Validation of Plasma Free Hemoglobin Testing for Extracorporeal Membrane Oxygenation (ECMO) Patients

Circuit-induced hemolysis is relatively common in extracorporeal membrane oxygenation (ECMO) patients. Intravascular release of cell-free hemoglobin can lead to complications and requires timely recognition. Validation of plasma free hemoglobin (PFH) measurement using a direct spectrophotometric method is presented.

We evaluated a method modified from Kahn et al. (Ann Clin Lab Sci 1981;11:126-31) on a stand-alone spectrophotometer (Cary 60) and compared its performance to the semiquantitative H-index on an Abbott Alinity c, including precision, linearity, recovery, reference interval verification, interference, and stability. Method comparison was performed relative to the H-index and the same method on a different spectrophotometer (Beckman DU 720). Lipemia interference was performed on the Cary 60, Cary 3500, and Beckman DU 720. Surrogate biomarkers for hemolysis detection were also investigated in ECMO patients.

The PFH method on the Cary 60 demonstrated imprecision ranging from 1% (96.0 mg/dL) to 4% (3.0 mg/dL), linearity to 100 mg/dL, and recovery >80% for values >2 mg/dL hemoglobin-spiked plasma. Dilution expanded the reportable range to the maximum dilution tested (1000 mg/dL). Lipemia interfered with PFH measurement by the direct method, but the same method on the Cary 3500 was resistant to lipemia. Bilirubin did not cause significant interference. Direct and H-index methods were comparable with a mean difference of 5.03 mg/dL (95% CI -1.38, 11.44). Lactate dehydrogenase was the most reliable surrogate biomarker for hemolysis. with AUC of 0.921 (0.894, 0.949) at >50 mg/dL.

PFH measurement by a direct spectrophotometric method is more precise and sensitive compared to the H-index; however, PFH measurement is susceptible to lipemia unless performed on a high-end spectrophotometer.