Evaluation of a new pediatric continuous oximetry catheter

Abstract

To evaluate the accuracy of central venous oxygen saturation recordings from a new in-line pediatric oximetry catheter. Prospective, observational study. Eighteen pediatric patients who needed central venous access for monitoring and/or treatment between January 2006 and June 2006 in the pediatric intensive care unit of the Hospital for Sick Children in Toronto, Canada. Measurements were done at the baseline and then every 4-8 hrs. The monitor was calibrated in vivo at the baseline and then daily. In vitro calibration of the monitor was also performed in the last five patients. The hemoglobin value was updated when there was a significant change. The maximum duration of sampling was 72 hrs (if indicated). There were 131 measurements in 17 patients; each subject had a different number of paired measurements (median 5). Three patients were excluded due to violation of the protocol, and 113 measurements were left in analysis. The mean difference of catheter value from the laboratory value was -1.01 (median 0). The interquartile range was 5. The difference of both methods was evenly distributed as per a Bland-Altman plot, with one patient's data lying outside of the comparable limits of ± 1.96 sd from the mean differences. The relationship of the difference between the catheter data and the lab data to the independent variables (age, weight, gender, catheter tip, diagnosis, and signal quality index) was estimated by using the multiple regression analysis (version 9.1, SAS Institute, Cary, NC). All variables were eliminated. The Pearson correlation coefficient between lab-mixed venous oxygen saturation and oximetry catheter readings for measurements was 0.88. In this limited number of patients, use of the PediaSat venous oximetry catheter was safe and had good agreement with co-oximetry-measured values.