Spectrum of myocardial contusion.

During a consecutive period of 26 months, 42 patients with blunt chest trauma were diagnosed as having a myocardial contusion on the basis of an abnormal electrocardiogram (ECG) in 36 patients, elevated creatine phosphokinase (CPK) in 39 patients, and positive CPK-Muscle Brain (CPK-MB) isoenzyme in 33 patients. Using these screening modalities, the incidence of myocardial contusion in patients with blunt chest trauma increased from 7 per cent when viewed retrospectively to 17 per cent when viewed prospectively. Eight patients had cardiac index determinations only; of these, three were less than 2.9 1/min/M2. An additional 21 patients underwent a standard fluid challenge of 500 cc of 5 per cent plasmanate infused over 30 minutes allowing construction of a Starling Curve. Five patterns of ventricular function curves were observed. Six patients had biventricular dysfunction, six patients had isolated right ventricular dysfunction, three patients had isolated left ventricular dysfunction, three patients had an "unslope-peak-downslope" pattern, and three patients had normal ventricular function studies. Multiple gated acquisition (MUGA) scans were abnormal in ten patients and normal in 12 patients. Major morbidity and mortality due to myocardial contusion occurred in 17 per cent of the (7/42) patients; of these, three had biventricular dysfunction, one had left ventricular dysfunction, and two had a low cardiac index. This experience suggests that screening tests are sensitive in detecting myocardial contusion in blunt chest trauma, but are not predictive of major morbidity or mortality. Only direct hemodynamic measurement with construction of a Starling Curve was useful in determining the severity of the myocardial injury and identifying those patients at greatest risk.