Interanalyzer Analytical Variation of a High-Sensitivity Cardiac Troponin T Assay Can Exceed the Cutoff of the European Society of Cardiology 1-Hour Algorithm for Ruling Out Non–ST-Segment Elevated Myocardial Infarction

The European Society of Cardiology (ESC) guidelines (1) recommend that patients presenting with chest pain can be ruled out for non–ST-segment elevation myocardial infarction (NSTEMI) if their baseline Roche Diagnostics high-sensitivity cardiac troponin T (hs-cTnT) result is either <5 ng/L (if chest pain started at least 3 h earlier) or <12 ng/L and the change 1 h later is <3 ng/L. Because of the rounding of results, the analytical variation of the Roche hs-cTnT assay should not exceed 2 ng/L at concentrations <12 ng/L for use in the 0- to 1-h algorithm. A large clinical biochemistry diagnostic laboratory usually has multiple analyzers performing a hs-cTnT assay to meet service demand. Zero- and 1-h samples for hs-cTnT from the same patient are likely to be processed on different analyzers. Although the lot-to-lot analytical variation of the Roche hs-cTnT assay has already been investigated (2), the analytical variation between separate Cobas modules over the likely duration of a patient’s initial investigation is less understood. The aim of this study was to determine whether the interanalyzer analytical variation of the Roche hs-cTnT assay over an 8-h period is satisfactory for the adoption of the ESC 0- to 1-h algorithm. A serum pool with a hs-cTnT concentration of approximately 6 ng/L was made from anonymous and previously separated redundant serum obtained from the blood sciences department at St George’s Hospital. Intra- and interanalyzer precision was assessed on the Roche Cobas modules in use in the central laboratory (2 Cobas e602 and 1 Cobas e801) and the emergency department stat laboratory (1 Cobas e411) at St George’s hospital. The serum pool was not recentrifuged before analysis on any of the Cobas modules. Intra- and interanalyzer repeatability (short-term precision) was assessed by analysis of the serum pool 25 times on each analytical cell on each Cobas module within 1 h (n = 150 on 1 Cobas e411, 2 Cobas e801, and 3 Cobas e602 cells). Interanalyzer analytical variation over an 8-h period was assessed by replicating realistic sample flow. Ten aliquots of the serum pool were analyzed in the stat laboratory on the Cobas e411 to represent a hs-cTnT baseline at 0 h. Five aliquots were then processed on our preanalytical modular Roche Cobas 8000 in the central laboratory at each 2-h interval; the samples were sent for analysis on a Cobas e602 or e801 (depending on system demand). The mean, SD, and 95% CIs were calculated using the Analyse-it software (3) for each Cobas module and time point. The Cobas e411 module demonstrated the poorest intra-analyzer repeatability, with a CV of 9.6% and a range of 3 ng/L (n = 25). The Cobas e602 and e801 modules demonstrated improved intra-analyzer repeatability, with CVs <5.7% and ranges ≤2 ng/L. After exclusion of the Cobas e411, the interanalyzer repeatability within 1 h gave a CV of 7.7% and a range of 2 ng/L (n = 125). Results of the 8-h interanalyzer analytical variation experiment are shown in Table 1. When 0-h baseline samples were processed on the Cobas e411 and subsequent samples were analyzed on the Cobas e602/e801 (n = 30), results increased by up to 3 ng/L. After exclusion of the 0-h measurements on the Cobas e411, no 2 results over the 2-, 4-, 6-, and 8-h periods on the Cobas e602/e801 modules (n = 20) differed by >1 ng/L. Interanalyzer analytical variation of the Roche high-sensitivity cardiac troponin assay over an 8-h period. Interanalyzer analytical variation of the Roche high-sensitivity cardiac troponin assay over an 8-h period. A previous study reported that Roche hs-cTnT assay lot variations between 2013 to 2018 were large (2). This finding could potentially render the use of the 0- to 1-h algorithm inappropriate if the lot numbers between the different Cobas modules vary. If hospitals adopt these ESC rule-out algorithms, it is important that laboratories monitor the performance of the Roche hs-cTnT assay by using third-party internal quality control material with appropriate concentrations at the clinical cutoff levels (5 and 12 ng/L). In conclusion, the interanalyzer analytical variation between separate Cobas e601 and e802 modules is satisfactory for the implementation of the ESC 0- to 1-h NSTEMI rule-out algorithm. However, the independent Cobas e411 module should not be used to provide the baseline or repeat hs-cTnT measurement for use in this algorithm because the imprecision might lead to reporting a false increase of >2 ng/L, failing to correctly rule out NSTEMI. Therefore, we recommend that all hs-cTnT requests intended for the use in the ESC 0- to 1-h algorithm be processed on the more precise Cobas e602 and e801 modules. This recommendation is still subject to the monitoring of significant reagent lot-to-lot variation and the appropriate use of low-concentration internal quality control. ESC, European Society of Cardiology; NSTEMI, non–ST-segment elevation myocardial infarction; hs-cTnT, high-sensitivity cardiac troponin T. All authors confirmed they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved. K.C. Flowers, statistical analysis, administrative support, provision of study material or patients; C. Hunt, statistical analysis, administrative support. Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest: P.O. Collinson, The Journal of Applied Laboratory Medicine, AACC. None declared. None declared. None declared. None declared. None declared. None declared.