The German CPU Registry: Dyspnea independently predicts negative short-term outcome in patients admitted to German Chest Pain Units

Background: While dyspnea is a common symptom in patients admitted to Chest Pain Units (CPUs) little is known about the impact of dyspnea on their outcome. The purpose of this study was to evaluate the impact of dyspnea on the short-term outcome of CPU patients. Methods: We analyzed data from a total of 9169 patients admitted to one of the 38 participating CPUs in this registry between December 2008 and January 2013. Only patients who underwent coronary angiography for suspected ACS were included. 2601 patients (28.4%) presented with dyspnea. Results: Patients with dyspnea at admission were older and frequently had a wide range of comorbidities compared to patients without dyspnea. Heart failure symptoms in particular were more common in patients with dyspnea (21.0% vs. 5.3%, p < 0.05) at admission. Importantly, in patients presenting with dyspnea the 3 month mortality was fourfold higher compared to patients without dyspnea (8.6% vs. 2.1%, p < 0.05, OR death: 4.40 95% CI 3.14-6.03). Interestingly, the mortality estimated from the GRACE risk score was below the actual mortality assessed after the 3 month follow-up. After adjustment for the GRACE risk score or for heart failure, dyspnea remained highly predictive of death and myocardial infarction within 3 months (OR death adjusted for heart failure: 2.99 95% CI 1.99-4.47 and OR death adjusted for GRACE risk score: 3.37 95% CI 2.27-4.99). Conclusion: Dyspnea is a common symptom in CPU patients. Our data show that dyspnea is associated with a fourfold higher 3 month mortality which is underestimated by the established ACS risk scores. To improve their predictive value we therefore propose to add dyspnea as an item to common risk scores. (C) 2014 Elsevier Ireland Ltd. All rights reserved.