Abstract P335: Building Assessment Tools for Unrecognized Myocardial Infarction: The REasons for Geographic and Racial Differences in Stroke Study
Introduction: People with unrecognized myocardial infarction (UMI) have elevated risk for cardiovascular events and death. Population-wide electrocardiogram screening for detection of UMI would be resource intensive. We therefore sought to develop assessment tools for targeted UMI screening.
Methods: Using logistic regression and a backward selection procedure, we developed two models among 16,653 participants without coronary revascularization or known myocardial infarction in the REasons for Geographic and Racial Differences in Stroke (REGARDS) study. UMI was defined as electrocardiogram evidence (using the Minnesota Code system) of myocardial infarction without a self-reported history (n = 740). The basic model used demographics, self-reported medical history, blood pressure, and body mass index; the expanded model used information on 51 potential variables including health behaviors, clinical measurements, and participant-reported health scales and symptoms.
Results: The basic model had a c-statistic of 0.638 (95% CI 0.617 − 0.659) and included age, race, smoking status, body mass index, systolic blood pressure, transient ischemic attack, deep vein thrombosis, falls, diabetes, and hypertension. The expanded model had a c-statistic of 0.654 (95% CI 0.634−0.674) and included age, race, income, smoking status, self-reported hypertension, unrecognized dyslipidemia, perceived stress score, waking at night due to breathlessness, body mass index and body mass index squared, HDL cholesterol squared, systolic blood pressure, albumin:creatinine ratio, heart rate, and heart rate squared. The Framingham coronary heart disease risk score had a c-statistic of 0.587 (95% CI 0.566−0.609).
Conclusions: Even with a large number of potential risk factors, the models did not provide sufficient discrimination to warrant use as an assessment tool. Identifying individuals at high risk of UMI for further screening using an assessment tool may be challenging.
- © 2012 by American Heart Association, Inc.