The Use of Infra-Red Spectroscopy to Assess Intact Skin of Various Skin Tones

Background Pressure injuries (PI) continue to be a serious medical complication in the acute care setting. Industry leaders advocate for the use of objective technologies to provide accurate and quantifiable skin assessments for all patients across diverse skin tones. The Infra-Red Spectroscopy Scanner (IRSS) is a novel non-invasive decision support system (DSS) electro-optic technology that uses near infra-red spectroscopy to examine and assess physical and chemical changes prior to visual skin breakdown in real-time. Methods A 2-phase prospective open-label pilot device study was launched under protocol IRM-02-1123 (USA version 3.0, November 1, 2023). In both Phase 1 and Phase 2, Certified Wound Ostomy Continence Nurses (CWOCNs) collected the study data, including visual and tactile skin assessments and IRSS device scans. Phase 1 supported device calibration and validation, study-personnel training, safety assessment, and algorithm refinement. Phase 2 evaluated device safety and efficacy performance against CWOCN visual and tactile assessment. Adults with a Stage 1 PI or deep tissue PI were eligible; Phase 1 also allowed at-risk adults with Fitzpatrick skin types IV-VI and a Braden score below 18. Six individual scans were collected at each eligible anatomical site. Efficacy performance was evaluated at the anatomical site level. For Phase 2, the conservative threshold was prespecified, while the balanced threshold was evaluated as a post hoc approach prioritizing overall accuracy. Results Across Phases 1 and 2, 26 participants contributed 219 anatomical site observations and 1314 individual device scans. Of the 219 anatomical sites, 188 (85.8%) were clinically classified as healthy and 31 (14.2%) as PIs. Forty-six percent of participants had Fitzpatrick skin types I-III and 54% had types IV-VI. In Phase 1, the IRSS device performance demonstrated an accuracy of 81%, sensitivity of 91%, and specificity of 80% and an area under the curve (AUC) of approximately 92.2%. In Phase 2, the balanced threshold yielded 87.7% accuracy, 90.0% sensitivity, 87.3% specificity, and an AUC of approximately 96%, demonstrating strong discrimination between anatomical sites clinically classified by CWOCNs as healthy tissue versus PI. No device-related adverse events were identified. Conclusion These interim results from the ongoing Phase 2 study demonstrate promising anatomical site agreement between the IRSS and CWOCN classification. Phase 2 enrollment and data collection are ongoing, and continued enrollment will provide additional data to further evaluate device performance across skin-tone groups and anatomical sites.