Incidence and significance of cardiac troponin I release in severe trauma patients

Abstract

Background: The incidence and significance of troponin I release and its mechanism are unknown in severe trauma patients. The characteristics of this release were prospectively studied in such patients and correlated with presence of shock, existence of myocardial contusion, and outcome.

Methods: During a 24-month period, serial electrocardiogram recordings and troponin I measurements were performed in all trauma patients admitted at a surgical intensive care unit. The diagnosis of a significant myocardial contusion was made on electrocardiographic criteria. According to the time course of troponin I, three groups of patients were defined a priori: very transient (</= 12 h) and limited release (troponin I < 2 microg/l), transient (</= 36 h) and significant release (troponin I >/= 2 microg/l), and sustained (> 36 h) and significant release (troponin I > 2 microg/l). In the last group, coronary artery angiography was performed.

Results: The incidence of troponin I release was 12% (95% confidence interval [CI], 9.6-14.4%) in 728 patients. A significant myocardial contusion was found in 35 patients (5%; 95% CI, 3.4-6.6%) and may occur in the absence of chest trauma and without troponin I release. Sensitivity, specificity, and positive and negative predictive values of troponin I for the diagnosis of myocardial contusion were 63, 98, 40, and 98%, respectively. Troponin I release was observed in 54 early (> 48 h) survivors (7%; 95% CI, 5.6-9.6%) without preexisting coronary artery disease. A sustained and significant release of troponin I (17 patients) was frequently associated with chest trauma (82%) and constantly with electrocardiographic abnormalities. A coronary artery injury was found in 7 patients (2 major and 5 minor vascular injuries) (1% of the whole group; 95% CI, 0.4-2.0%). Mortality was similar in early survivors with (15%; 95% CI, 7-27%) or without (12%; 95% CI, 9-14%) troponin I release. The odds ratio for late mortality was 1.32 (95% CI, 0.61-2.85) in patients with troponin I release.

Conclusions: Serial electrocardiogram recordings and troponin I assessments may be proposed for initial screening in high-risk trauma patients to detect anatomical cardiac injuries through the time course of circulating protein. Troponin I release does not have a prognosis value in trauma patients.