Agreement of somatic and renal near-infrared spectroscopy with reference blood samples during a controlled hypoxia sequence: a healthy volunteer study

Purpose: O3 (R) Regional Oximetry (Masimo Corporation, California, USA) is validated for cerebral oximetry. We aimed to assess agreement of somatic and renal near-infrared spectroscopy with reference blood samples. Methods: O3 sensors were placed bilaterally on the quadriceps and flank of 26 healthy volunteers. A stepped, controlled hypoxia sequence was performed by adding a mixture of nitrogen and room air to the breathing circuit. O3-derived oxygen saturation values were obtained at baseline and at six decremental saturation levels (5% steps). Blood samples (radial artery, iliac vein (somatic reference) and renal vein) were obtained at each step. Reference values were calculated as: 0.7 x venous saturation + 0.3 x arterial saturation. The agreement between O3-derived values with blood reference values was assessed by calculating root-mean-square error accuracy and Bland-Altman plots. Results: The root-mean-square error accuracy was 6.0% between quadriceps oxygen saturation and somatic reference values. The mean bias was 0.8%, with limits of agreement from -7.7 to 9.3%. These were 5.1% and 0.6% (-8.3 to 9.5%) for flank oxygen saturation and somatic reference values, respectively, and 7.7% and -4.9% (-15.0 to 5.2%) for flank oxygen saturation and renal reference values. The kidney depth was 3.1 +/- 0.9 cm below the skin. Conclusion: O3 regional oximetry can be used on the quadriceps and flank to monitor somatic saturation, yet has a saturation-level dependent bias. O3-derived values obtained at the flank underestimated renal reference values. Additionally, it is unlikely that the flank sensors did directly measure renal tissue.

Automated summary

This study validates the use of O3 regional oximetry for monitoring muscle saturation in the quadriceps and flank. However, it is important to note that there is a saturation-level dependent bias and the values obtained at the flank underestimate renal reference values.