Diagnostic and prognostic value of cardiac troponin I assays in patients admitted with symptoms suggestive of acute coronary syndrome

Fred S Apple,Heidi E Quist,MaryAnn M Murakami
DOI: https://doi.org/10.5858/2004-128-430-DAPVOC
Abstract:Context: Increasing numbers of patients are presenting to emergency departments with symptoms suggestive of an acute myocardial infarction. Objective: To demonstrate the comparative performance of the Ortho Vitros Troponin I and Beckman Access AccuTnI assays used to detect myocardial infarction and to develop risk stratification schemes for all-cause death in patients who presented with myocardial ischemia symptoms that were suggestive of acute coronary syndrome (ACS). Design: The prospective enrollment of patients with ACS and the measurement of serial plasma samples by 2 commercial cardiac troponin I (cTnI) assays. Setting: A metropolitan medical center that admitted patients with ACS during a 2-month period. Patients: The study population consisted of 200 consecutively admitted patients who presented with symptoms that were suggestive of ACS. Results: Correlation scatterplots showed no significant bias between cTnI assays based on 659 specimens across the dynamic range of each assay. Only minor differences in slopes and intercepts were observed between assays when correlations were based across selected concentration ranges. The receiver operating characteristic curve areas for the detection of myocardial infarction were not significantly different (Ortho,.991; Beckman,.995). At the 99th percentile (Beckman, 0.04 microg/L; Ortho, 0.08 microg/L), each assay demonstrated 100% sensitivity with 78% and 80% specificity, respectively. Kaplan-Meier survival curves and the log-rank test were used to compare time-to-event data. Patients with increased baseline cTnI values had higher odds ratios of death than did those with normal concentrations. For Ortho, the 99th percentile cutoff was 5.9, and the 10% coefficient of variation cutoff was 10.3; for Beckman, the 99th percentile cutoff was 31.4, and the 10% coefficient of variation cutoff was 15.3. Conclusions: Comparable diagnostic and risk stratification abilities were demonstrated in patients with ACS by the Ortho Vitros and Beckman Access cTnI assays, with no significant analytic bias between cTnI assays.
What problem does this paper attempt to address?