Reliability using the universal classification of acute myocardial infarction compared to ST-segment classification

Abstract

To study the inter-physician reliability using the universal classification (UC) of acute myocardial infarction (AMI) compared to the ST-segment classification (STC). The UC is based on clinical, electrocardiographic (ECG), and pathophysiologic characteristics compared to the STC, which is mainly ECG based. In this registry of consecutive patients with AMI presenting to a tertiary hospital, we studied the inter-physician reliability [weighted kappa (wK)] using the UC and the STC. Two physician investigators independently classified each patient with AMI according to the UC and STC, and a third senior physician investigator resolved any disagreement. The study included Type 1=226 (89.7%), Type 2=16 (6.3%), Type 3=3 (1.2%), Type 4a=1 (0.4%), Type 4b=4 (1.6%), Type 5=2 (0.8%), ST-segment-elevation AMI (STEMI)=140 (55.6%), and non-ST-segment-elevation AMI (NSTEMI)=112 (44.4%). Inter-physician reliability using the UC was very good (wK=0.84, 95% CI 0.68-0.99) and using the STC was good (wK=0.78, 95% CI 0.70-0.86). Of patients with Type 1 AMI, 57.1% were STEMI and 42.9% were NSTEMI. In contrast, of patients with Type 2 AMI, 18.8% were STEMI and 81.2% were NSTEMI. The UC is a reliable method to classify patients with AMI and performs better than the STC in this study. Validation of the two classifications should be performed in large prospective studies.