The study of the influence of hemolysis was determined experimentally for twenty two biochemical parameters on the analyzer Cobas 6000 ce (Roche Diagnostics). The addition method of hemolysate was used to create an increasing concentration of hemoglobin ranging from 0 to 2000 μmol/L. The limit of 10% variation was chosen to define the influence of hemolysis on the measurement. The parameters studied were classified into several categories: the parameters for which hemolysis does not influence the measurement: albumin, uric acid, calcium, C-reactive protein, myoglobin, NT -pro BNP, S100 protein, and urea; parameters impacted positively leading to an overestimation of the result: aspartate aminotransferase, total cholesterol, creatine kinase, creatinine, lactate dehydrogenase, magnesium, magnesium, total protein, triglycerides; and negatively impacted settings so causing an underestimation of the result: alanine amino- transferase, gamma glutamyl transferase, lipase, alkaline phosphatase, troponin T hypersensitive. Certain parameters influence of hemolysis varies depending on the magnitude of the measured parameter this interference being observed for normal values but disappearing for pathological values: creatinine, cholesterol, alkaline phosphatase, triglycerides, or the inverse interference is greater than for conventional pathological values: lipase, alanine amino-transferase. Knowledge of this variability interference allows the biologist to adapt its methods of reporting in the case of haemolysed samples.
Keywords: hemoglobinemia; hemolysis; interference; overestimation; underestimation.