Alternate Methodology of Pulse Oximeter Validation

Pulse oximeters are non-invasive devices that use light to display an estimate (denoted by SpO2) of the true arterial blood oxygen saturation (SaO2). The accuracy of the pulse oximeter is determined from desaturation sequences (reduction in SpO2 from 100 to 70%) performed in healthy volunteers. The current standard desaturation sequence is a step-wise reduction in 5% SpO2 increments, achieved by reduction in inspired oxygen delivered by a gas blender such as the ROBD (Reduced Oxygen Breathing Device). This mixes oxygen and nitrogen (components of air) to produce the desired gas mixture. Throughout the desaturation sequence arterial blood gas samples are drawn at each step. The blood gas samples are processed through a co-oximeter to measure the SaO2. For each SaO2 measurement there will be a corresponding SpO2 recorded. All data points from all subjects are amalgamated together to allow calculation of how well SpO2 corresponds to the SaO2 throughout the range of 70 - 100% oxygen saturation. Accuracy, A[RMS], is the unit of measure and this incorporates both bias and precision.

When the ROBD is used in a desaturation sequence, the subjects' arterial blood carbon dioxide (CO2) level will initially be normal (normocapnia). With the reduction in SpO2 during the desaturation sequence, the subjects' breathing rate will increase and this will cause the a physiological reduction in the blood CO2 value (hypocapnia). The extent of hypocapnia will vary from one subject to another. It is not known whether the presence of hypocapnia may have an effect upon the accuracy measurement.

The RespirAct is a gas blender which, in contrast to the ROBD, can control both the oxygen level as well as the carbon dioxide level. The end-tidal oxygen level (PAO2) is continuously displayed by the RespirAct during the desaturation sequence. The PAO2 closely approximates to arterial blood oxygen (PaO2) - the degree of closeness will be measured in the study. Through an accepted formula, the SaO2 can be derived from the measured PAO2. The derived SaO2 values can subsequently be used to measure the accuracy of the pulse oximeter. This A[RMS] can then be compared with the A[RMS] calculated from the blood gas co-oximeter PaO2 / SaO2 values. If the accuracies are comparable then this would remove the need for arterial line placement during pulse oximeter validation studies.

Eligible healthy volunteers will complete five (5) oxygen desaturation sequences with adjustment of sequence type and/or CO2 level. Two sequences will be delivered by step-wise reduction and two will be delivered by gradual slope. Two sequences will be delivered at normocapnia and two will be delivered at hypocapnia values. The ROBD sequence will be a step-wise reduction with CO2 value set by the subjects' own breathing control.