Evaluation of conventional troponin I testing for the detection of myocardial dysfunction in children

Abstract

Objectives: Troponin is a marker of myocardial injury but is not well studied in children. Our primary objective was to ascertain the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of conventional troponin I for the detection of acute myocardial dysfunction in previously healthy children. Our secondary objective was to identify clinical predictors of myocardial dysfunction in the setting of elevated troponin. Study Design: This was a retrospective chart review in a single, paediatric, tertiary care centre of troponin tests performed in all admitted children over a 4-year period. Demographics, symptoms, signs, chest x-ray, ECG, and echocardiogram abnormalities were documented. Myocardial dysfunction was presumed to be absent when the patient had a normal cardiac assessment, with or without echocardiography, and did not re-present. Results: From January 2014 through December 2017, 566 patients had troponin tested as a screen for myocardial injury. Troponin was positive in 38 of 566 cases (6.7%). Myocardial dysfunction was detected in 9 of 566 cases (1.6%). Troponin was elevated in six of nine cases of myocardial dysfunction. The sensitivity of conventional troponin I for detecting acute myocardial dysfunction was 66% (95% confidence interval [CI] 30 to 93%). The specificity was 94% (95% CI 92 to 96%). PPV was 16% (95% CI 6 to 31%) and NPV 99% (95% CI 98 to 100%). An abnormal ECG was more prevalent in patients with a true positive versus a false-positive troponin result (P=0.03). Conclusion: Troponin testing identified few cases of myocardial dysfunction. We found the test to have only 66% sensitivity. Troponin testing as a screen for myocardial injury in children has limited utility.