A comparison of the performance of 20 pulse oximeters under conditions of poor perfusion.

The performance of 20 pulse oximeters with finger probes was evaluated by comparison of their readings with directly measured arterial blood oxygen saturations. The samples were taken from patients who had undergone cardiac surgery under hypothermic cardiopulmonary bypass and had poor peripheral perfusion. The mean difference (bias, accuracy), standard deviation (precision) and drop-out rate for each pulse oximeter was determined. An overall ranking of performance of each pulse oximeter was calculated using five criteria (accuracy, precision, number of readings within 3% of standard, percentage of readings given within 3% of standard, expected overread limit in 95% of cases). Two pulse oximeters achieved a combination of accuracy and precision such that 95% of measurements would be expected to be within 4% of the co-oximeter value; these two also had the lowest drop-out rate.