Related Experiment Videos
A next-generation prediction risk model for acute myocardial infarction: Derivation and validation in a multi-centre
Jose David Amorocho-Morales1, Sergio Parra Guevara1,2, Elias Quintero-Muñoz1
1Universidad de La Sabana, Chía, Colombia.
Insights
This study developed a new model for predicting short-term acute myocardial infarction (AMI) risk using electronic health records. The model shows strong accuracy and calibration, offering a valuable tool for cardiovascular disease risk stratification.
Area of Science:
- Cardiology
- Medical Informatics
- Public Health
Background:
- Cardiovascular disease is a leading cause of death globally.
- Accurate, event-specific risk prediction is crucial, especially where long-horizon models are less effective.
- Routine electronic health record (EHR) data offers a rich resource for developing predictive models.
Purpose of the Study:
- To develop and validate a probabilistic model for estimating 6- and 12-month risk of acute myocardial infarction (AMI).
- To explore 5- and 10-year risk prediction horizons.
- To utilize routinely collected EHR data from a Colombian cardiovascular cohort.
Main Methods:
- The study analyzed 382,589 patients and 3.9 million encounters, adhering to TRIPOD+AI guidelines.
- A modeling strategy combined a calibrated gradient-boosting classifier with an interpretable survival ensemble (Cox regression, random survival forests, discrete-time hazards).
- Outcomes included prediction accuracy, discrimination, calibration, and concordance with existing risk scores.
Main Results:
- The classifier achieved an Area Under the Curve (AUC) of 0.869.
- 6- and 12-month survival models demonstrated C-indices of 0.836 and 0.846, respectively.
- Strong calibration (Observed/Expected ratio = 0.998) and moderate concordance with legacy scores were observed, indicating significant short-term re-ranking.
Conclusions:
- The developed model serves as a practical tool for population health stratification of short-term AMI risk.
- The model is particularly valuable in resource-constrained settings.
- Recalibration to local incidence rates and prospective evaluation are recommended for deployment.
Aim:
Cardiovascular disease remains the leading global cause of death, and the need for accurate, event-specific risk prediction is particularly critical in regions where long-horizon models perform poorly. We developed and internally validated a probabilistic model to estimate 6- and 12-month risk of acute myocardial infarction, with exploratory 5- and 10-year horizons, using routinely collected electronic health record data from an integrated cardiovascular cohort in Colombia.
Methods:
The study followed TRIPOD + AI guidance and analysed 382,589 patients contributing 3.9 million encounters. The modelling strategy combined a calibrated gradient-boosting classifier with an interpretable survival ensemble incorporating Cox regression, random survival forests, and discrete-time hazards. Primary outcomes were prediction accuracy, discrimination, calibration, and concordance with legacy score scales.
Results:
The classifier achieved an AUC of 0.869, while 6- and 12-month survival models reached C-indices of 0.836 and 0.846. Calibration was strong, with predicted vs observed AMI counts nearly identical (O/E = 0.998). Concordance analyses demonstrated only moderate alignment with Framingham and PROCAM, indicating substantial re-ranking at short horizons compared with legacy long-term models. External, label-delayed validation (n = 5602) showed monotonic risk separation across predefined priority bands.
Conclusion:
This model provides a practical population-health stratification tool for short-term AMI risk, with particular value in resource-constrained settings. Recalibration to local incidence rates is recommended before deployment. Prospective evaluation is warranted to assess real-world clinical and operational impact.
