UMOX™: A New Device for Preoxygenation

This study was as a single center, crossover, randomized trial. Following approval from research and ethics committee from our hospital and written consent from healthy subjects recruited, 50 consenting volunteers were enrolled over a three-month period, from: September 29, 2008 to November 5, 2008.

Subjects enrolled were of both genders, smokers and non-smokers, American Society of Anesthesiologists (ASA) physical status I or II, 18 to 65 years old. Nonrecruitment depended on the presence of any craniofacial abnormality, facial hair (beard or moustache) or absence of teeth. Well-controlled non-severe asthma was not a criterion for exclusion.

The trial was conducted in an operating room, using the Dräger Fabius GS (Dräger, Canada), anesthesia machine in a usual set up with the operating table. The investigators used a, patient circuit system consists of 1.5-L lime soda absorber connected to disposable corrugated tubing and a 2-L breathing bag. Gases were analyzed through a Marquette Medical Systems™ SAM module (GE Healthcare), volumes monitored with a GE Respirator Mechanics-M (GE Healthcare) and SpO2 measured with a Nellcor™ (Covidien) pulse oximetry positioned on a finger. Preoxygenation procedures were explained to the subjects and ample time was allowed so they could become familiar with the mask and mouth piece. Set up for the measure was conducted by the same experimenter. Subjects laid supine and preoxygenation was performed with fresh gas flow: 100% oxygen (O2) 10L/minute, administered trough facemask or mouth piece sequentially in a random order. Randomization was established with a random assignment table (ww.randomization.com). The envelope was open as the subject entered the operating room (OR). At the beginning and between each step, the anesthesia circuit was flushed with 100% (O2). Each subject went through all steps in random order, each requiring a 5-min period of: 1) tidal breathing using the facemask (TVFM); 2) tidal volume breathing with the UMOXTM mouthpiece (TVMP); 3) tidal volume breathing with the UMOXTM mouthpiece and verbal instructions to favor mouth breathing (TVMPMB); (4) tidal volume breathing with the UMOXTM mouthpiece and a nose clip (TVMPNC). The experimenter conducting the steps instructed participants to create a seal with their lips around the mouthpiece and readjusted the facemask when the subject signalled a discernable leak. Each step was interrupted by 10 minutes of breathing normal room air to allow for return to original respiratory status. Continuous recording of experimental measures, displayed on the anesthesia monitoring system, was done on digital films (Canon PowershotS5 IS™). For the purpose of our trial, 10 seconds intervals were considered adequate for following preoxygenation trends. End-tidal O2 was measured continuously for 5 min. We have also evaluated the level of satisfaction of the subject with each technique using a 1-4 scale.Participants were asked to asses their comfort by showing the associated number with their fingers. Finally we described the performance of the facemask ventilation technique as reported by the experimenter on a scale of 1 to 4.Primary outcome was the end-tidal O2 at the end of the 5-minutes preoxygenation period. Secondary outcomes included endtidal O2 as a function of time, the number of subjects reaching at least 87% EtO2, subject satisfaction and ease of ventilation.

Based on the few studies available on different preoxygenation technques, the end-tidal oxygen fraction standard deviation can range from 5% to about 20% depending on the device used for preoxygenation3,4,14,15. Assuming a 12% value for standard deviation, sample size calculations were made for Type I and Type II errors of 5% and 20%, respectively (power of 80%). From these numbers and taking account that we would rely on multiple comparisons, the power analysis suggested that 50 subjects would allow us to find a difference between groups around 9%. The 9nvestigators analyzed the EtO2 final value with one-way ANOVA followed by the Tukey's Multiple Comparisons test. Percentage of participants reaching at least 87% EtO2 are described with survival curves compared with Kaplan-Meier log rank test. Tolerance is presented using Kruskal-Wallis with Dunn's Multiple Comparisons test.