Objective: Obtaining an accurate medical history is essential in the assessment of patients, particularly in emergency department (ED) patients with acute chest pain, as there can be a time imperative for diagnosis and commencement of treatment. We aimed to evaluate reliability of patient-reported compared with physician-adjudicated medical history by assessing patient's recall and communication of personal events and its influence on the accuracy of the medical history.
Methods: A total of 776 patients presenting at ED with suspected cardiac chest pain were recruited. Data collection included self-reported patient history, electrocardiogram testing, and troponin I measurements. Independent assessment of risk factors and medical history was adjudicated by cardiologists. Diagnosis of acute coronary syndrome (ACS) at 30 days after presentation was assessed. Cohen's kappa measured patient-cardiologist agreement. Cardiologist adjudicated events were taken as true to assess accuracy.
Results: A total of 83 participants (10.7%) were diagnosed with ACS at 30 days after presentation. "Previous coronary artery bypass grafting" showed highest agreement (K = 1.00) between patient-reported and cardiologist-adjudicated events. Lowest agreement between patient-reported and cardiologist-adjudicated events was found for "prior ventricular dysrhythmia" (K = 0.33). Accuracy of reported "prior congestive heart failure" differed significantly between patients with and without diagnosed ACS at 30 days (92.8% and 97.5%, respectively).
Conclusions: Accuracy of patient's recall and communication of medical history and risk factors was substantial but not perfect in the assessment of patients with ACS in the ED context. Our study reinforces the importance in the utilization of medical records and collateral information to address possible discrepancies in the medical history and improve patient care.