Background
The icteric index (I-index) is a practical and low-cost method for screening hyperbilirubinemia and potentially reducing unnecessary total bilirubin (TBil) testing. This study was designed to determine a safe I-index to detect TBil above the upper reference range to avoid unnecessary TBil orders. The reflexive addition of TBil in samples where I-index is above the cut-off value and TBil measurement is not initially planned will be a secondary gain of the study.
Methods
This study included 267185 samples with TBil test results and corresponding I-index values. TBil and I-index values were measured with the Roche Cobas 8000 (c701) and Cobas 6000 (c501) analyzers. Statistical analyses were performed using SPSS 15.0, with ROC curve analysis used to determine the optimal I-index cut-off for TBil screening.
Results
An I-index cut-off of 34.2 μmol/L provided the lowest false negativity (0.45 %), highest specificity (97.1 %), and negative predictive value (99.5 %). Using this cut-off point, if we had reached 100 % NPV, 87.4 % of TBil tests in the group receiving inpatient/outpatient treatment with Cobas-8000 and 89.8 % in the emergency group with Cobas 6000 could have been saved.
Conclusions
This is the study with the largest sample size. The sample size is critical in determining a reliable I-index cut-off point. We conclude that I-index should not replace TBil since the NPV was not 100 %; however, it seems reasonable to reflexively add TBil testing in samples where the I-index is above the cut-off value and TBil measurement was not initially planned.
扫码关注我们
求助内容:
应助结果提醒方式:
