2.1

Study population

This is a retrospective cohort study of 348 consecutive patients from a single tertiary hospital with a discharge diagnosis of AMI from December 31, 2004, until December 31, 2006, who met the study criteria. Patients were followed for a minimum of 24 months for clinical outcomes until December 31, 2007. We performed a retrospective classification of AMI based on the 2007 Universal Definition consensus document .

2.2

Inclusion and exclusion criteria

We included male and female patients, age >30 years old, who met the UC criteria for AMI , had angiographic documentation of ≥50% obstructive coronary artery disease (CAD), and completed at least 24 months of follow-up. We excluded patients: (a) with elevations of cTn in the absence of overt ischemic heart disease , (b) with metastatic cancer, (c) on comfort care only, or (d) refusing standard care for AMI.

2.3

Specific aims and study hypothesis

We aimed to study the long-term outcomes of patients with AMI according to the UC and STC and determine whether patients with ST segment elevation acute myocardial infarction (STEMI) and non-ST-segment elevation acute myocardial infarction (NSTEMI) have different survival compared to Type 1 and Type 2 AMI. We tested the hypothesis that the long-term outcome of patients with AMI is better predicted by the new UC than the traditional STC.

2.4

Primary and secondary study outcomes

The primary outcome was major adverse cardiovascular events (MACE) (composite of all causes of mortality and recurrent nonfatal AMI). The secondary outcomes were the individual components of MACE.

2.5

Study follow-up and outcomes adjudication

Trained hospital personnel used the American College of Cardiology National Cardiovascular Data Registry’s instrument for patients with AMI ( http://www.accncdr.com/WebNCDR/Common ). The date of death was confirmed by cross-checking the National Death Index (NDI) database ( http://www.cdc.gov/nchs ), Social Security Death Index (SSDI) Web site ( http://ssdi.rootsweb.com ), and by reviewing death notes in the medical record. To ensure we captured all patients who died, we waited until December 31, 2008, to cross-check all patients in these databases and adjudicate the death outcome. Only one patient in our study did not have a social security number and was lost to follow-up. The date of the recurrent AMI was confirmed by cross-checking the institutional roster list of patients admitted with AMI and the billing records. Two physician investigators independently reviewed the medical records to adjudicate the primary outcomes, complete the case report form, and classify the type of AMI. When disagreement existed, a third physician investigator resolved the conflict.

2.6

Definition and classification of AMI

AMI was defined according to the 2007 Universal Definition of AMI . The UC defines AMI when there is evidence of myocardial necrosis (rise and/or a fall of troponin with at least one value above the 99th percentile of the upper reference limit, >0.001 μg/L) in the clinical setting consistent with myocardial ischemia where any of the following apply: (a) symptoms of ischemia, (b) ECG changes indicative of ischemia, (c) development of pathologic Q waves on the ECG, or (d) image with evidence of new loss of viable myocardium or new regional wall motion abnormality.

The UC defines several types of AMI as follows, Fig. 1 : Type 1 as “spontaneous myocardial infarction related to ischemia due to a primary coronary event such as plaque erosion and/or rupture, fissuring, or dissection”; Type 2 as “myocardial infarction secondary to ischemia due to either increased oxygen demand or decreased supply, e.g., coronary artery spasm, coronary embolism, anemia, arrhythmias, hypertension, or hypotension”; Type 3 as “sudden unexpected cardiac death, including cardiac arrest, often with symptoms suggestive of myocardial ischemia, accompanied by presumably new ST elevation, or new left bundle-branch block or evidence of fresh thrombus in a coronary artery by angiography and/or at autopsy, but death occurring before blood samples could be obtained or at a time before the appearance of cardiac biomarkers in the blood”; Type 4a as “myocardial infarction associated with percutaneous coronary intervention (PCI) with increase in cTn greater than 3×99th percentile of the upper reference limit”; Type 4b as “myocardial infarction associated with stent thrombosis as documented by angiography or at autopsy”; and Type 5 as “myocardial infarction associated with coronary artery bypass graft with increase in cTn greater than 5×99th percentile of the upper reference limit.”

Types 3, 4a, 4b, and 5, while identified and tabulated, were not part of the survival analysis because of the paucity of patients in these subgroups. Standard definitions were used for STEMI and NSTEMI according to currently published guidelines .

2.7

Statistical analysis

Continuous variables were summarized as mean±S.D. and compared using the t test, while categorical variables were summarized as frequencies and compared using the chi-square test. Other statistical tests were used as appropriate. All tests were two-sided with P <.05 considered significant. Multivariate adjusted cumulative risk of MACE curves were constructed by the Kaplan–Meier method, and statistical differences between the curves were assessed by log-rank test. Multivariate Cox proportional hazards regression model was used to identify independent predictors of MACE for STEMI, NSTEMI, Type 1, and Type 2 subgroups of AMI. Statistical analysis was performed with SPSS/PC (version 14.0, SPSS, Chicago, IL, USA) software package by an independent statistician.

The study was carried out according to the principles of the Declaration of Helsinki and was approved by the institutional review board. Informed consent was not necessary due to the observational nature of the study. The authors had full access to the data and take full responsibility for its integrity. All authors have read and agree with the manuscript as written .