BACKGROUND: Among patients hospitalized for acute ischemic stroke, abnormal serum troponins are associated with higher risk of short-term mortality. However, most findings have been reported from European hospitals. Whether troponin elevation after stroke is independently associated with death among a more heterogeneous US population remains unclear. Furthermore, only a few studies have evaluated the association between the magnitude of troponin elevation and subsequent mortality, patterns of dynamic troponin changes over time, or whether troponin elevation is related to specific causes of death. METHODS: Using data collected in the American Heart Association's 'Get With The Guidelines' stroke registry between 2008 and 2012 at a tertiary care US hospital, we used logistic regression to evaluate the independent relationship between troponin elevation and mortality after adjusting for demographic and clinical characteristics. We then assessed whether the magnitude of troponin elevation was related to in-hospital mortality by calculating mortality rates according to tertiles of peak troponin levels. Dynamic troponin changes over time were evaluated as well. To better understand whether troponin elevation identified patients most likely to die due to a specific cause of death, investigators blinded from troponin values reviewed all in-hospital deaths, and the association between troponin elevation and mortality was evaluated among patients with cardiac, neurologic, or other causes of death. RESULTS: Of 1,145 ischemic stroke patients, 199 (17%) had elevated troponin levels. Troponin-positive patients had more cardiovascular risk factors, more intensive medical therapy, and greater use of cardiac procedures. These individuals had higher in-hospital mortality rates than troponin-negative patients (27 vs. 8%, p < 0.001), and this association persisted after adjustment for 13 clinical and management variables (OR 4.28, 95% CI 2.40-7.63). Any troponin elevation was associated with higher mortality, even at very low peak troponin levels (mortality rates 24-29% across tertiles of troponin). Patients with persistently rising troponin levels had fewer anticoagulant and antiatherosclerotic therapies, with markedly worse outcomes. Furthermore, troponin-positive patients had higher rates of all categories of death: neurologic (17 vs. 7%), cardiac (5 vs. <1%), and other causes of death (5 vs. <1%; p < 0.001 for all comparisons). CONCLUSIONS: Ischemic stroke patients with abnormal troponin levels are at higher risk of in-hospital death, even after accounting for demographic and clinical characteristics, and any degree of troponin elevation identifies this higher level of risk. Troponins that continue to rise during the hospitalization identify stroke patients at markedly higher risk of mortality, and both neurologic and non-neurologically mediated mortality rates are higher when troponin is elevated.
BACKGROUND: Among patients hospitalized for acute ischemic stroke, abnormal serum troponins are associated with higher risk of short-term mortality. However, most findings have been reported from European hospitals. Whether troponin elevation after stroke is independently associated with death among a more heterogeneous US population remains unclear. Furthermore, only a few studies have evaluated the association between the magnitude of troponin elevation and subsequent mortality, patterns of dynamic troponin changes over time, or whether troponin elevation is related to specific causes of death. METHODS: Using data collected in the American Heart Association's 'Get With The Guidelines' stroke registry between 2008 and 2012 at a tertiary care US hospital, we used logistic regression to evaluate the independent relationship between troponin elevation and mortality after adjusting for demographic and clinical characteristics. We then assessed whether the magnitude of troponin elevation was related to in-hospital mortality by calculating mortality rates according to tertiles of peak troponin levels. Dynamic troponin changes over time were evaluated as well. To better understand whether troponin elevation identified patients most likely to die due to a specific cause of death, investigators blinded from troponin values reviewed all in-hospital deaths, and the association between troponin elevation and mortality was evaluated among patients with cardiac, neurologic, or other causes of death. RESULTS: Of 1,145 ischemic strokepatients, 199 (17%) had elevated troponin levels. Troponin-positive patients had more cardiovascular risk factors, more intensive medical therapy, and greater use of cardiac procedures. These individuals had higher in-hospital mortality rates than troponin-negative patients (27 vs. 8%, p < 0.001), and this association persisted after adjustment for 13 clinical and management variables (OR 4.28, 95% CI 2.40-7.63). Any troponin elevation was associated with higher mortality, even at very low peak troponin levels (mortality rates 24-29% across tertiles of troponin). Patients with persistently rising troponin levels had fewer anticoagulant and antiatherosclerotic therapies, with markedly worse outcomes. Furthermore, troponin-positive patients had higher rates of all categories of death: neurologic (17 vs. 7%), cardiac (5 vs. <1%), and other causes of death (5 vs. <1%; p < 0.001 for all comparisons). CONCLUSIONS:Ischemic strokepatients with abnormal troponin levels are at higher risk of in-hospital death, even after accounting for demographic and clinical characteristics, and any degree of troponin elevation identifies this higher level of risk. Troponins that continue to rise during the hospitalization identify strokepatients at markedly higher risk of mortality, and both neurologic and non-neurologically mediated mortality rates are higher when troponin is elevated.