Measuring the effects of hypoxia training on diver performance

The study will take place in-person at the Exercise Metabolism Laboratory at the University of Auckland. A two-part, randomized, single blind study will be conducted. There are two trial arms requiring 20 participants each (n = 40). Unblinded participants will be randomized 1:1 to the hypoxia exposure or no exposure arms. At a later date, blinded participants from both groups will be exposed to a “test event” involving exposure to hypoxia and asked to perform a simulated self-rescue task if hypoxic symptoms are perceived. Objective: Examine the role of a prior hypoxic exposure on symptom recognition and self-rescue during a second, later hypoxic exposure. Hypoxia exposure arm: Participants (n = 20) randomized to hypoxia exposure will participate in a single, individual, open-label hypoxia exposure session of 1 hours duration. A commercially available rebreather will be used. The oxygen controller will be set to maintain a partial pressure of inspired oxygen at 0.21 atmospheres (equivalent to room air). At the appropriate time in the experimental protocol the oxygen supply will be switched off to mimic a gradual change into hypoxia that can occur in diving. For example, this mimics the unfortunately common scenario where a diver enters the water breathing air, then initiates a dive after forgetting to switch on the oxygen supply to their rebreather. As several liters of breathable gas remain in the rebreather circuit, it takes several minutes for the diver to metabolize the oxygen content of this residual gas. This leads to a gradual change into hypoxia that will be mimicked in our protocol. Participants will be performing a card recognition task, similar to those regularly conducted during commercial and military aviation hypoxia awareness training. Participants will be asked to verbalize the number and suit of the card within four seconds, when a new card will appear. Participants will be familiarized with the test under normoxic conditions to confirm 100% accuracy and test re-test reliability. Participants will be seated and instrumented with a 3-lead electrocardiogram and finger pulse oximeter. End tidal PCO2 (ETCO2) and inspired PO2 and PCO2 will be continuously measured by sampling from the breathing circuit throughout the exposure. A portable EEG system will be applied. A functional near infrared spectroscopy (fNIRS) system will be combined with the EEG electrodes using an EEG-compatible fNIRS cap. A non-invasive blood pressure finger cuff will be used to continuously measure beat-to-beat blood pressure. In those participants who separately consent, an arterial cannula will be placed in the radial artery of the non-dominant wrist. An arterial blood gas sample will be taken at the start of the baseline card-recognition task. An additional blood gas sample will be taken when the hypoxic exposure is terminated. The drawing of each blood sample will not interfere with the participant’s ability to carry out the card-recognition task. All arterial blood gas samples will be run through a portable iSTAT arterial blood gas analyser which will be in the laboratory during the experiment. This will provide immediate results of PaO2, PaCO2 and oxygen saturation for all blood samples. The arterial line and blood sample collection will be performed by an anaesthetic registrar or consultant anaesthetist. The session will begin with the participant breathing from the rebreather loop with a constant oxygen concentration and performing a two-minute baseline card-recognition task. The oxygen supply will then be switched off to create a hypoxic mixture. The exposure will be terminated on committal of the second card recognition error after the peripheral oxygen saturation falls below 60%. Alternatively (for safety reasons) the exposure will be terminated if the participant commits three errors prior to reaching a peripheral oxygen saturation of 60%. Five minutes after the exposure, the participant will be asked to recall the total number of errors he or she made and to rate the severity of potential hypoxia symptoms on a visual analogue scale. Test event: The "test event" will take place approximately five weeks after the initial hypoxic exposure. Participants in both original groups (no exposure versus hypoxic exposure) will be exposed to the hypoxia test event. Participants will be informed that there is a 50% chance of being randomized to breathe hypoxic gas, and a 50% chance of being randomized to breathe air, however, participants will also be informed that their allocations will be concealed until after the final measurement to avoid expectation bias. The research team will debrief all participants and reveal their allocations prior to concluding the study. The set-up will be similar to that of the initial hypoxia exposure, including the use of EEG and fNIRS. Additionally, a virtual reality headset with embedded eye tracking technology will be employed. Participants will view a video of an underwater swim and will be asked to perform a virtual fish survey, similar to a scientific diving objective, to identify a specific fish species and to count these fish in a realistic and immersive underwater environment. This is intentionally a different cognitive task to that undertaken by those experiencing the initial hypoxia exposure in the first allocation to avoid confounding due to practice effect. Prior to beginning the event, participants will be instructed to perform "self-rescue" if any impairment is perceived. Self-rescue will involve operation of the mouth-piece bailout valve; a task which involves both hands and a certain degree of manual dexterity. Participants will be allowed to become immersed in the task, after which they will either be switched to hypoxic gas breathing or continue breathing a normoxic gas. Participants will be videoed, and accuracy of the distractor task and time to self-rescue initiation after the switch to hypoxic gas breathing will be recorded. If a participant ceases recording fish counts or reaches a peripheral oxygen saturation of less than 60% without bailing out, the participant will be prompted to bail-out with a visual stimulus inside the VR environment, if no response occurs the event will be terminated. Five minutes after the exposure, participants will be asked to rate the severity of potential hypoxia symptoms on a visual analog scale. Both the initial hypoxia exposure and the later test event are recorded on video using two cameras. One records the participant and their interaction with the distractor task (card recognition or simulated dive), while the other records the entire experimental setup and the actions of the investigators. Fidelity to the planned intervention will be assessed by auditing the recordings for errors.