Is the high-sensitivity troponin I assay accurate in identifying patients with suspected acute coronary syndromes who are at low risk of cardiac events?
The high-sensitivity troponin I assay used in this study was fairly accurate in identifying patients with chest pain who are at low risk of cardiac events. (Level of Evidence = 1c−)
These authors conducted a two-part study in adults presenting to the emergency department with suspected acute coronary syndromes. Each of the participating departments had been randomized into the standard care arms of trials designed to determine if high-sensitivity troponin I levels will influence care and improve outcomes. In the first part, the derivation cohort, the researchers evaluated nearly 4,900 adults, 16% of whom had an acute myocardial infarction (MI) at the time of emergency department presentation and an additional handful who had a subsequent acute MI or died (1% and 2%, respectively) within 30 days. From this cohort, they determined that to achieve a 99.5% negative predictive value (determined a priori) a threshold value for high-sensitivity troponin I should be 5 nanograms per liter. In the second part of the study, the validation cohort, the researchers used two different groups and followed more than 1,400 patients for up to one year after enrollment. Approximately 0.5% of the patients with a low high-sensitivity troponin I level had an acute MI—nearly 100% negative predictive value. The researchers evaluated the performance across many subgroups (sex, age, prior heart disease, and so forth) and found no significant differences. After one year, 0.3% of the patients with a low high-sensitivity troponin I level had an incident acute MI and 0.3% had a cardiac death compared with 1.3% and 2.2% in those with elevated high-sensitivity troponin I level, respectively.
Study design: Cohort (prospective)
Funding source: Government
Setting: Emergency department
Reference: ShahASAnandASandovalYet alHigh-STEACS investigatorsHigh-sensitivity cardiac troponin I at presentation in patients with suspected acute coronary syndrome: a cohort study. Lancet2015; 386( 10012): 2481– 2488.