Predicting Mortality in Patients Hospitalized With Acute Myocardial Infarction: From the National Cardiovascular Data...

Area of Science:

• Cardiology
• Health Services Research

Background:

• In-hospital mortality risk prediction is crucial for quality assessment and patient prognostication.
• Contemporary risk models are needed for acute myocardial infarction (MI) due to evolving patient demographics and treatments.

Purpose of the Study:

• To develop and validate a contemporary risk-standardized model for predicting in-hospital mortality in acute MI patients.
• To create a simplified risk score for individual patient stratification.

Main Methods:

• Utilized data from 313,825 acute MI hospitalizations (2019-2020) from the National Cardiovascular Data Registry Chest Pain-MI Registry.
• Developed a risk-standardized model using stepwise logistic regression with 1000 bootstrapped samples on a 70% development cohort.
• Considered 23 patient characteristics at presentation for model inclusion and developed a simplified risk score.

Main Results:

• The final model included 14 variables, identifying out-of-hospital cardiac arrest, cardiogenic shock, and ST-segment elevation MI as key predictors.
• The model demonstrated excellent discrimination (C-statistic=0.868) and good calibration across subgroups.
• A simplified risk score (0-25) correlated with mortality risk, ranging from 0.3% to 49.4%.

Conclusions:

• A contemporary risk model accurately predicts in-hospital mortality for acute MI patients.
• The model is suitable for hospital quality risk standardization and bedside patient prognostication.
• The simplified risk score facilitates individual risk stratification at the point of care.