Interpretation of Positive Troponin Results Among Patients with and Without Myocardial Infarction

Abstract

Measuring cardiac troponins is integral to diagnosing acute myocardial infarction (AMI); however, troponins may be elevated without AMI, and the use of multiple different assays confounds comparisons. We considered characteristics and serial troponin values in emergency department chest pain patients with and without AMI to interpret troponin excursions. We compared serial troponin in 124 AMI and non-AMI patients from the observational Performance of Triage Cardiac Markers in the Clinical Setting (PEARL) study who presented with chest pain and had at least one troponin value exceeding the 99th percentile of normal. Because 8 assays were used during data collection, we employed a method of scaling the troponin value to the corresponding assay's 99th percentile upper reference limit to standardize the results. In 81 AMI patients, 96% had elevated troponin at the first test following initial elevation, compared to 73% of the 43 non-AMI patients (P < 0.001). Scaling troponin to the 99th percentile of normal yielded a median value that was 4.8 [2.2, 14.1] times higher than the 99th percentile cutpoint among AMI patients, compared to 2.3 [1.5, 6.5] times higher among non-AMI patients (P = 0.04). The rise in serial scaled troponin values distinguished the AMI patients. Scaling to the 99th percentile was useful for comparing troponin when different assays were utilized.