Simulation software applied to short term cardiovascular disease risk assessment

Several risk score tools are available in literature to predict death/myocardial infarction for coronary artery disease (CAD) patients, within a short period of time. However, these tools present some weaknesses: i) eventual lack of performance; ii) inability to deal with missing risk factors; iii) incapacity to incorporate new risk factors (clinical knowledge). This work addressed these problems, proposing a Bayesian classifier strategy through two main phases: i) common representation of individual risk assessment tools; ii) combination of individual models. The validation of the strategy was carried out through the combination of three well known current risk score tools (GRACE, TIMI, PURSUIT) applied to a real patient datasets (Santa Cruz hospital 460 patients). Software was developed to support the proposed methodology. The physician can calculate the CVD risk of a patient as well as perform a set of configurations to adjust the parameters of the global model to a specific population.