Pulse Oximeter Accuracy in Healthy Humans During Hypoxia

This study is to determine the accuracy of pulse oximeters, which measure blood oxygen by shining light through fingers, ears or other parts of the body, without requiring blood sampling. The aim of this project is to test the accuracy of pulse oximeters during mild, moderate and severe hypoxia. This is done by comparing the reading of the pulse oximeter during brief, steady state hypoxia with a gold-standard measurement of blood oxyhemoglobin saturation (arterial blood sample processed in a laboratory hemoximeter).The data obtained is submitted by pulse oximeter manufacturers to the FDA for device approval. Studies normally involve 10-14 normal adult volunteer paid subjects, with 20-25 1-2 ml arterial blood samples from each subject obtained at different steady-state levels of hypoxia from 70-100%. Blood samples obtained from an arterial line are measured in a hemoximeter to determine true saturation value. Some sponsors may not require any blood sampling. Some sponsors may require additional blood samples, up to 35 samples may be drawn. Readings from the test pulse oximeters are recorded and compared to these "gold-standard" blood values. The pulse oximeter probes may be located at a variety of sites on the subjects, including fingers, toes, ears, forehead, scalp and bridge of the nose. These detector-probes are all non-invasive. The level of hypoxia is measured and controlled by the investigator. A computer program that displays a prediction, breath by breath, the arterial oxygen saturation, (SaO2) of the study subject. SaO2 is computed from end-expired Po2 and Pco2 as determined by mass spectrometer gas analysis. This information permits the inspired gas mixture of air, plus CO2 and nitrogen, to be adjusted by an operator watching the value computed after each expiration on an analog meter. This computer-estimated saturation is adjusted by the operator to one of 6 levels of predicted saturation, and is held stable for about 30 seconds at each level. Two or three "runs" are conducted per subject. Each "run" lasts 10-15 min and 4 to 5 plateaus are tested per run. The manufacturer may choose the target values. Plateaus are typically sought at 96%, 87%, 77%, 92%, 82% and 72%. Other manufacturers have asked for an equal number of data points but with all points between 70% and 100%. Some sponsors may also request additional variables to be measured during testing including high or low CO2, where subjects will be asked to hyperventilate; low perfusion, where subjects will be asked to lay flat, or with their head up or down; testing pulse oximeters during motion, or with the subject's hand fixed to a motion machine. The subject's will be informed by the study staff and in the consent form if any of these additional procedures apply.