Detecting paraprotein interference on a direct bilirubin assay by reviewing the photometric reaction data

Abstract

Abstract Background: The direct bilirubin (D-Bil) assay on the AU Beckman Coulter instrumentation can be interfered by paraproteins, which may result in spurious D-Bil results. In a previous work, we took advantage of this fact to detect this interference, thus helping with the identification of patients with unsuspected monoclonal gammopathies. In this work, we investigate the possibility to detect interference based on the review of the photometric reactions, regardless of the D-Bil result. Methods: The D-Bil assay was carried out in a set of 2164 samples. It included a group of 164 samples with paraproteins (67 of which caused interference on the assay), as well as different groups of samples for which high absorbance background readings could also be expected (i.e. hemolyzed, lipemic, or icteric samples). Photometric reaction data were reviewed and receiver operating characteristics (ROC) curves were used to establish a cut-off for absorbance that best discriminates interference. Results: The best cut-off was 0.0100 for the absorbance at the first photometric point of the complementary wavelength in the blank cuvette. Once the optimal cut-off for probable interference was selected, all samples analyzed in our laboratory that provided absorbance values above this cut-off were further investigated to try to discover paraproteins. During a period of 6 months, we detected 44 samples containing paraproteins, five of which belonged to patients with non-diagnosed monoclonal gammopathies. Conclusions: Review of the photometric reaction data permits the systematic detection of paraprotein interference on the D-Bil AU assay, even for samples for which reasonable results are obtained.