The relationship between hemolysis and I-index: investigating its use as a screening test in hyperbilirubinemia.

Abstract

Hemolysis (H-index) and icteric indexes (I-index) are widely used in clinical laboratories to control interferences. Since I-index correlates with bilirubin, it can aid in deciding whether to analyze bilirubin levels. The study aims to evaluate I-index use for this purpose, considering H-index interference on I-index. This retrospective study included H-index, I-index and total bilirubin results, all analyzed using the Mindray BS800M. Total bilirubin was analyzed using the Vanadate oxidase method. Patients were categorized according to their H-index values in multiples of 0.5. Group-0 had the lowest hemolysis (<0.5 g/L), and Group-10 had the highest hemolysis (>5 g/L), with a total of 11 groups. ROC analysis determined optimal I-index cut-offs to differentiate normal and abnormal bilirubin levels, calculating PPV, NPV, specificity and sensitivity. Indexes of 55,306 patients and total bilirubin results of 26,781 patients were evaluated. There was no significant difference between the groups in pairwise comparison for bilirubin results but significant differences were observed in I-index values. When the I-index values were examined, there were significant differences between Group 0,1,2,3,4, and all groups from 0 to 10 (p < 0.001). As hemolysis increased, the I-index cutoff values decreased. While the cut-off values for Group-0 and Group-4 ranged from 16.70 to 14.15 μmol/L, in Group-5 (hemolysis 2.5-3 g/L), the cut-off was 6.50 μmol/L, in Group-6 and Group-7, it was -2.35 μmol/L. In Group-10, this value dropped to -42.55 μmol/L. Our findings suggest using the I-index to spot hyperbilirubinemia and save on testing. However, different cut-off values should be calculated according to the hemolysis value.