Pilot Study: Using Reflectance as a Proxy for Skin Pigmentation to Assess Pulse Oximeter Accuracy

This study used a novel approach to measure skin pigmentation via reflectance rather than assigning individuals to subjective groups to investigate correlation between skin pigmentation and pulse oximeter reading accuracy.

A prospective observational study with a concurrent retrospective medical record review of clinical data was conducted at a Midwestern academic medical center. Pediatric patients admitted in the intensive care unit with arterial blood gas readings recorded during their hospitalization were eligible for the study. Photovolt, a reflectance meter, was used to obtain skin reflectance measurements. Arterial gas, oxygen saturation as measured by pulse oximetry (Spo2), and hemoglobin values closest in time to the skin readings were obtained. Linear and robust regression were performed for statistical analysis.

Thirty-two patients were enrolled in the study. The median age of the patients was 15 months (range: 6 days to 7 years). Reflectance ranged from 12% to 40% with 12% indicating the most pigmented skin of participants. Spo2 overestimated the true O2 across participants for arterial oxygen saturation (Sao2) near 90%. At the end of low reflectance (12%-21%) and low Sao2 (75%-78%), Spo2 overestimated O2 by 6%. Each 10-unit decrease in reflectance was associated with an increase in Spo2%-Sao2% of 1% (95% CI: 0%-2%; P = .006), indicating overestimation of Sao2 status in darker skin.

Pulse oximeters overestimated Spo2 reading in patients with lower skin reflectance value (darker skin tones) at lower oxygen status range. Additionally, the difference between Spo2 and Sao2 grew significantly different as objectively measured skin reflectance decreased. This suggests that in the pediatric population, darker skin tone as measured by skin reflectance is associated with an overestimation in oxygen status via pulse oximetry.