Objective: Patient history and physical examination are widely accepted as cornerstones of diagnosis in modern medicine. We aimed to assess the value of individual historical and examination findings for diagnosing acute myocardial infarction (AMI) and predicting adverse cardiac events in undifferentiated Emergency Department (ED) patients with chest pain.
Methods: We prospectively recruited patients presenting to the ED with suspected cardiac chest pain. Clinical features were recorded using a custom-designed report form. All patients were followed up for the diagnosis of AMI and the occurrence of adverse events (death, AMI or urgent revascularization) within 6 months.
Results: AMI was diagnosed in 148 (18.6%) of the 796 patients recruited. Following adjustment for age, sex and ECG changes, the following characteristics made AMI more likely (adjusted odds ratio, 95% confidence intervals): pain radiating to the right arm (2.23, 1.24-4.00), both arms (2.69, 1.36-5.36), vomiting (3.50, 1.81-6.77), central chest pain (3.29, 1.94-5.61) and sweating observed (5.18, 3.02-8.86). Pain in the left anterior chest made AMI significantly less likely (0.25, 0.14-0.46). The presence of rest pain (0.67, 0.41-1.10) or pain radiating to the left arm (1.36, 0.89-2.09) did not significantly alter the probability of AMI.
Conclusions: Our results challenge many widely held assertions about the value of individual symptoms and signs in ED patients with suspected acute coronary syndromes. Several 'atypical' symptoms actually render AMI more likely, whereas many 'typical' symptoms that are often considered to identify high-risk populations have no diagnostic value.
Copyright 2009. Published by Elsevier Ireland Ltd.