Prediction of late mortality after myocardial infarction from variables measured at different times during...

Area of Science:

• Cardiology
• Internal Medicine
• Clinical Prognostics

Background:

• Acute myocardial infarction (AMI) poses significant long-term mortality risks.
• Accurate prognostication is crucial for patient management post-discharge.
• Identifying early predictors can aid in risk stratification and intervention.

Purpose of the Study:

• To evaluate the prognostic importance of variables collected at different time points during hospitalization for AMI.
• To determine if adding data from later in the hospital course improves prediction of 1-year cardiac mortality.
• To identify key factors for early risk stratification in AMI patients.

Main Methods:

• Discriminant function analysis was used to identify predictors of 1-year cardiac mortality.
• Data were analyzed from 818 patients discharged after AMI.
• Variables were assessed from initial history, first 24 hours, entire hospitalization, and at discharge.

Main Results:

• Early factors (within 24 hours) predicting death included blood urea nitrogen, prior MI, age, abnormal apex, and sinus bradycardia.
• Adding data from later hospitalization or discharge did not significantly improve predictive accuracy.
• Left ventricular ejection fraction and complex ventricular arrhythmias were not independent predictors in subgroups.
• Prediction models correctly identified 55-60% of deaths and 79-81% of survivors.

Conclusions:

• Prognosis after AMI can be accurately predicted early in the hospital course.
• Late-stage hospitalization data may be redundant for prognostic evaluation.
• Early identification of high-risk patients (25% of cohort) is feasible, with a 28-30% mortality prediction.