Predictors of short- and long-term mortality in hospitalized veterans with elevated troponin.

BACKGROUND: Cardiac troponin elevation is associated with mortality. We compared the mortality risk related to elevated troponin from acute coronary syndrome (ACS) and non-ACS causes in a hospitalized elderly veteran population. METHODS AND RESULTS: As part of a quality initiative at our Veterans Affairs hospital, all patients with elevated troponin were evaluated by a cardiologist to determine if ACS was present and to recommend management. We selected a sample (n = 761) of consecutive patients studied between February 2006 and February 2007 and examined all-cause mortality over extended follow-up. Nearly all were men (99.1%), and about half had coronary disease (n = 385, 50.5%) and diabetes (n = 339, 44.4%). ACS patients had lower mortality that non-ACS patients. Mortality began to diverge at 30 days; at 1 year it was 42.0% versus 29.0% (odds ratio [OR]: 0.56, 95% confidence interval [CI]: 0.41-0.78) and at 6 years 77.7% versus 58.7% (OR: 0.41, 95% CI: 0.30-0.56). Cox regression models for mortality at multiple time points yielded several independent factors associated with mortality; however, the distribution of the factors was not sufficient to explain the observed difference in mortality. CONCLUSIONS: In this elderly, male veteran population, mortality related to an elevated troponin was higher at 1 and 6 years for non-ACS patients compared with ACS patients. Factors independently associated with a higher mortality risk were predominantly markers of general systemic illness, but did not elucidate the reasons why troponin elevation secondary to non-ACS causes carries this higher risk. A better understanding of these cardiac troponin elevations and implications for future mortality requires additional investigation. Journal of Hospital Medicine 2016;11:773-777. © 2016 Society of Hospital Medicine.