Treatment of Acute Hypercapnic Respiratory Failure With OptiflowTM or Optiflow+DuetTM Nasal Cannula in COPD-patients

The newly developed OptiflowTM+Duet with different sizes of the two nasal prongs and is primarily intended for use in an acute setting. Compared to the original OptiflowTM nasal cannula, the OptiflowTM +Duet is predicted to deliver a doubled to tripled airway pressure, depending on flow. Whether the OptiflowTM+Duet nasal cannula effectively can reduce PaCO2 and normalize pH in patients with COPD exacerbation and acute hypercapnic failure is unknown.

Aims To investigate the effect of HFNC in combination with either OptiflowTM or OptiflowTM+Duet nasal cannula on PaCO2 levels and pH in patients with COPD exacerbation and acute hypercapnic failure and compare the results of treatment with the two different nasal cannulas.

To describe adherence to treatment with high flow and either OptiflowTM or OptiflowTM+Duet nasal cannula.

Study design The study will be carried out as a prospective, multicenter, randomized controlled trial including patients from pulmonary wards in three Danish hospitals: Bispebjerg Hospital, Copenhagen, Viborg Hospital, Viborg, and Aalborg University Hospital, Aalborg, Denmark.

First patient in and study duration The study is initiated Jan 31. 2024. All sites will be ready to include patients from June 30, 2024.

The duration of the study is expected to be twelve months but will be determined by inclusion of study participants. The study group has decided to evaluate after the first six months' inclusion in the study; our main concern could be that these experienced NIV-sites will have a lower NIV-treatment failure rate than could be expected based on recent literature, and thereby a low inclusion rate in the study. This could potentially bias outcome, with a negative selection bias in the included patients. A lower-than-expected inclusion rate will open for inclusion of further sites in the study, of course based on an amendment to the protocol.

Equipment Patients COPD and acute hypercapnic will be treated with HFNC for respiratory support in case of treatment failure with NIV, as defined below. This is standard procedure in all three departments. In this study AirvoTM 3, ("Fisher \& Paykel Healthcare", Auckland, New Zealand) will be used as the device for delivering high flow, and OptiflowTM /OptiflowTM+Duet nasal cannulas ("Fisher \& Paykel Healthcare", Auckland, New Zealand) will be used as the flow-conducting system, leading the heated, humidified, oxygenated air to the patients. In addition, Nonin Model 3231 USB External Pulse Oximeter (Nonin Medical Inc, Plymouth, MN, USA), will be used for oxygen saturation measurement, carried out on the AirvoTM 3 equipment, as well as for monitoring respiratory rate. Data from the AirvoTM 3 software is stored within the device and will be downloaded at the end of each patient's use of the device.

Study procedure Patients who do not tolerate NIV-treatment, that is patients who have been informed of NIV being the target treatment for their respiratory problem, yet desist treatment (e.g. due to mask intolerance, claustrophobia, intolerance of positive airway pressure), will be offered the change to HFNC-treatment, which is usual procedure at all sites. HFNC-treatment should be introduced as soon as possible after treatment failure is a reality. If patients accept, they will be invited to participate in this study. Patients will receive written information. They will be informed verbally about the purpose of the study and their rights as research participants, orally and in writing concomitantly. Unfortunately, informed consent will have to be obtained immediately, as a delay in treatment will influence prognosis (further discussed under "Ethical Considerations"), yet the informant will, during the oral information, offer the patient to be of assistance with including a bystander, either by phone or otherwise online connection, during the oral information. Furthermore, the informant will actively offer the patient to confer with an involved bystander without members of the study team present, both before and after delivering oral information. Information will be given to the patient in the ward, it will however be secured that the communication is not disturbed by personnel or other patients.

The patient will be informed that a consent to access case records includes access to the case file, The Common Medicine Chart by the research responsible, sponsor and those who represent them to perform the study and potentially by authorities that oversee scientific procedures to secure correct data inclusion, quality control, and required monitoring.

Patients will be randomized to either OptiflowTM or OptiflowTM+Duet nasal cannulas. The size of the cannula is fitted by the ward staff according to written instructions. HFNC treatment with allocated nasal cannula, flow 40-60 (prescribed by the responsible clinician) will be initiated. Titration of flow should be to the maximal tolerated flow and of FiO2 till target SO2 is reached (as prescribed by the responsible clinician or by default 88-92%). Arterial puncture (registering pH, PaO2, PaCO2, HCO3, SaO2 and Base Excess) will be drawn at baseline and repeated after two hours (±30 minutes or at least 30 minutes after stable flow- and FiO2 has been reached) and at termination of the HFNC. Should pH reduce ≤ 0.2 after two hours compared to baseline the arterial puncture should be repeated after one hour; if pH reduces < 0.2 after two hours it is considered treatment failure and the patients will terminate study participation and continue treatment at the responsible physician's discretion. If HFNC is terminated before two hours' treatment, arterial puncture will be repeated at termination. If further arterial blood gas analyses are drawn during HFNC-treatment by the responsible clinicians' discretion, data will be included in the study material. Patients will remain in study till it is decided by the treating physician to terminate HFNC-treatment. In case a clinical decision is made for the patient to continue home treatment with HFNC, the patient will remain in study till the decision is noted in the patient's case file.

For patients who desist HFNC-treatment after initiation, duration of use and, if possible, reason for desisting use, will be registered.

Patients who candidates for invasive ventilations will be excluded from the study if the arterial blood gasses further deteriorate after initiation of HFNC.

If the patients show clinical signs of deterioration at any time point during the study (i.e. cognitive deterioration, respiratory fatigue), they should be clinically evaluated immediately and an arterial blood gas should be part of this evaluation.

In patients who at inclusion are suspected of COPD, the diagnosis must be confirmed with spirometry and diagnostic work up according to national clinical guidelines within three months of hospital discharge.

When patients improve clinically, that is, normalization of pH, stabilization of oxygenation above target saturation and are allowed breaks from HFNC by clinical decision, they should be weaned.

Data registration From Airvo3: Airflow, FiO2, respiratory rate, oxygen saturation. Data will be downloaded from each of the devices at the end of every included patients' use of the device, to a subject-specific USB.

From patients' case records: Age, sex, time since COPD diagnosis, latest lung function within six months of inclusion in study (forced expiratory volume in the first second (FEV1), Forced vital capacity (FVC), (liters/%), as well as FEV1/FVC), body mass index, long term oxygen treatment prior to admission +/-, (if +, duration and prescribed oxygen flow (liters/minute)), home HFNC-treatment (if + duration, flow, FiO2), smoking status (current, former, quitter, never; if current duration and pack years, if former/quitter duration, time since termination and pack years). A pack year is defined as 20 cigarettes/day/year. In addition, number of moderate- and severe exacerbations within the past twelve months1. Patients' comorbidities are registered, according to the COPD-specific comorbidity test (COTE) index. Furthermore: Size of nasal cannula and time with HFNC-treatment is registered. Lastly, duration of admission (full days) and death during admission (+/-) will be registered.

From the Patient Prescription Registry data on inhaled medication at time of admission (long-acting beta-2-agonists/ long-acting muscarinic antagonists/inhaled corticosteroids +/- short-acting beta-2 agonists +/-) will be retrieved.

From laboratory files: From the ongoing admission: Data from arterial blood gasses as described above. C-reactive protein, white blood cell count, hemoglobin, erythrocyte middle cell volume (MCV) and erythrocyte middle cell hemoglobin concentration (MCHC). Laboratory files are, in all three regions, embedded in the electronic case files described above.

Data will be registered by the responsible research clinician at each site or by a study staff member, delegated by the study responsible physician at the respectable sites.

Biologic material Blood from arterial blood gasses will be used in this study. The blood tests performed during the study would also be carried out as standard procedures during treatment for acute hypercapnic failure. Should further blood gas analysis be carried out at the responsible clinicians' discretion during the treatment period. There is no biobank in this study.

Data protection and data registration Data from the AirvoTM3 will be downloaded from the device and transferred for data processing on "Fisher \& Paykel Healthcare"'s server in New Zealand. Processed datassets will thereafter be stored on a secured server at Aalborg University Hospital. All data will be registered in a database created in the Aalborg University Hospital Licensed Research Electronic Data Capture (RedCap) (Vanderbilt University, Nashville, Tennessee, USA). Data are kept on a separate, secured server. Data will not be shared with a third party and will remain within Denmark.

The study is approved by the Danish Ethical committee, the Data Protection Agency and registered in the North Jutland Regional Research database.

The study is investigator initiated and funded by an unrestricted grant from "Fisher \& Paykel Healthcare". The contract between Investigator and "Fisher \& Paykel Healthcare", including a data-management agreement, is approved by the Legal Unit at Aalborg University Hospital. Collaboration- and data management agreements with the adjacent sites are also approved by the Legal Unit at Aalborg University Hospital.

Sample size calculation Power calculations were based on results- and the power calculation carried out in the study by Cortegiani et al. This study anticipated a standard deviation for PaCO2 equal to 15 mmHg (1.99kPa) and a non-inferiority limit of 10 mmHg for PaCO2 (1.33 kPa). At confidence intervals of 95%, a margin of error of 5, and with inclusion of extra 10 patients per variable in the multiple regression analysis (PaCO2 baseline, flow, cannula size, time, and respiratory rate) a total of 138 patients, equally randomized to OptiflowTM /OptiflowTM+Duet nasal cannulas (69/69).

Statistical analysis plan

• Describing baseline information and investigating potential differences between the two groups of patients: Mean/standard deviations, median/interquartile ranges, dependent on normality, judged by Shapiro-Wilk's test.
• Describing adherence to HFNC treatment: Percent of total population
• To investigate changes in PaCO2: Multiple regression analysis, correcting for PaCO2 at baseline, HFNC-flow, cannula size, duration of HFNC-treatment, and respiratory rate.
• To investigate changes in pH: Unpaired t-test, with respect for deviation from normal distribution and thereof need of calculation of SE (boot-strap- or Mann-Whitney test if non-parametric p-value).
• To investigate pH-neutrality +/-: Fisher's exact test
• To investigate time to pH neutral: Cox regression analysis