Miscoding of hospital discharges as acute myocardial infarction: Implications for surveillance programs aimed at elucidating trends in coronary artery disease