Comparative analysis of arterial oxygen saturations during exercise by pulse oximetry, photometric measurements, and calculation procedures.

Pulse oximetry allows non-invasive monitoring of arterial oxygen saturation (SO2). To study the validity of pulse oximetry, comparative measurements were performed. During exhaustion limited exercise SO2-values measured by pulse oximetry (SO2puls), calculated SO2-values (algorithms of Kelman, Severinghaus, and Siggaard-Andersen--SO2calc), and as "golden standard" photometric measured SO2-values (SO2meas) were compared. Fourteen triathletes performed a stepwise cycling exercise test in the supine position. SO2calc was determined on the basis of capillary actual blood gas values. SO2puls was measured continuously with a finger probe attached to the second finger. The SO2puls- and SO2calc-values differed from the SO2meas-values (p less than 0.05); however, the differences were of no clinical relevance. Performing linear regression analysis, only SO2puls correlated significantly (r = 0.47, p less than 0.001) with SO2meas. Pulse oximetry is able to replace invasive measurements of arterial oxygen saturation in athletes. It is superior to SO2-calculations and permits reliable, valid and non-invasive continuous monitoring of SO2.