The (Dis) Utility of a Change in Troponin I for Diagnosis of Non-ST-Segment Elevation Myocardial Infarction in an Observation Unit.

Abstract

Observation units (OUs) may be an efficient and effective setting to diagnose and risk stratify patients with coronary ischemia and myocardial infarction (MI). Given improved cardiac troponin I (cTnI) assays and expanded utilization of OUs, it is not uncommon for patients with mildly elevated cTnI to be evaluated in OUs. We investigated the serial cTnI results in OU patients to determine whether absolute or relative cTnI changes were useful for the diagnosis of MI. This was a retrospective study of 260 patients placed in the OU from a single center in 2007, with an initial cTnI in the indeterminate range of 0.04-0.2 ng/ml (Siemens ultrasensitive), and a second cTnI was drawn at 6 hours. The diagnosis of MI was determined based on the third universal definition of MI by consensus review of 2 cardiologists, with adjudication by a third cardiologist in case of disagreement. Of the 260 patients, 25 (9.6%) were determined to have MI at OU presentation. The optimal absolute and relative change in cTnI for MI diagnosis by receiver operating characteristic curve analysis were 0.02 ng/ml and 40%, respectively. There was initial cardiologist disagreement in 60% (15/25) of MI cases despite full review of serial cTnI and cardiac testing results. At 30 days, there were 3 adverse events: 2 deaths and 1 MI. The diagnosis of MI in OU with low-level cTnI elevation is problematic. Furthermore, there is only marginal diagnostic utility of serial changes in cTnI in this patient population.