The influence of interference by hemolysis, icterus and lipemia on the results of routine chemistries may lead to wrong interpretations. The H-, I- and L-indices that can be measured by the Beckman LX-20 instrument (Beckman Coulter) in serum or plasma samples are a reliable semi-quantitative measure of the size of these interferences. A survey carried out in 16 Dutch clinical laboratories on the use of these indices demonstrated that in several of these laboratories, the influence of interferences is largely underestimated. Therefore, a multicenter study was carried out in which we examined the interference of hemolysis, icterus and lipemia on 32 analytes. On the basis of biological variation, we decided on cutoff indices above which analytically significant interference exists. We found analytically significant interference by hemolysis, icterus or lipemia, in 12, 7 and 15 of the 32 analytes studied, respectively. Flagging of results on the basis of analytically significant interference, however, results in too many clinically insignificant comments. On the basis of clinical significance, we conclude that significant interference by hemolysis, icterus or lipemia is present in only 5, 6 and 12 of the analytes studied, respectively. Use of the cutoff indices presented here facilitates optimal use of the LX-20 indices to prevent reporting of wrong results due to interference.

The influence of interference by hemolysis, icterus and lipemia on the results of routine chemistries may lead to wrong interpretations. On Synchron LX-20 instruments (Beckman Coulter) serum or plasma indices can be used as reliable semi-quantitative measures of the magnitude of such interference. In an article recently published in this journal, we presented the results of a multicenter study carried out in Dutch hospitals in which we determined cutoff indices for analytes above which analytically significant interference exists. Clinically significant interference cutoff indices were also derived for these analytes. In this article, we describe the handling of patient samples with clinically significant interference by hemolysis, icterus or lipemia. We investigated several possible approaches for correction of the result: dilution of the interference; mathematical correction in the case of hemolysis; treatment with ferrocyanide to destroy bilirubin; and removal of lipids in lipemic patient samples. We concluded, that mathematical correction of potassium or lactate dehydrogenase results in hemolytic samples can only be carried out if intravascular hemolysis is ruled out. Hemoglobin quantification in serial patient samples, combined with measurement of haptoglobin, represents a useful tool to rule out in vivo hemolysis. We derived an algorithm for this situation. We do not simply recommend mathematical correction, unless it is clinically acceptable. We present formulas for potassium and lactate dehydrogenase: corrected potassium=measured potassium-(hemolytic index increment x 0.14); corrected lactate dehydrogenase=measured lactate dehydrogenase-(hemolytic index increment x 75). The dilution studies indicated that dilution is only applicable for bilirubin, C-reactive protein and iron. The results of treatment with ferrocyanide were poor, and we do not recommend this method. Removal of lipids using high-speed centrifugation or LipoClear (StatSpin Inc.), a non-toxic and non-ionic polymer, is a very effective approach, although C-reactive protein, creatine kinase-MB (CK-MB) and cholesterol cannot be removed using LipoClear. For all interferants (hemoglobin, bilirubin, lipids), relatively simple algorithms are derived that can easily be implemented in the clinical laboratory.