Patient Parameters
Eight variables, measured within 24 hours of admission
Clinical
Recorded to a ceiling of 90 in the source cohort
Echocardiography
Peak early diastolic mitral inflow
Normal below 34
Posterior wall, diastole
Tissue Doppler, septal mitral annulus
Tissue Doppler, lateral tricuspid annulus
Medical Disclaimer: For research and educational purposes only. This model has not been externally validated outside the development centre, and it is not FDA, EMA or TGA cleared. It must not be used for standalone clinical decision-making. Always consult a qualified cardiologist.
Risk Assessment
Time to event survival prediction
About this Model
A parsimonious Gradient Boosted Cox model using the eight highest ranked predictors from a 59-variable model, trained on 1,568 post-MI patients (Royal Brisbane and Women's Hospital, pre-2018).
Validated on an independent temporal holdout of 1,634 patients (2018 onwards), achieving a C-index of 0.858, not significantly different from the full 59-variable model.
Absolute probabilities come from the Breslow baseline hazard estimated on the training cohort, so no information from the validation cohort enters the prediction.
Explore other models
AF Post MI
New onset atrial fibrillation after myocardial infarction
Estimates the risk of developing atrial fibrillation in the months and years after an MI, to help decide who warrants extended rhythm surveillance.
6 variables · C-index 0.83 · n = 2,596
Both calculators come from the same group at Royal Brisbane and Women's Hospital and run entirely in your browser.