This was a single-center, parallel-group, prospective, double-blind, randomized clinical trial. The study was conducted from April 2022 to July 2022 at the Tibet Autonomous Region People’s Hospital, Lhasa, Tibet, China (3650 m above sea level). This hospital is a Class A tertiary comprehensive hospital and referral center in the Tibet Autonomous Region.

This study was registered at ClinicalTrials.gov (NCT05304923) before enrollment on 31/03/2022. Ethical approval for this study (approval number: ME-TBHP-22-02) was provided by the institutional review board of Tibet Autonomous Region People’s Hospital, Lhasa, Tibet, China (Chairperson Prof Ciren Dawa) on 21 April 2022. Written informed consent was obtained from all the participants. All personnel on the study team completed standardized training for procedural sedation and the use of SJOV. A separate investigator monitored the safety of the participants and recorded their outcomes. An independent monitoring group saved and analyzed the data as planned. This trial followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guidelines.

The patients were recruited from the Tibet Autonomous Region People’s Hospital. Patients could be eligible for the trial if they fulfilled the following inclusion criteria: a. 18 years or older; b. underwent routine gastrointestinal endoscopy under procedural sedation; c. consented to participate in this trial. The exclusion criteria included infection of the upper airway; anatomical abnormalities of the face, nose, and upper airway; coagulopathies; anticipated or known difficult airway; known allergy to propofol, soybeans, and egg; and absence from the high-altitude environment during the past 3 months. As per the instructions of the center, patients with severe disease will not be allowed to accept procedural sedation; thus, patients with severe pulmonary disease were also excluded from this study.

The participants were randomly allocated to either SJOV or nasal cannula oxygen supply in a 1:1 ratio using block randomization with variable block sizes of four or six randomized. Randomization was stratified according to whether upper endoscopy or colonoscopy (1:1) was planned. An independent research assistant generated randomization using a computer-generated list and assigned allocations. The group assignments were concealed from the investigators who enrolled the participants until after enrollment using sequentially numbered opaque envelopes. Participants and statisticians who assessed the outcomes were blinded. Since the clinicians and investigators had to conduct sedation and record the outcomes throughout the procedures, they could not be blinded.

In all patients, a loading dose of propofol (1 mg/kg) was administered, followed by an incremental bolus (10 mg) to titrate to a deep sedation level. To control for bias, no other sedation techniques were used in this study. Deep sedation was defined as a purposeful response after repeated or painful stimulation [1] and was approximately equivalent to the Observer’s Assessment of Alertness/Sedation (OAA/S) scale of 1 or 2 (1–5, 1 = deep sleep, 5 = alert) [9, 10]. Sedation was performed by the same anesthesiologist who had completed standardized training for procedural sedation and the use of SJOV. Another member of the study team supervised all the procedures. The OAA/S scale was administered every 2 minutes, and incremental boluses were titrated to maintain deep sedation (OAA/S scale< 3). Before sedation, each patient was oxygenated via a nasal cannula for two minutes to ensure that the peripheral oxygen saturation (SPO2) was ≥95%. No premedication was used in this study. After sedation, patients were positioned in the left lateral recumbent.

In the nasal cannula oxygen supply group, oxygen supplementation at 2 l/min was delivered via a nasal cannula. In the SJOV group, SJOV was conducted using a Wei nasal jet tube (WNJ, Well Lead Medical Co. Ltd., Guangzhou, China), which was connected to a manual jet ventilator (Well Lead Medical Co. Ltd., Guangzhou, China) via its jet port (Fig. 1) [8, 11]. The initial settings of SJOV were as follows: driving pressure (DP) 15 psi; respiratory rate (RR) 20 bpm; inspiratory-to-expiratory (I/E) ratio 1:2, and gas supply, 100% oxygen [8]. The WNJ was inserted through the nostril toward the vocal cords, and chest excursion was visible when the ideal placement was achieved, albeit not necessarily [8].

Supraglottic jet oxygenation and ventilation using a Wei nasal jet tube . SJOV was conducted using a WNJ connected to a manual jet ventilator via its jet port. The WNJ is a regular nasopharyngeal airway with two modified nozzles. One is the jet nozzle connected to the manual jet ventilator. Another nozzle is used to monitor breathing connected to a PaCO2 monitor (breathing was not monitored in this study). The regular nasal cannula could also supply up to 6 l/minute of oxygen. Only the jet nozzle was used for oxygen administration in this study. SJOV, supraglottic jet oxygenation and ventilation; WNJ, Wei nasal jet tube

After the target sedation level was reached, the WNJ was placed in lieu of the nasal cannula to initiate SJVO in the SJOV group. Before the patients recovered from sedation, the WNJ was replaced with a nasal cannula to ensure that the patients were blinded to the assignment.

When SPO2 was < 95%, increasing oxygen supplementation was performed first. In the nasal cannula oxygen supply group, the oxygen flow rate was increased to 6 l/min. In the SJOV group, the RR was adjusted to 30 bpm, and the DP was increased to 20 psi. If SPO2 fell below 90%, the airway was opened with a jaw-thrust maneuver in both groups [12]. If hypoxemia could not be corrected within 1 min or SPO2 was < 75%, mask ventilation was considered in both groups [8, 9]. Tracheal intubation can be performed, if necessary. No other additional airway devices were used in this study.

Blood pressure, heart rate, and pulse oximetry were recorded before oxygen supplementation as the baseline and repeated every 2 min. Hemodynamic management and administration of additional analgesics (alfentanil in this study) were performed at the discretion of the anesthesiologists based on the patients’ clinical signs. The total propofol dose, total alfentanil dose, procedure duration, and hypoxia-related interventions were recorded. Patients in both groups were monitored for 20 min after awakening.

The primary outcome was moderate hypoxia during sedation, defined as an SPO2 of 75–89% for < 60 s [8]. The secondary outcomes were adverse events except for moderate hypoxia, which included respiratory-related complications including pulmonary aspiration, mild hypoxia (SPO2 = 90–95%) [8], and severe hypoxia (SPO2 < 75% or < 90% for > 60 s) [8]; cardiovascular-related complications including hypotension (systolic blood pressure < 90 mmHg), hypertension (systolic blood pressure > 160 mmHg), bradycardia (heart rate < 50 beats/min), tachycardia (heart rate > 120 beats/min); and fatal complications including severe anaphylactic reactions, myocardial infarction, cardiac arrest, and death [8, 13, 14]. Adverse events related to SJOV were also recorded 20 min after the patient awoke, which comprised pharyngalgia, xerostomia, nasal bleeding, and barotrauma.

The sample size was estimated based on the primary outcome of hypoxia. There have been no reports of hypoxemia during deep sedation during gastrointestinal endoscopy at high altitudes. In a study on hypoxemia during moderate sedation for gastrointestinal endoscopy at low altitudes [8], the incidences of hypoxia were 9% when a nasal cannula was used to supply oxygen and 3% when SJOV was used. The incidence of sedation-related hypoxia in gastrointestinal endoscopy varies considerably [8, 12, 14, 15]. Given the deeper sedation level and higher altitude, we assumed that 30% of patients in the nasal cannula oxygen supply group might encounter hypoxia, which was still within the reported range [12, 14] and consistent with our unpublished preliminary data. Considering the ventilation competence of SJOV [5, 11, 16], the incidence of hypoxia might not deviate from the reported rate (3%). We assumed an incidence of 5% in the SJOV group, which corresponded to an absolute difference of 25% between the groups. With these estimates, a 2-sided α level of 5%, and an attrition rate of 10%, a total of 36 patients per group (72 in total) were required to have 80% power to show a significant difference between groups.

The patients were analyzed according to the randomization group. Multiple imputations were planned for missing data. Continuous variables were expressed as the median with interquartile range (IQR) or mean with standard deviation (SD), as appropriate. Categorical data are presented as numbers and proportions and were compared using the chi-square test or Fisher’s exact test, as appropriate. A 2-sided P value of less than 0.05 was considered statistically significant. A modified Poisson regression model was used to adjust the primary outcome for the stratification variable (upper endoscopy or colonoscopy) and baseline characteristics [age, sex, ethnic group, BMI (body mass index), history of OSAHS (obstructive sleep apnea-hypopnea syndrome), erythrocytosis, original hypoxia before oxygenation, procedure time, total propofol dose, and total alfentanil dose]. Subgroup analyses were not conducted because of the limited sample size. Sensitivity analyses were established a priori and performed using alternate definitions of the primary outcome. The alternate definitions (Supplementary Table 1) comprised mild to moderate hypoxia (SPO2 of 90–95% or 75–89% for < 60 s), moderate to severe hypoxia (SPO2 < 89%), and mild to severe hypoxia (SPO2 < 95%). Because of the potential for type I errors due to multiple comparisons, findings for analyses of secondary outcomes should be interpreted as exploratory. All statistical analyses were performed using IBM SPSS Statistics (version 25.0; IBM Corp., Armonk, NY, USA).