Abstract Background/Introduction The use of a single low high-sensitivity cardiac troponin measurement to rule out myocardial infarction (MI) at presentation is advocated by international guidelines. Studies have not systematically compared high-sensitivity cardiac troponin (hs-cTn) I and T assays to determine which values should be used to rule out MI in patients with suspected acute coronary syndrome. Methods In a pooled analysis from three prospective observational cohort studies the lowest cTn concentration was determined that gave a negative predictive value (NPV) of ≥99.5% for MI or death at 30 days on presentation to the emergency department. EDTA or lithium heparin plasma or serum samples collected at presentation were measured using five hs-cTnI assays (Abbott ARCHITECT i2000, Beckman Coulter Access 2, Siemens Atellica IM, Ortho-Clinical Diagnostics VITROS 5600, LSI Medience PATHFAST) and one hs-cTnT assay (Roche Cobas e411). The diagnostic outcome of MI or death at 30 days was adjudicated according to the Fourth Universal Definition of MI. Optimal thresholds were determined to identify the highest proportion of patients that could be considered for discharge for a NPV of ≥99.5% for each assay. Results Among 941 patients, MI or death at 30 days occurred in 113 (12%). Optimized thresholds varied between assays ranging from <2 to <7 ng/L (Table) that identified between 22% and 60% of patients as low risk (0-1%) for MI or death at 30 days. Conclusions Our data identify differences in the effectiveness of thresholds between hs-cTnI and hs-cTnT assays for the same safety performance to rule out MI on presentation using a single measurement. In some cases, these thresholds differ from those used in clinical practice. Due to lack of assay standardization, optimal decision thresholds should be defined and validated for each high sensitivity cTn assay.
Read full abstract