Pulse oximeter accuracy and precision affected by sensor location in cyanotic children

Children's digits are often too small for proper attachment of oximeter sensors, necessitating sensor placement on the sole of the foot or palm of the hand. No study has determined what effect these sensor locations have on the accuracy and precision of this technology. The objective of this study was to assess the effect of sensor location on pulse oximeter accuracy (i.e., bias) and precision in critically ill children. Design: Prospective, observational study with consecutive sampling. Setting: Tertiary care, pediatric intensive care unit. Patients: Fifty critically ill children, newborn to 2 yrs of age, with an indwelling arterial catheter. Forty-seven of 50 (94%) patients were postcardiac surgery. Interventions: None. Measurements and Main Results: Co-oximeter-measured arterial oxygen saturation (SaO₂) was compared with simultaneously obtained pulse oximetry saturations (SpO₂). A total of 98 measurements were obtained, 48 measurements in the upper extremities (finger and palm) and 50 measurements in the lower extremities (toe and sole). The median SaO2 was 92% (66% to 100%). There was a significant difference in bias (i.e., average SpO₂SaO₂) and precision (±1 SD) when the sole and toe were compared (sole, 2.9 ±3.9 vs. toe, 1.6±2.2, p =.02) but no significant difference in bias and precision between the palm and the finger (palm, 1.4 ± 3.2 vs. finger, 1.2 ± 2.3, p = .99). There was a significant difference in bias ± precision when the SaO₂ was <90% compared with when SaO₂ was >90% in the sole (6.0 ± 5.7 vs. 1.8 ± 2.1, p = .002) and palm (4.5 ± 4.5 vs. 0.7 ± 2.4, p = .006) but no significant difference in the finger (1.8 ± 3.8 vs. 1.1 ± 1.8, p = .95) or toe (1.9 ± 2.9 vs. 1.6 ± 1.9, p = .65). Conclusions: The Philips M1020A pulse oximeter and NellcorMAX-N sensors were less accurate and precise when used on the sole of the foot or palm of the hand of a child with an SaO₂ <90%. (Pediatr Crit Care Med 2008; 9:393-397)