Prognostic value of NIRS-derived reoxygenation rate in patients receiving extracorporeal membrane oxygenation

Microcirculatory dysfunction plays a critical role in the pathophysiology of critically ill patients receiving extracorporeal membrane oxygenation (ECMO). However, reliable bedside indicators reflecting microcirculatory function and their prognostic value remain unclear. This study aimed to evaluate the association between a near-infrared spectroscopy (NIRS)–derived dynamic parameter, the reoxygenation rate (ReO₂, %/s), and hospital mortality in ECMO patients, and to compare its predictive performance with conventional clinical indicators. This retrospective cohort study included 57 adult patients who received ECMO support between October 2022 and November 2024. Tissue oxygenation at the thenar eminence was assessed using near-infrared spectroscopy (NIRS) 24 h after ECMO initiation, and dynamic parameters were derived from a vascular occlusion test. The primary outcome was hospital mortality. Multivariable logistic regression, ROC curve analysis, DeLong test, and restricted cubic spline analysis were performed. Among the 57 patients included, 32 (56.1%) died during hospitalization. Non-survivors had significantly lower ReO 2 than survivors (0.74 vs. 1.17%/s, P < 0.001). In multivariable logistic regression analysis, ReO 2 was independently associated with hospital mortality (OR 0.196, 95% CI 0.044–0.881, P = 0.034). ROC analysis showed that ReO₂ had the numerically highest AUC (0.810, 95% CI 0.686–0.917), with significantly better discrimination than lactate ( P = 0.019) but no significant difference compared with baseline StO₂ or APACHE II score. Restricted cubic spline analysis revealed a significant nonlinear association between ReO 2 and mortality ( P for nonlinearity = 0.006). Patients with a ReO 2 ≤ 0.92%/s had significantly higher mortality than those with a ReO 2 > 0.92%/s (78.1% vs. 32.0%, P < 0.001). ReO 2 , a NIRS-derived indicator of microcirculatory function, may be associated with hospital mortality in ECMO patients. ReO₂ may represent an exploratory marker of impaired microvascular reperfusion, although prospective validation is required before clinical application. Not applicable.