[Hemolysis interference on clinical chemistry tests analyzed on DxC 700 AU (Beckman Coulter®) and kaliemia rendering algorithm].

The influence of hemolysis was evaluated for 26 clinical chemistry parameters on DxC 700AU (Beckman Coulter®). Ten sample pools were prepared and separated into six aliquots. These aliquots were overloaded with hemolysis in increasing amounts to reach levels equivalent to the maximum hemolysis thresholds H1 (+), H2 (++), H3 (+++) and H4 (++++). Each aliquot is compared to its reference aliquot (not hemolyzed) and a ratio is calculated for each parameter. We proposed that there was a significant difference if, for a given analyte and threshold, more than 20% of the ratios are above the total acceptable limit variability. A significant difference was found for TGP, TGO, cholesterol, creatine kinase (CPK), lactate deshydrogenase (LDH), phosphorus and potassium at H1 (+), chlorine, iron, γ-glutamyltransferase (GGT), and magnesium at H2 (++), amylase and alkaline phosphatase (PAL) at H3 (++++), and prealbumin at H4 (++++). No interference was found until H4 included for uric acid, calcium, creatinine, lipase, glucose, HDL-cholesterol, triglycerides, urea, sodium and immunoglobulins A, G and M. The overestimation of kalemia was calculated as a function of hemolysis, ranging from 0.28 mM +/–0.047 (upper H1 threshold) to 1.37 mM +/–0.126 (upper H4 threshold). Its estimation makes it possible to propose a result rendering algorithm of kalemia according to the hemolysis index. Evaluation of the automates hemolysis indexes is highly recommended for each laboratory. It can allow for some critical parameters the establishment of a decision tree facilitating the result rendering, after clinicobiological consultation.