The analytical performance assessment and clinical practice of three sensitive cardiac troponin Ⅰ assays / 中华检验医学杂志

ABSTRACT

Objective To assess the analytical performance of three sensitive cardiac troponin Ⅰ assays and compare the clinical application to provide help in choosing the detection method for clinical laboratory. Methods A total of 474 serum samples were collected from apparently healthy subjects and a total of 112 serum samples were collected from patients presenting with symptoms suggestive of acute myocardial infarction. The functional sensitivities of three assays from Abbott, Beckman-Coulter and Ortho were determined ( CV = 10% ). The reference ranges have been established. The analytic performance was compared according to the assessment mode described by Apple. The relationship was compared among the different assays. The preliminary clinical application value for different detection methods has been evaluated and validated with self-established reference ranges. Results The functional sensitivities ( CV = 10% ) of the cTnI assays for Abbott, Beckman-Coulter and Ortho were 0. 030, 0. 04 and 0. 013 μg/L, respectively.The 99th percentiles of cTnI in healthy volunteers were 0. 021, 0. 02 and 0. 026 μg/L respectively. The analytical data of ROC curve showed that the area under curve (AUC) of the cTnI assays for Abbott,Beckman-Coulter and Ortho for diagnosis of AMI was 0. 852,0. 909 and 0. 910,respectively. There was no statistical difference between any two methods(Z1 = 1.18 ,Z2 = 1.21 ,Z3 =0. 026,all P ＞0. 05). There were good consistency between the 99th percentile obtained from our laboratory and suggested by manufacturers (Kappa value were 1. 000, 0. 730 and 0. 893 respectively, all P ＜ 0. 01 ). Conclusions The analytical performance of two cTnl assays is "clinical accepted" ,the other one is accepted according to guideline. All of them could detect cTnI in apparently healthy subjects. There exist differences among three assays, but their diagnostic characteristics for AMI are not significantly different.