The intent of this study is to determine the difference in pharyngeal oxygen concentration in patients who have a natural airway (not intubated) using commonly available oxygen delivery systems.

The investigators will test the hypothesis that oxygen concentration during the period of inspiration (FiO2) in the pharynx is dependent on oxygen delivery system design, even at high flow (15 liters/minute) oxygen delivery. Specific measurements include oxygen concentration at subjects' lips and pharynx when breathing 100% oxygen and room air via a simple mask, non-rebreather mask, OxyMaskTM, and anesthesia mask with headstrap and Jacson Rees circuit.

A mean difference of 10% pharyngeal FiO2 between any of the masks will be considered clinically important. The expected standard deviation of the within-subject FiO2 is 3.5%. With a significance criterion of 0.05, 10 subjects would provide more than 90% power to detect a mean difference of 10%.

Oxygen therapy refers to the administration of oxygen at higher concentrations than room air. According to the most recent American Association for Respiratory Care Clinical Practice Guidelines, indications for supplemental oxygen include hypoxemia, severe trauma, acute myocardial infarction, short term therapy, and postoperative recovery. Though there are few contraindications to oxygen therapy, potential complications exist, including ventilatory depression in hypercarbic patients, absorption atelectasis, and oxygen toxicity (particularly in some patients receiving specific chemotherapeutic agents). In the spontaneously breathing hypoxemic patient without an endotracheal tube, it is important to be able to deliver high concentrations of oxygen in order to either avoid intubation or bridge to intubation. A number of oxygen delivery devices are available that have different properties that make them useful for different situations.

Variable performance (low-flow) systems: Low-flow devices are variable performance because the delivered flow rates are less than the patient's peak inspiratory flow rate (PIFR) and the oxygen is diluted by room air entrained from around the devices. Higher PIFR (e.g. respiratory distress) results in a decrease in the FiO2 delivered through a variable performance device. Rebreathing of carbon dioxide is also possible with low-flow masks at lower oxygen flow rates. Variable performance systems include the nasal cannula, simple mask, partial rebreathing mask, and non-rebreathing mask.

Nasal cannulas are commonly used for stable patients because they are more comfortable, less irritating to skin, and less anxiety-inducing than masks; they also allow the patient to eat, talk, and use incentive spirometry. Nasal cannula can provide 24-40% FiO2 with flow-rates up to 6 L/min. The simple mask is commonly used in the immediate postoperative setting for supplemental oxygen. Several small holes on either side of the mask as well as the imperfect seal allow room air to be entrained and exhaled air to escape from the mask. Simple masks can provide 35-50% FiO2 at flow-rates 5-10 L/min.1 Long-term use of masks can lead to skin irritation and pressure sores.The partial rebreather consists of a simple mask with a reservoir bag, and provides 40-70% FiO2 at flow-rates 6-10 L/min.1 The partial rebreather is no longer used at many hospitals because the range of oxygen delivery can be encompassed by the simple mask and the non-rebreather. The non-rebreathing mask is similar to the partial rebreather but includes one-way valves that prevent exhaled air from returning to the reservoir bag. The non-rebreather can deliver 60-80% FiO2 at 10-15 L/min.1

Fixed performance (high-flow) systems: By delivering oxygen at a flow rate greater than PIFR, fixed performance systems make it possible to provide a specified FiO2 throughout the respiratory cycle. The two commonly used devices are the Venturi mask and OxyMaskTM. Venturi masks use the Bernoulli principle to deliver oxygen at high flows through a Venturi valve that has a narrow constriction followed by a wider area with vents that entrain room air. Because of the relative ease of use and consistent oxygen delivery, they are useful for patients with COPD or others with chronic hypoxemia. The Venturi mask is able to deliver an FiO2 of 24-50%.1 Venturi masks will not be tested in this study.

The OxyMaskTM is a relatively new device that uses the Venturi diffuser with a five-pronged open mask design, which allows the patient to talk and may cause less claustrophobia than the more closed masks. OxyMaskTM is able to deliver 25-80% FiO2 at flow rates of 1.5-15 L/min. A closed mask with a perfect seal could theoretically deliver 100% FiO2. For the purposes of this study, a Jackson Rees circuit will be used to deliver oxygen to an anesthesia mask. The Jackson Rees circuit (also called the Mapleson F circuit) is routinely used during transport to deliver oxygen from the oxygen tank to intubated patients. This circuit consists of tubing connected to a manual ventilation bag that can be squeezed to deliver pressurized gas flow to the patient. The bag is equipped with an adjustable pressure limiting valve, which allows the provider to control the pressure delivered by squeezing the bag.

Measurement of FiO2: It has been established that FiO2 achieved with nasal cannula does not differ significantly whether the subject breathes through their mouth or nose. This property has not been established for masks. With a closed mask, FiO2 achieved would also be the same regardless of mouth or nose-breathing. With an open mask, the assumption must be made that oxygen concentration is equal at the nose and the mouth in order to ensure equivalence of mouth and nose-breathing. Also, previous studies on oxygen delivery systems have used a variety of methods for measuring FiO2. From a review of the literature of the oxygen delivery devices discussed above, only the nasal cannula has been studied by measurement of oxygen at the level of the pharynx, whereas the other masks have been studied by measurement of oxygen at the lips.

Study Rationale: The studies on OxyMaskTM determined FiO2 based on gas sampled at the lips. However, due to the possibility of gas mixing and/or nasal breathing with the open mask design, it is unclear if the gas sampled at the lips is the same gas that arrives at the trachea, which is the most clinically relevant location. The gas inspired from the trachea is what is actually seen by the alveoli and participates in oxygenation of the blood. Gas mixing would seem more likely with an open mask design such as the OxyMask. A study comparing pharyngeal FiO2 between masks would more accurately demonstrate the effective FiO2 than the currently available data. This in turn, will better inform perioperative clinicians of the best approach to maximize oxygen delivery to our critically ill post-operative patients who require supplemental oxygen to treat or prevent systemic hypoxia.

Study Design: The investigators will measure steady state FiO2 in the pharynx and at the lips simultaneously on each subject as they undergo each of the following oxygen delivery conditions sequentially: 1) room air; 2) high flow oxygen and simple mask; 3) high flow oxygen and non-rebreather mask; 4) high flow oxygen and OxyMask; and 5) high flow oxygen and anesthesia mask with head-strap and Jackson Rees circuit.

Study Population: The study will include up to 20 subjects. A mean difference of 10% pharyngeal FiO2 between any of the masks will be considered clinically important. The expected standard deviation of the within-subject FiO2 is 3.5%. With a significance criterion of 0.05, 10 subjects would provide more than 90% power to detect a mean difference of 10%.

Setting: The study will take place in a fully equipped operating room with an anesthesia machine that has been approved for patient use by Infection Control and Biomedical Engineering. It is located in the South OR area of Kohler Pavilion at OHSU and is equipped with the Datex Ohmeda anesthesia machine and Poet Gas Analyzer that will be used for the study. The study will be conducted after hours to avoid any conflict with surgical patient care. The oxygen sensors will be calibrated before each subject is tested.

Recruitment: Subjects will be recruited from OHSU's perioperative service. Advertisement flyers will be distributed in each operating room location.

Consent: The consent process will be undertaken by individuals with appropriate human subjects protection and HIPPA education, and performed in person on the OHSU campus or on the phone. Subjects may withdraw from the study at any time.

Procedures: The subjects will each participate on one study day and will require one hour per subject. Subjects will remain NPO for at least 6 hours prior to their study period. Subject characteristics will be recorded. The anesthesia machine will be turned on and the oxygen sensor on the gas analyzer will be calibrated. Baseline vital signs (HR, BP, SpO2) will be measured. FEV1 and FVC will be measured with the anesthesia machine and a standard circuit, using the Y-piece as the mouthpiece. The following protocol for gas sampling will be used: One gas sampling line will be attached at one end to the gas analyzer and at the other end to an 8 French suction catheter, which will be lubricated with 2% lidocaine jelly. This nasal catheter will be inserted through a nare, with the tip position immediately behind the uvula. Placement will be confirmed by visualizing the tip of the catheter using a penlight and having the subject open the mouth widely and say, "Ah." If required, a tongue depressor will be used. If the catheter cannot be visualized, then it will be inserted to a standard depth of 9-10cm. The nasal catheter will be taped to the subject's face. A second gas sampling line will be taped to the subject's face and positioned so that sampling occurs at the patient's lips. After placement of the two sampling catheters is confirmed, the subject will breathe room air normally for 90 seconds and FiO2 will be measured over a period of 30 seconds. The subject will not be instructed whether to use their nose or mouth to breathe, since patients in the hospital typically not instructed which way to breathe. Afterward, testing of the various oxygen masks will be performed with oxygen set at high flow (15L/min) through simple mask, non-rebreather mask, OxyMask, and anesthesia mask with headstrap and Jackson Rees circuit. Since the auxiliary oxygen dial on the anesthesia machine only goes up to 10L/min, a 15L/min adjustable oxygen valve will be attached directly to the main oxygen supply line in the operating room. After placement of each mask and starting oxygen high flow, the subject will breathe normally for 90 seconds and then FiO2 will be measured over the next 30 seconds, similarly to the room air control. At the end of each trial period, each subject will be asked to take a single vital capacity breath (starting with maximum exhalation and followed by maximum inhalation). Between testing of each mask, there will be a five minute period of breathing room air as a washout period and to confirm stability of hemodynamic status (measurement of BP and HR). Demographic data to be obtained includes height (cm), weight (kg), age, gender, and self-declared ethnicity. Blood pressure, heart rate, and oxygen saturation will be recorded at baseline and prior to and following each mask-trial. Baseline FEV1/FVC measurement will be conducted on all patients prior to initiating the study and recorded. FiO2 data will be downloaded digitally from the gas analyzer to an excel spreadsheet. If this is not possible, then oxygen percentage will be manually recorded during each breath to determine the range of oxygen percentages (minimal and maximal) during both phases of ventilation, during the 30 second measurement period. Respiratory rate during measurement time will also be recorded. At the end of the 30 second sampling period, a final measurement will be taken during a forced vital capacity breath.

Data Analysis: Phase of ventilation will be defined using the capnography tracing, which is displayed concurrently with measurement of inspired oxygen. Maximal and minimal oxygen concentration will be averaged for inspiration, during the 30 second mask breathing trial, for each subject. These values will be averaged across the group for each intervention to define statistics for the entire group. Values recorded during normal tidal volume ventilation will be analyzed separately from values recorded during each forced vital capacity ventilation. The Analysis of Variance test and appropriate post-hoc analysis will be used to determine the difference between FiO2 administered by each mask tested.

Inclusion Criteria:

• 18 to 70 years of age
• Male and female volunteers
• ASA physical status I, II and III
• Capable and willing to provide written informed consent in English

Exclusion Criteria:

• Acute cardiopulmonary disease, as defined by blood pressure greater than 150/90, HR greater than 120 and room air oxygen saturation less than 92.
• Allergy to lidocaine or adhesive tape
• History or physical exam finding of nasal polyps
• Currently taking oral or parenteral anticoagulant medications (other than aspirin)
• History of frequent nose bleeds
• Current symptoms of nasal congestion
• Physical examination findings of rales or wheezing
• Facial hair that prevents forming a seal with an anesthesia mask

Subjects who meet the inclusion criteria and none of the exclusion criteria who are to undergo a procedure not requiring general anesthesia.

All subjects consented