Effect of probe design on accuracy and reliability of pulse oximetry in pediatric patients

Study Objective: To determine if the traditional board-wrap disposable puls e oximeter probe is more accurate or reliable for oxygen monitoring in chil dren than the reusable clip-type probe. Design: Prospective observational clinical study. Setting: Operating room of a large university hospital. Patients: 18 children 11 years of age or younger, who were scheduled for ge neral anesthesia with placement of an intraarterial catheter. Interventions: Pulse oximetry values were obtained using both band and clip probes of three pulse oximeters (Nellcor; Hayward CA; Novametrix, Wallingf ord, CT; Ohmeda, Boulder, CO) and compared with simultaneous hemoximetry va lues. Dropout rate (percent of down time) also was recorded for each probe- machine combination. Measurements and Main Results: Data were analyzed using bins and precision and t-test. p < 0.05 is considered significant. Bias (mean SpO(2) - SaO(2)) is less than 2% for all probe-machine combinations. The range of error obs erved between SpO(2) and SaO(2) for all data points was greatest using the Nellcor band (27.8%) and least using the Ohmeda band (11.4%). In cyanotic c hildren, the greatest bins and precision were observed with Nellcor band (N = 11, -5.12 +/- 9.74) and the best agreement with Novametrix band (N = 17, 0.08 +/- 4.21). The difference in bias for any test units above or below t he median weight of 13.75 kg did not vary by more than 1%. Dropout rate was minimal for all units in nonbypass situations. After cardiopulmonary bypas s, no data dropout was observed with Ohmeda band; observed down time with t he other units varied between 34% and 55%. Conclusions: The type of probe s elected has little effect on accuracy of pulse oximetry in children. After cardiopulmonary bypass, using the Ohmeda band combination may improve the l ikelihood of obtaining consistent reading and decreasing down time. (C) 199 9 by Elsevier Science Inc.