A Comparison of Three Methods for Defining Acute Myocardical Infarction

• Published in: Academic emergency medicine Vol. 10; no. 5; p. 534
• Main Author: Nagurney, J. T
• Format: Journal Article
• Language: English
• Published: Des Plaines Wiley Subscription Services, Inc 01-05-2003
• ISSN: 1069-6563; 1553-2712
• DOI: 10.1197/aemj.10.5.534

OBJECTIVE: To compare three methods of defining acute myocardial infarction (AMI): the traditional 1985 WHO criteria, the recent 2000 ESC/ACC criteria, and chart criteria (CC). METHODS: type of study: single-center, retrospective descriptive study. Setting: a level one 70,000-visit university ED. Subjects: any adult admitted to the hospital through our ED over a three month period with at least one elevated cardiac biomarker (TnI, TnT, CKMB) recorded in the hospital database. Observations: all laboratory data, medical records, and ECG results were reviewed. Published criteria for WHO 1985 (1985) and ESC/ACC 2000 (2000) definitions of AMI were applied to each subject by two trained, blinded coders. A subject was considered positive (pos) by CC if the medical record indicated that they had sustained an AMI during that admission. Analysis: a multi-rating system kappa statistic was calculated, as well as the percent overlap among criteria, with 95% confidence intervals. An interobservor kappa statistic on the two abstractors for key variables was also determined. RESULTS: Preliminary results show that there were 399 eligible subjects, and data available on 396. Among them, all three criteria were pos in 15% ((95% CI 11-19%), negative (neg) in 27% (95% CI 23-31%), and disagreed in 58% (95% CI 53-63%). The multi-rating system kappa statistic for all 3 criteria was 0.19(95% CI 0.14-0.25). There were 277 subjects who were 2000 criteria pos. Among them, 21% (95% CI 16-26%) were pos by all three criteria but 52% (95% CI 46-58%) were neg by both 1985 and CC. An additional 16% (95% CI 12-20%) were CC pos and 1985 neg and 11% (95% CI 7-15%) were 1985 pos and CC neg. CONCLUSION: The agreement for the definition of AMI among the new 2000 ESC/ACC criteria, traditional 1985 WHO criteria, and providers (CC) is poor. These new 2000 ESC/ACC criteria appear to include large numbers of patients who would not be considered to have had an AMI by traditional criteria or by their providers.