Dynamic in vivo response characteristics of three oximeters: Hewlett-Packard 47201A, Biox III, and Nellcor N-100.

Pulse oximeters (Biox III, Nellcor N-100) and a transmittance oximeter [Hewlett-Packard 47201A (HP)] were compared for SaO2 measurement and responsiveness during dynamic changes in arterial oxygen saturation and heart rate. Five sleep apnea syndrome patients were studied because they had large oscillations in SaO2 and heart rate in sleep. During sleep, each patient exhibited a series of rapid (18.0 +/- 8.3 s, mean +/- SD) oscillations in oxygen saturation (92.1 +/- 2.6% to 74.2 +/- 7.7%). Oxygen saturation measurements were sampled simultaneously from each oximeter by computer (at 2 Hz). Accuracy was assessed by comparing pulse and transmittance oxygen saturation measurements at the peak and trough of each apnea-related oscillation. Oximeter response was defined in terms of the "delay" or absolute time difference between the pulse oximeters and the transmittance oximeter for the determination of the peak and trough saturations. Linear regression analysis was used to establish accuracy and response relationships between pulse oximeter sensors (reusable ear, reusable digit, disposable digit, and disposable nasal sensors) and the transmittance oximeter sensor (reusable ear sensor). Pulse oximeter response delay was highly correlated with heart rate. Pulse oximeter SaO2 measurement and response characteristics varied considerably with sensor type (disposable, reusable) and sensor location (ear, nose, and digit). One must be aware of these differences in clinical and research application.