Methodology: The prospective study involved 14 neonates and 11 infants, all under the age of 3 months, undergoing heart surgery under hypothermic cardiopulmonary bypass. Accuracy was determined by the degree of agreement between the pulse ox and the hemoximeter values .Bias refers to the difference between the arterial saturation determined by the hemoximeter (SaO2) and the pulse oximetry saturation (SpO2). The hemoximeters were calibrated weekly and optimized daily. Blood samples were collected for direct oximetry using indwelling arterial catheters. These samples were collected anaerobically, stored in ice and analyzed 30 minutes after collection. Pulse oximeter readings were recorded with both the hand and the foot as probe sites. Before initiation of the cardiopulmonary bypass most of the subjects received 1mg/kg phenoxybenzamine intravenously and in 10 infants pulse oximetry readings were taken before and after administration of the drug. (Parvathi, McDougall and Peter 1996).
Results: Pulse oximetry readings using the foot probe and the hand probe were similarly affected by temperature. With a foot temperature of above 29 0c, 94.7% of the measurements were within the acceptable range (+/-3%) with the bias increasing at lower temperatures. At foot temperatures below 27 0c about 45.5% of the readings were outside the acceptable range. At hand temperatures below 27 0c the readings with unacceptable bias were greatly increased (47.7%). The difference performance of the pulse oximeter at the two different sites was not statistically significant. In ten infants with peripheral temparatures below 27 0c oximetry readings before and after the administration of phenoxybenzamine were taken with the post administration readings taken at mean time of 19.5% (10-25 mins range) after administration. Before administration of phenoxybenzamine the mean bias was 4.15 +/- 5.47% and decreased to 0.86 +/-2.86 after drug administration the bias. The phenoxybenzamine readings were beyond the acceptable limits for the 10 cases but 7 improved to the acceptable limit with time (Parvathi, McDougall and Peter 1996).
Conclusion: The performance of the pulse oximeter was independent of the site of the probe but was affected by the temperature of the probe site with 95% of the observations showing good agreement with the hemoximeter saturation values at 29 0c. The accuracy of pulse oximetry for monitoring oxygen saturation in infants is compromised at temperatures below 27 0c. This study also showed that phenoxybenzamine improves the accuracy of pulse oximetry under conditions of hyperthermia and low peripheral temperatures (Parvathi, McDougall and Peter 1996).
Parvathi, L., McDougall and P., Peter,L., (1996). Accuracy of Pulse Oximetry in Hypothermic Neonates and Infants Undergoing Cardiac Surgery. Critical Car Medicine , 507-511.