Extracorporeal membrane oxygenation (ECMO) can induce massive intravascular hemolysis, commonly quantified by plasma free hemoglobin (fHb) measurement. While manual spectrophotometry using second derivative spectrophotometry (sDS) is the reference method, a recent assessment tool based on the hemolysis index (HI) was developed for fHb determination on different automated systems. This study compares the performance of two HI methods-on Roche Cobas c502 and Abbott Architect c16000-versus sDS in 83 ECMO patients. Method validation was performed on both automated HI systems, including determination of linearity, precision, and analytical interference from lipemia and icterus. Plasma fHb measurements in 83 ECMO patients from two hospital centers were compared between the two HI-derived methods and two manual sDS methods (at 415 nm and 578 nm). Both HI-automated methods showed excellent linearity (0.04-20 g/L) and reproducibility (CVs between 2.3 % and 6.3 %). However, they exhibited sensitivity to interference by high lipemic index, leading to underestimation of fHb for Abbott Architect and overestimation for Roche Cobas. In non-lipemic ECMO samples with fHb levels >0.10 g/L, a very good correlation was found between HI and sDS, especially at 578 nm. However, discrepancies were observed in lipemic samples, leading to clinically relevant biases for fHb values >0.50 g/L. The Roche Cobas and Abbott Architect HI methods are reliable tools for monitoring fHb in ECMO patients, with good reproducibility and linearity. However, in the presence of significant lipemia, confirmation by sDS is recommended to ensure accurate assessment of hemolysis. This approach facilitates an efficient and automated monitoring of fHb, essential for daily managing ECMO-related complications and improving patient outcomes.< Réduire