Pulse oximetry in critically ill children.

To confirm the clinical applicability of a commercial pulse oximeter, we compared arterial hemoglobin saturation values determined by in-vitro oximetry and pulse oximetry in 15 critically ill children. One hundred ninety-two paired hemoglobin saturations were determined by both noninvasive pulse oximetry and direct measurement of arterial blood samples. The correlation between these two methods of measurement was statistically significant (r = 0.895; p less than 0.001). The mean percentage difference between the two measurements was 1.8%. Pulse oximetry was found to be safe and less cumbersome than other methods of monitoring arterial oxygen content. Overall, pulse oximetry was precise and provided a clinically satisfactory noninvasive method for continuously monitoring arterial hemoglobin saturation in critically ill children.