A validation method for near-infrared spectroscopy based tissue oximeters for cerebral and somatic tissue oxygen saturation measurements

摘要

We describe the validation methodology for the NIRS based FORE-SIGHT ELITE^® (CAS Medical Systems, Inc., Branford, CT, USA) tissue oximeter for cerebral and somatic tissue oxygen saturation (StO2) measurements for adult subjects submitted to the United States Food and Drug Administration (FDA) to obtain clearance for clinical use. This validation methodology evolved from a history of NIRS validations in the literature and FDA recommended use of Deming regression and bootstrapping statistical validation methods. For cerebral validation, forehead cerebral StO2 measurements were compared to a weighted 70:30 reference (REF CXB) of co-oximeter internal jugular venous and arterial blood saturation of healthy adult subjects during a controlled hypoxia sequence, with a sensor placed on the forehead. For somatic validation, somatic StO2 measurements were compared to a weighted 70:30 reference (REF CXS) of co-oximetry central venous and arterial saturation values following a similar protocol, with sensors place on the flank, quadriceps muscle, and calf muscle. With informed consent, 25 subjects successfully completed the cerebral validation study. The bias and precision (1 SD) of cerebral StO2 compared to REF CXB was −0.14 ± 3.07%. With informed consent, 24 subjects successfully completed the somatic validation study. The bias and precision of somatic StO2 compared to REF CXS was 0.04 ± 4.22% from the average of flank, quadriceps, and calf StO2 measurements to best represent the global whole body REF CXS. The NIRS validation methods presented potentially provide a reliable means to test NIRS monitors and qualify them for clinical use.