A comparison of a new reflectance oximeter with the Hewlett-Packard ear oximeter.

The purpose of this study was to characterize the accuracy and dynamic response characteristics of traditional transmittance pulse oximeters and a new reflectance pulse oximeter with regard to the current standard in oximetry, the Hewlett-Packard ear oximeter. Studies were performed with 15 healthy male and female subjects. A rebreathing technique was employed to produce a steady fall in oxygen saturation and to maintain constant, eucapnic, end-tidal CO2 levels. The oximeters' analog outputs were recorded by a Western Graphtec Linearecorder and by an IBM-PC utilizing an analog-to-digital converter for data collection at 5 Hz. The agreement between two clinical methods of measurement was used as the statistical technique of comparing new technology (pulse oximetry) with the current standard (Hewlett-Packard ear oximetry). The mean of the difference between the Hewlett-Packard and the pulse oximeter with reflectance sensor was 0.28%, with a range of difference of +/- 3.49%. This value was similar to that of the transmittance oximeters (Criticare 501+: 0.92 +/- 2.11), (Physio-Control: 3.15 +/- 2.04), (Ohmeda 3700: 2.05 +/- 2.06). There was virtually no response-time difference between the reflectance oximeter with the sensor placed on the subject's forehead and the Hewlett-Packard ear oximeter. The authors conclude that the accuracy of the reflectance oximeter is within the limits of clinical acceptance for monitoring and trending of arterial oxygen saturation in healthy subjects.