Analytical outlier rates for Roche fifth generation cardiac troponin T stat using biotin sensitive and insensitive reagent lots.

Abstract

We compared analytical outlier rates for cardiac troponin T (cTnT) between biotin sensitive and insensitive testing reagents, and between testing platforms Roche Cobas e411 and e601. Plasma samples from hospitalized patients were tested on a Roche Cobas e411 analyzer using lots of Roche Elecsys Troponin T Gen 5 Stat reagent labelled free of biotin interference at ≤20 ng/mL (biotin sensitive reagent, n = 1997); compared to samples tested on a Roche Cobas e601 analyzer using the same reagent labelled free from biotin interference at ≤1200 ng/mL (biotin insensitive, n = 2555). Samples with measurable cTnT (≥6 ng/L) were aliquoted, re-centrifuged and retested on the same analyzer. Analytical outliers were defined as repeat value that differed from the initial value by a critical difference (CD): where CD = z x 2 x SD_analytical, with a z-value of 3.719. The total number of outliers that were observed for the biotin-sensitive reagent on the e411 was 68/1997 (3.4%), compared to 208/2555 (8.1%) on the biotin-insensitive reagent measured on the e601 (p < 0.0001). Outliers were observed in 45 (2.6%) of samples with initial values <100 ng/L using the biotin-sensitive reagent compared to 138 (6.4%) of samples using biotin insensitive (p < 0.0001). Among these samples, outliers with repeat values differing by ≥10 ng/L occurred for 0.8% and 1.3% of samples tested with the biotin sensitive and insensitive reagent, respectively. cTnT analytical outliers occurred more than twice as often using biotin-insensitive reagent on the Cobas e601 analyzer compared to biotin-sensitive reagent on the Cobas e411.