This is a four-center randomized controlled trial, which has been approved by the Ethic Committees of East Hospital (Shanghai, China) and Changhai Hospital (Shanghai, China). This trial is registered with Chictr.org.cn (ChiCTR2100055038). The report of the protocol followed the SPIRIT guideline [43].
Study population
Patients who need bronchoscopy examination will be recruited from outpatient clinics or inpatient units in the four hospitals.
Inclusion criteria
Adult patients who have one of the following high-risk factors: 1) hypoxemia with the PaO2/FIO2 ≤ 300 mmHg before bronchoscopy examination; 2) hypercapnia with PaCO2 ≥ 45 mmHg and baseline chronic pulmonary disease, including COPD, and bronchiectasis, etc.; 3) narrow trachea confirmed by radiology before bronchoscopy, which may have been caused by tracheomalacia, trachea tumor or granuloma, or aspirated foreign body, etc.; 4) morbidly obese patients with body mass index (BMI) ≥ 30 kg/m2.
Exclusion criteria
Patients will be excluded if any of the following criteria is met: 1) age ≥ 90 years; 2) pregnancy; 3) reluctance to participate; 4) estimated duration of bronchoscopy is less than 10 minutes; 5) contraindication to using HFNC, including nasopharyngeal obstruction and blockage.
Study procedures
See study diagram in Fig. 1.
Recruiting and consent process
The nurse responsible for scheduling bronchoscopy examinations for patients in the bronchoscopy suite will screen the patients. The study investigator will be notified of eligible patients and will introduce this study to the patients. Study purposes, procedures, risks, and benefits will be discussed with the patients in a private room. Consent forms will be given to patients for full consideration until the examination day. On the examination day, arterial blood gas will be assessed for patients who were enrolled in the study due to hypoxemia and/or hypercapnia on the screen day. If the patients still meet the inclusion criteria and sign the consent form, they will be enrolled for randomization. The other two types of patients (narrow airway and morbidly obese) will be randomized to receive HFNC or COT during bronchoscopy examination after signing the consent form in the presence of the study investigator.
Randomization and treatment
An independent statistician will be responsible for generating the randomization sequence, which will be placed in a series of sequentially numbered, sealed, and opaque envelopes. After the patient signs the consent form, an envelope will be opened to assign an oxygen device for the patient.
OH-70C (micomme, HuNan, China) will be used to provide HFNC. The initial flow will be set at 30 L/min, and the flow will be titrated up by 5 L/min until the patient could not tolerate it. FIO2 will be set at 0.45. The temperature will be set at 37° C. Nasal cannula will be chosen based on the patient’s prong size: the nasal cannula size should be less than 50% of the nasal prong size. As the bronchoscope will be inserted via one nasal prong, one side of the nasal cannula will be occluded and left outside of the nose while the other will be placed in one nasal prong (Fig. 2). Similarly, in the group with COT, a single-prong nasal cannula (Healthcare Medical Supplies Co., Ltd. Hangzhou, China) will be chosen, and oxygen flow will be set at 6 L/min. Pre-oxygenation will last for five minutes. Patient comfort, SpO2, PtCO2, and vital signs including heart rate, respiratory rates, and blood pressure before and after pre-oxygenation will be recorded (Fig. 1).
Bronchoscopy examination and hypoxemia management
Topical anesthesia will be provided prior to bronchoscopy as routine treatment. Sedation and/or anesthesia will be utilized by an anesthesiologist at the bedside. The timing to use, the name, and the dose of sedation will be recorded. Bronchoscopy examination will be performed by experienced pulmonologists in the bronchoscopy suite. SpO2, PtCO2, and vital signs will be continuously monitored while FIO2 for HFNC and oxygen flow for COT will be maintained until the patients experience the first desaturation, defined as SpO2 ≤ 90% for 10 seconds. The timing and the lowest SpO2 in the first desaturation will be recorded (Fig. 1). To improve oxygenation, FIO2 will be increased to 1.0 in the HFNC group while in the COT group, COT will be changed to HFNC with FIO2 at 1.0 and flow at the maximal level the patient can tolerate. This management will be recorded as treatment escalation. If SpO2 is still beyond the safe range, bronchoscopy will be withdrawn until SpO2 recovers (Fig. 1). The process and the reason for the suspension will be recorded. After the recovery, whether bronchoscopy will be resumed or not will be determined by pulmonologists based on the clinical situation. The length of the bronchoscopy examination will be recorded.
Termination criteria
Bronchoscopy examination will be terminated if any of the criteria is met: 1) any of the adverse events including hemodynamic instability, cardiac arrest, massive bleeding, and pneumothorax is observed or reported; 2) with maximal oxygen delivery, SpO2 cannot be maintained at a level ≥ 90%.
All the rescue treatments and resuscitation equipment are prepared in the bronchoscopy suite. Any patient who experiences adverse events will be rested and monitored in the general unit, until the patient recovers.
Post Bronchoscopy monitoring
After bronchoscopy, SpO2, PtCO2, and vital signs will be continuously monitored until the patients return to pre-bronchoscopy status or are stable. The recovery time will be recorded as well. After the patients recover, patient comfort will be questioned.
Outcomes
The primary outcome is to compare the incidence of desaturation which is defined as SpO2 < 90% between patients using HFNC and those using COT. The secondary outcomes include the timing of the appearance of the first desaturation (named desaturation time), the lowest SpO2and the highest PtCO2 during the entire bronchoscopy examination, the need for treatment escalation, the number of bronchoscopy interruptions, and the recovery time between the two groups.
Sample size calculation (https://clincalc.com/stats/samplesize.aspx)
This study is a superiority study. The incidence of desaturation during endoscopy examination was reported to vary from 5.88–43.2% using HFNC in contrast to 28.8–89.7% using COT in the seven previously published RCTs [22–26, 29, 30]. While in the study [26] with a patient population close to that of our study during bronchoscopy examination, the incidence of desaturation of HFNC vs COT group was 13.3 vs 33.3%, respectively. However, this previous study only included hypoxemic patients while we will enroll other types of patients, so the incidence of hypoxemia in the HFNC group in that study might be higher than the figure in ours. Thus, we assumed that when the maximal concentration of oxygen is provided by the two assigned devices, that is, when FIO2 is set at 1.0 for HFNC and oxygen flow at 6 L/min for COT, the probability of desaturation is 10% in the HFNC group and 30% in the COT group, respectively. According to statistical calculation, the sample size should be 124, with confidence level (α) of 95%, power (1-ß) of 80% and margin (Δ) of 0.2.
Data collection
The parameters that are continuously monitored by the bedside monitor before, during, and after bronchoscopy examination, including vital signs, SpO2, and PtCO2, and the incidence of desaturation and bronchoscopy suspension/termination will be observed and recorded by an independent investigator. Demographic information of patients (age, gender, height, weight, body mass index [BMI], race, smoking history, diagnosis, and reason for bronchoscopy examination) will also be collected.
Statistical analysis
Categorical variables, including the incidence of desaturation, treatment escalation, suspension or termination of bronchoscopy, etc., will be presented as percentages and analyzed by Chi-square test. Continuous variables, including the desaturation time, recovery time, the lowest SpO2, the highest PtCO2, etc., will be presented as mean ± standard derivation (SD) or median and interquartile rage (IQR) and compared by independent t-test or Mann Whitney test, depending on the normality of distribution, which will be tested by the Kolmogorov-Smirnov test. A two-sided P-value of < 0.05 will be regarded as statistically significant for all tests. Data analysis will be conducted with SPSS software (SPSS 23.0; Chicago, IL).
Patient and public involvement
Patients and/or the public were not involved in the study design.