'Chest pain typicality' in suspected acute coronary syndromes and the impact of clinical experience

Abstract

Background Physicians rely upon chest pain history to make management decisions in patients with suspected acute coronary syndromes, particularly where the diagnosis is not immediately apparent through electrocardiography and troponin testing. The objective of this study was to establish the discriminatory value of "typicality of chest pain" and the effect of clinician experience, for the prediction of acute myocardial infarction and presence of significant coronary artery disease. Methods This prospective single-center observational study was undertaken in a UK General Hospital emergency department. We recruited consecutive adults with chest pain and a nondiagnostic electrocardiogram, for whom the treating physician determined that delayed troponin testing was necessary. Using their own clinical judgment, physicians recorded whether the chest pain described was typical or atypical for acute coronary syndrome. Physicians were defined as "experienced" or "novice" according to postgraduate experience. Acute myocardial infarction was adjudicated using a high-sensitivity troponin (hs-cTn) assay, whereas coronary artery disease was adjudicated angiographically. Results Overall, 912 patients had typicality of chest pain assessed, of whom 114/912 (12.5%) had an acute myocardial infarction and 157/912 (17.2%) underwent angiography. In patients undergoing angiography, 90/157 (57.3%) had hs-cTn elevation, of whom 60 (66.7%) had significant coronary artery disease. Sixty-seven of 157 (42.7%) patients had angiography without hs-cTn elevation; of these, 31 (46.2%) had significant coronary artery disease. For the diagnosis of acute myocardial infarction, chest pain typicality had an area under the curve (AUC) of 0.54 (95% confidence interval [CI], 0.49-0.60). For the prediction of significant coronary artery disease with hs-cTn elevation AUC: 0.54 (95% CI, 0.40-0.67), and without hs-cTn elevation AUC: 0.45 (95% CI, 0.31-0.59). When assessed by experienced physicians, specificity for the diagnosis of acute myocardial infarction was higher at 65.8% (95% CI, 63.1%-68.7%) vs 55.4% (95% CI, 53.9%-56.8%) for novices. Conclusions Subjective interpretation of "typicality of chest pain" is of limited discriminatory value in the assessment of suspected acute coronary syndromes, in the context of a nondiagnostic electrocardiogram. Greater clinical experience improves accuracy as a rule-in tool but does not improve overall discriminatory ability.