Pulse oximeter validation study

Introduction: The aim of this study was to assess a) the reliability of our pulse oximeters for SpO2 and HR signals, b) to detect malfunctions, c) and to provide recommendations on SpO2 testing. Methods: A pulse oximeter tester (Metron DEAG) was to test 12 Konica Minolta pulse oximeters over a split 6 month period. The suitability of the analyser was measured against a reference standard (OxSim-1 pulse oximeter simulator). The ranges that were simulated include SpO2 of 85%, 95% and 98% at heart rates (HR) of 80, 40 and 140 beats per minute (b.p.m) respectively. Results: The Bland-Altman analysis did not reveal any significant difference at all 3 ranges for both analysers. The mean differences at SpO2 of 85% and HR of 40 b.p.m was -0.37% and ±2SD of (0.67- -1.39). At 95% 80 b.p.m mean difference of +0.67±2SD (+1.18 - +0.15) and 98% at 140 b.p.m a mean difference of 0.08% ±2SD (+0.44- -0.29). The difference between the two analysers was correlated better towards the higher range of SpO2. There was no variability in the simulated HR during the testing of all the pulse oximeters. Discussion: The results of the study show that the oximeters appear to be remarkably stable over time. Conclusions: The change in SpO2 over Period 1 (mean simulated readings of 86.1%, 95.8% and 98.0%) and Period 2 (mean simulated readings of 86.2% and 95.7% and 98.0%) was negligible. The P values for all three ranges (85% P value of 0.65, 95% P=0.59 and 98% P=0.47) were > 0.05 and therefore no significant differences in the data existed. We conclude that regular testing of pulse oximeters does not help predict failure or problems that aren9t picked up by the regular operators.