Performance of Near-Infrared Spectroscopy in Detecting Acute Tourniquet-Induced Upper-Extremity Ischemia Across Different Skin Phenotypes

Abstract

Diagnosing acute tissue ischemia is challenging, particularly in patients with higher skin melanin content. We investigated whether near-infrared spectroscopy (NIRS) is effective and consistent in detecting upper extremity ischemia across various skin phenotypes. Volunteers underwent tourniquet-induced upper extremity ischemia. Skin color was evaluated by the Fitzpatrick scale (FP, range: I-VI) and the Von Luschan scale (vL, range: 1-36). A NIRS probe was placed on one finger. The tourniquet was inflated to 250 mmHg and perfusion was restricted for 7 minutes, followed by a 10-minute monitored reperfusion period. The percent tissue oxygenation (StO2) was recorded. A total of 55 volunteers were enrolled (22 self-identified as Caucasian, 21 African American, 7 Asian, 2 Latinx, and 2 Biracial). Average starting and ending StO2 for the cohort was 72.2% and 45.9%, respectively. However, there was variability based on skin melanin content. Increasing vL correlated with lower starting StO2, smaller StO2 decrease, and shorter time to reach ischemic steady state. High skin melanin (FP scale IV-VI) was associated with significantly lower starting StO2 (-7.1%) and shorter time to reach ischemic steady state (-0.3 mins). African Americans had lower starting StO2 (-8.6%) and 7.8% lesser total StO2 decrease than other groups. NIRS can rapidly detect acute onset tissue ischemia in the upper extremity. However, given the lower starting StO2 and smaller total StO2 decrease after tourniquet-induced ischemia for patients with higher skin melanin, using NIRS for clinical detection of acute ischemia may be more challenging in these patients. These inconsistencies may limit use of NIRS clinically for spot identification of ischemia. Although NIRS has utility in tracking tissue oxygenation, variable performance with different skin melanin content raises concerns as to whether different cutoff/threshold levels are needed for different groups, and whether NIRS is reliable for spot checks in acute events.