Lipemi-interferens vid mätning av Hb på Sysmex XN-10 och HemoCue Hb 201+

Abstract

Anemia can arise from either loss of erythrocytes or impaired production of new erythrocytes. In order to discover and evaluate the treatment of anemic patients, correct Hb measurements are important. A common method to measure Hb concentration is photometry in combination with chemical conversion of the Hb. Like all light-dependent methods this suffers from a vulnerability to turbidity that scatters light. Lipemia is a common cause of turbidity caused by e.g. recent intake of high fat foods, diabetes mellitus, liver or kidney disease, alcoholism and some drugs. Manufacturers of Hb analyzers use different methods to counter the influence of interference from lipemia on measurements. Sysmex XN-10 analyzers use a fat dissolving sheath fluid in its photometric channel (HGB) and HemoCue measures absorbance at a second wavelength to compensate for turbidity. Sysmex XN-10 also has an optic channel (HGB-O) for counting reticulocytes by measuring their nucleic acid and Hb content. At the same time this channel measures Hb equivalents of erythrocytes and gives a calculated value of Hb content in the entire sample. The aim of this study was to compare the photometric and the optical channels for measuring Hb concentration in whole blood. Both the Sysmex XN-10 channels were compared with HemoCue Hb 201+ when measuring Hb concentrations in lipemic samples. Plasma Hb concentration was determined for the corresponding samples in order to investigate correlation between elevation in Hb concentration with and without simulated lipemia and in the plasma after centrifugation. Samples analyzed at Skånes University Hospital in Lund during the month of November 2018 (n = 392) using both HGB and HGB-O on XN-10 were compared using Spearman's signed correlations coefficient. Lipemia was simulated by using the fat emulsion Intralipid in a total of 32 samples. Samples collected and analyzed on the previous day was used for the study. Each sample was split into one part with added Intralipid to form a lipemic sample and one part with NaCl-solution of the same volume as Intralipid in the lipemic sample. The differences between lipemic and non lipemic samples was tested for significance by the non-parametric Wilcoxons signed ranks test for each of the methods. Significance between the three methods was tested by using Kruskal-Wallis and Dunn's tests. Level of significance was set to p < 0.05. The results showed good correlation between earlier test run on both HGB and HGB-O with a Spearman correlation score of 0,982.  A significant difference was found between lipemic and non lipemic samples with the photometric method (p < 0,001) but not the optical method (p = 0,11) on XN10. HemoCue Hb 201+ also showed a significant difference (p < 0,001) between lipemic and non lipemic samples but a lower median than HGB and less deviation than HGB-O. The median of HGB-O indicated that it was influenced the least by lipemia of the three methods but had the greatest deviation of the differences. The greater deviation of HGB-O values may have been caused by hemolysis since the method measures intra cellular Hb. HemoCue shows according to this study the slightest deviation of the three methods and a less heightened median value compared to HGB which confirms the methods suitability as complement to HGB when dealing with lipemic samples.