Eliminating Diagnostic Drift in the Validation of Acute In-Hospital Myocardial Infarction--Implication for Documenting Trends across 25 Years: The Minnesota Heart Survey

Abstract

Long-term trends in epidemiologic studies of acute myocardial infarction (AMI) require application of a consistent diagnostic algorithm. Typically an algorithm includes chest pain, cardiac enzymes, electrocardiographic findings, and autopsy results. The Minnesota Heart Survey (MHS) has determined trends for incident AMI and for in-hospital and long-term outcomes over a 25-year period (1970-1995). However, dramatic changes have occurred that seriously challenge the ability of the MHS and other epidemiologic studies to use a consistent diagnostic algorithm. These include newer and more sensitive cardiac biomarkers, introduction of diagnosis-related groups, and change in International Classification of Diseases coding. In the MHS, the electrocardiogram is the only diagnostic element consistently available and consistently classified over this 25-year period. The authors identified eight dichotomous Minnesota Code criteria that provided a consistent diagnostic method from 1970 to 1995 as documented by extensive cross-validation. These criteria were combined into a logistic score and used to define incident, recurrent, and attack AMI rates over this 25-year period. For both men and women, AMI rates determined by electrocardiogram are parallel to rates based on the International Classification of Diseases and parallel over adjacent survey periods to the standard MHS algorithm. The electrocardiogram classified by Minnesota Code provides the only consistent long-term diagnostic tool for AMI trends over this 25-year period.