Diagnostic value of serum troponin in stable patients at risk for blunt cardiac injury.

Abstract

There are no gold standard criteria for diagnosing blunt cardiac injury (BCI). While the combination of electrocardiogram (ECG) and serum troponin is considered the most sensitive screening method for those being considered for discharge, little evidence exists regarding the diagnostic value and clinical utility of initial and serial serum troponin levels in stable patients who are being admitted with a possible BCI. Therefore, this study seeks to determine if troponin level is an independent predictor of adverse cardiac events in stable, admitted patients at risk for BCI. This was a five-year retrospective study using the trauma database at a University Level I Trauma Center. The study population included adult trauma patients presenting with a physician diagnosis of BCI or a sternal fracture who met prespecified stability criteria (SBP ≥ 90 mmHg, HR < 110bpm, shock index < 1, GCS ≥ 14). The data collected included all troponin values and ECG interpretations as well as echocardiogram interpretations, when performed. A patient was classified as having an adverse cardiac event if they were diagnosed with a new arrhythmia requiring treatment, had cardiac surgery, or suffered cardiac-related mortality. Sensitivity, specificity, positive and negative likelihood ratios, and diagnostic odds ratios were calculated to analyze the diagnostic performance of the different BCI screening modalities. Diagnostic tests were compared using the Exact Mcnemar's test and the exact binomial test. 350 patients met inclusion criteria. There were 12 adverse cardiac events; each were new arrhythmias requiring treatment with one patient also requiring synchronized cardioversion. Only one adverse cardiac event occurred in a patient with a normal ECG and troponin (n = 160). No patients with a normal ECG and abnormal troponin had an adverse cardiac event (n = 83). Patients with an abnormal ECG were more likely to have an adverse cardiac event (p < 0.001). In stable patients in this cohort, troponin level did not predict adverse cardiac events. Therefore, admitted patients at risk for BCI who meet stability criteria might be safely observed without measurement of serum troponin.