Electrocardiographic ST-segment elevation myocardial infarction in critically ill patients: an observational cohort analysis.

OBJECTIVE To investigate the specificity of the electrocardiographic diagnosis of ST-segment elevation myocardial infarction in the critical care unit setting. DESIGN Retrospective observational cohort analysis. SETTING An 880-bed tertiary care teaching hospital with 120 intensive care unit beds. PATIENTS The population included medical, surgical, trauma, and neurosurgical intensive care unit patients. INTERVENTIONS Electrocardiograms were systematically collected to include all consecutive recordings over a 15-month period in which the interpretation software indicated ***ACUTE MI***. Patient demographics, markers of intensive care unit complexity, and hospital mortality were ascertained. The electrocardiograms were then further evaluated by a blinded, board-certified cardiologist for agreement or disagreement with the interpretation software. Serum troponin measurements obtained within 96 hrs of electrocardiogram acquisition were used to determine the likelihood of myocardial infarction. MEASUREMENTS AND MAIN RESULTS Over the 15-month study period, the interpretation software diagnosed ST-segment elevation myocardial infarction in 67 of 2243 intensive care unit patients (2.99%) who had an electrocardiogram performed. In the final study population of 46 cases with electrocardiographic ST-segment elevation myocardial infarction, 85% had peak troponin elevation<5 ng/mL, a strong suggestion against clinical ST-segment elevation myocardial infarction. The cardiologist agreed with the computer interpretation in 39% (18 of 46) of cases, but of those 18 patients, only six showed a significant rise in the troponin level. The cardiologist disagreed with the computer interpretation in 60.9% (28 of 46) of cases and of those, one patient had a marked elevation of the cardiac troponin. CONCLUSIONS ST-segment elevation myocardial infarction in the intensive care unit is a relatively common electrocardiographic reading both by standard interpretation software and by expert evaluation. In contrast to nonintensive care unit patients who present with chest pain, the electrocardiographic ST-segment elevation myocardial infarction diagnosis seems to be a nonspecific finding in the intensive care unit that is frequently the result of a variety of nonischemic processes. The vast majority of such patients do not have frank ST-segment elevation myocardial infarction.