Cardiac troponins (cTn) have been used historically to estimate infarct size in ST elevation myocardial infarction (STEMI). Within a resource constrained health care environment, cTn could therefore be used for prioritisation of patients for cardiac imaging, in particular echocardiography. We aimed to determine how useful routinely collected cTn would be in predicting significant left ventricular (LV) impairment.

All patients in the All New Zealand Acute Coronary Syndrome Quality Improvement (ANZACS-QI) registry with their first episode of STEMI between January 2013 and November 2018, who had high sensitivity troponin T measured, were included. We excluded patients with no left ventricular ejection fraction (LVEF) assessment, known LV dysfunction, or prior myocardial infarction.

In total, 3,698 patients were included in the analysis. A higher mean hsTnT (admission and peak) was seen in patients with more severely impaired LV function but there was significant overlap in the range of hsTnT between the different LVEF categories. Cardiac troponins demonstrated poor discriminative ability to either predict or exclude significant LV impairment (LVEF <40%). At an optimal cutpoint of 3,405 ng/L, peak hsTnT had a sensitivity of 56.5% (95% confidence interval [CI] 42–62%), a specificity of 65.3% (95% CI 62–79%) and an area under the receiver operating curve of 0.62 (95% CI 0.60–0.64).

This is the largest study comparing clinically measured troponin levels and LV function in patients presenting with STEMI. A definite, but weak, association was seen between peak troponin and the degree of LV dysfunction, with significant overlap in troponin levels between levels of myocardial dysfunction. Routinely acquired troponin is not suitable for clinical use as a method of prioritising patients for cardiac imaging.