formal analysis, writing – original draft.
Funding
The study was supported by Duke‐NUS Medical School, the Estate of ‘Tan Sri Khoo Teck Puat’, and the SingHealth Foundation. Open access publication was supported separately by SingHealth Group Allied Health/SingHealth Group Pharmacy. None of the funders had any role in the study design, conduct, data analysis, interpretation, or decision to publish the findings.
Disclosure
No AI‐assisted technology was used for study design, data analysis, interpretation of results, or generation of scientific content.
Ethics Statement
The study was approved by the SingHealth Centralised Institutional Review Board Ref: 2022/2605 (21 November 2022) and underwent trial registration (ClinicalTrials.gov: NCT05752370). Eligible patients were identified through the inpatient physiotherapy service and written informed consent was obtained.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: Qualitative themes and subthemes of patient experiences in receiving early home pulmonary rehabilitation.
Table S2: Secondary outcomes for the PR‐only and HFO‐PR groups among baseline non‐COPD patients.
Figure S1: Exercise diary.
Acknowledgements
We extend our heartfelt appreciation to all the patients for their participation in this study. We also wish to acknowledge the Duke‐NUS Centre for Clinician‐Scientist Development, the Changi General Hospital Clinical Trials Research Unit with special mention to Zhang Yajie and Geraldine Lim, and the Departments of Physiotherapy, Respiratory and Critical Care Medicine, Care and Health Integration, and Health Services Research for their invaluable support.
Mok Y., Foong J. W., Wong H. S., et al., “High‐Flow Nasal Oxygen in Early Home Pulmonary Rehabilitation for Exertional Hypoxaemia: A Pilot Randomised Feasibility Trial,” Respirology 31, no. 10 (2026): 1029–1040, 10.1002/resp.70277.
Associate Editor: Amanda Jane Piper
Preliminary findings of this work were presented as an oral presentation in the “Emerging Clinical Trials in Rehabilitation and Chronic Care” session at the European Respiratory Congress 2025.
Senior Editor: Fanny Ko
[Correction added on 10 July 2026, after first online publication: The copyright line was changed.]
Data Availability Statement
De‐identified data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. Puhan M. A., Gimeno‐Santos E., Cates C. J., and Troosters T., “Pulmonary Rehabilitation Following Exacerbations of Chronic Obstructive Pulmonary Disease,” Cochrane Database of Systematic Reviews 12 (2016): CD005305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. California Pulmonary Rehabilitation Collaborative Group , “Effects of Pulmonary Rehabilitation on Dyspnea, Quality of Life, and Healthcare Costs in California,” Journal of Cardiopulmonary Rehabilitation 24, no. 1 (2004): 52–62. [DOI] [PubMed] [Google Scholar]
- 3. Hayton C., Clark A., Olive S., et al., “Barriers to Pulmonary Rehabilitation: Characteristics That Predict Patient Attendance and Adherence,” Respiratory Medicine 107, no. 3 (2013): 401–407. [DOI] [PubMed] [Google Scholar]
- 4. Jones S. E., Green S. A., Clark A. L., et al., “Pulmonary Rehabilitation Following Hospitalisation for Acute Exacerbation of COPD: Referrals, Uptake and Adherence,” Thorax 69, no. 2 (2014): 181–182. [DOI] [PubMed] [Google Scholar]
- 5. Keating A., Lee A., and Holland A. E., “What Prevents People With Chronic Obstructive Pulmonary Disease From Attending Pulmonary Rehabilitation? A Systematic Review,” Chronic Respiratory Disease 8, no. 2 (2011): 89–99. [DOI] [PubMed] [Google Scholar]
- 6. Zeng Q., Chen W., Xu S., et al., “Understanding Facilitators and Barriers to Adherence in Home‐Based Pulmonary Rehabilitation for Chronic Obstructive Pulmonary Disease Patients Using the Health Belief Model: A Qualitative Study,” Frontiers in Public Health 13 (2025): 1553744. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Thorpe O., Kumar S., and Johnston K., “Barriers to and Enablers of Physical Activity in Patients With COPD Following a Hospital Admission: A Qualitative Study,” International Journal of Chronic Obstructive Pulmonary Disease 9 (2014): 115–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Jenkins S. and Cecins N., “Six‐Minute Walk Test: Observed Adverse Events and Oxygen Desaturation in a Large Cohort of Patients With Chronic Lung Disease,” Internal Medicine Journal 41, no. 5 (2011): 416–422. [DOI] [PubMed] [Google Scholar]
- 9. Paneroni M., Spinello L., Salvi B., et al., “Prevalence of Exercise‐Induced Desaturation Among COPD Patients Enrolled in Early Inpatient Pulmonary Rehabilitation,” Chronic Obstructive Pulmonary Disease 13, no. 2 (2026): 136–146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Moreira M. A., Medeiros G. A., Boeno F. P., Sanches P. R., Silva Junior D. P., and Muller A. F., “Oxygen Desaturation During the Six‐Minute Walk Test in COPD Patients,” Jornal Brasileiro de Pneumologia 40, no. 3 (2014): 222–228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Chang C. H., Lin H. C., Yang C. H., et al., “Factors Associated With Exercise‐Induced Desaturation in Patients With Chronic Obstructive Pulmonary Disease,” International Journal of Chronic Obstructive Pulmonary Disease 15 (2020): 2643–2652. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Yohannes A. M., Casaburi R., Dryden S., and Hanania N. A., “Predictors of Premature Discontinuation and Prevalence of Dropouts From a Pulmonary Rehabilitation Program in Patients With Chronic Obstructive Pulmonary Disease,” Respiratory Medicine 193 (2022): 106742. [DOI] [PubMed] [Google Scholar]
- 13. Torres‐Sanchez I., Cabrera‐Martos I., Diaz‐Pelegrina A., Valenza‐Demet G., Moreno‐Ramirez M. P., and Valenza M. C., “Physical and Functional Impairment During and After Hospitalization in Subjects With Severe COPD Exacerbation,” Respiratory Care 62, no. 2 (2017): 209–214. [DOI] [PubMed] [Google Scholar]
- 14. Liu Y. and Gong F., “Determination of Whether Supplemental Oxygen Therapy Is Beneficial During Exercise Training in Patients With COPD: A Systematic Review and Meta‐Analysis,” Experimental and Therapeutic Medicine 18, no. 5 (2019): 4081–4089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Menadue C., Piper A. J., van't Hul A. J., and Wong K. K., “Non‐Invasive Ventilation During Exercise Training for People With Chronic Obstructive Pulmonary Disease,” Cochrane Database of Systematic Reviews 5 (2014): CD007714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Groves N. and Tobin A., “High Flow Nasal Oxygen Generates Positive Airway Pressure in Adult Volunteers,” Australian Critical Care 20, no. 4 (2007): 126–131. [DOI] [PubMed] [Google Scholar]
- 17. Parke R. L., Eccleston M. L., and SP M. G., “The Effects of Flow on Airway Pressure During Nasal High‐Flow Oxygen Therapy,” Respiratory Care 56, no. 8 (2011): 1151–1155. [DOI] [PubMed] [Google Scholar]
- 18. Fraser J. F., Spooner A. J., Dunster K. R., Anstey C. M., and Corley A., “Nasal High Flow Oxygen Therapy in Patients With COPD Reduces Respiratory Rate and Tissue Carbon Dioxide While Increasing Tidal and End‐Expiratory Lung Volumes: A Randomised Crossover Trial,” Thorax 71, no. 8 (2016): 759–761. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Eldridge S. M., Chan C. L., Campbell M. J., et al., “CONSORT 2010 Statement: Extension to Randomised Pilot and Feasibility Trials,” BMJ 355 (2016): i5239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Rochester C. L., Alison J. A., Carlin B., et al., “Pulmonary Rehabilitation for Adults With Chronic Respiratory Disease: An Official American Thoracic Society Clinical Practice Guideline,” American Journal of Respiratory and Critical Care Medicine 208, no. 4 (2023): e7–e26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Lahham A. and Holland A. E., “The Need for Expanding Pulmonary Rehabilitation Services,” Life (Basel) 11, no. 11 (2021): 1236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Whitehead A. L., Julious S. A., Cooper C. L., and Campbell M. J., “Estimating the Sample Size for a Pilot Randomised Trial to Minimise the Overall Trial Sample Size for the External Pilot and Main Trial for a Continuous Outcome Variable,” Statistical Methods in Medical Research 25, no. 3 (2016): 1057–1073. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Holmberg M. J. and Andersen L. W., “Adjustment for Baseline Characteristics in Randomized Clinical Trials,” Journal of the American Medical Association 328, no. 21 (2022): 2155–2156. [DOI] [PubMed] [Google Scholar]
- 24. Braun V. and Clarke V., “Using Thematic Analysis in Psychology,” Qualitative Research in Psychology 3, no. 2 (2006): 77–101. [Google Scholar]
- 25. Spitzer K. A., Stefan M. S., Priya A., et al., “Participation in Pulmonary Rehabilitation After Hospitalization for Chronic Obstructive Pulmonary Disease Among Medicare Beneficiaries,” Annals of the American Thoracic Society 16, no. 1 (2019): 99–106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. WM M. N. A., Williams G., Delacey D., George C., and Beasley R., “An Audit of Pulmonary Rehabilitation Program,” Clinical Audit 8 (2016): 7–12. [Google Scholar]
- 27. Guecamburu M., Verdy G., Cuadros J., et al., “Insufficient Pulmonary Rehabilitation Uptake After Severe Exacerbation of COPD: A Multicentre Study in the South West Region of France,” International Journal of Chronic Obstructive Pulmonary Disease 19 (2024): 1579–1589. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Camp P. G., Hernandez P., Bourbeau J., et al., “Pulmonary Rehabilitation in Canada: A Report From the Canadian Thoracic Society COPD Clinical Assembly,” Canadian Respiratory Journal 22, no. 3 (2015): 147–152. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Qualitative themes and subthemes of patient experiences in receiving early home pulmonary rehabilitation.
Table S2: Secondary outcomes for the PR‐only and HFO‐PR groups among baseline non‐COPD patients.
Figure S1: Exercise diary.
Data Availability Statement
De‐identified data that support the findings of this study are available from the corresponding author upon reasonable request.