Objective: Despite multiple inquiries, there are no available tests to definitively detect blunt myocardial injury. The evaluation of patients with chest wall injuries without other indications for intensive care unit (ICU) admission has ranged from a single emergency department electrocardiogram (ECG) to 72 hours of continuous electrocardiographic monitoring. Recently, signal-averaged ECG and serum cardiac troponin T have demonstrated clinical utility in the evaluation of ischemic heart disease. The purpose of this study is to determine the ability of these diagnostic tests to predict the occurrence of significant electrocardiographic rhythm disturbances for patients with chest wall injuries and no other indication for ICU admission.
Methods: We prospectively evaluated 71 consecutive adult patients admitted to a regional Level I trauma center with chest wall injuries not requiring ICU admission. We obtained admission signal-averaged ECG, serum troponin T level, standard ECG, and creatine phosphokinase (CPK-MB) level. Patients received continuous electrocardiographic monitoring, follow-up 12-lead electrocardiography, and serial monitoring of troponin and CPK-MB. Echocardiography was performed for patients with abnormal CPK-MB levels. Electrocardiographic events were graded as normal, abnormal but clinically insignificant, or clinically significant. Multiple stepwise logistic regression analysis was used to evaluate predictors for the development of clinically significant electrocardiographic events.
Results: On admission, 17 of 71 patients (23.9%) had normal sinus rhythm; 13 (18.3%) had a clinically significant finding. For 50 patients, follow-up ECG was abnormal; for 26, the findings were clinically significant. Of 17 patients with normal initial ECGs, 7 (41%) developed a clinically significant abnormality. Six patients received intervention for ECG findings. Eleven of 71 patients (16%) had positive troponin T; 5 of 71 (7%) had positive CPK-MB; 15 of 71 (21%) had positive signal-averaged ECG; and 4 of 13 had positive echocardiograms. Initial electrocardiographic abnormalities and a troponin T level > 0.20 microg/L were the only variables found to predict clinically significant electrocardiographic events. Sensitivity and specificity of troponin T in predicting clinically significant abnormalities were 27 and 91%, respectively.
Conclusions: 1. The best predictors for the development of significant electrocardiographic changes are an admission ECG abnormality and an elevated serum troponin T level. 2. Both tests have high specificity with low to moderate sensitivity. 3. Patients with normal ECGs may develop clinically significant events. 4. CPK-MB and echocardiograms continue to be poor predictors of significant electrocardiographic events.