Building a Knowledge-Based Tool for Auto-Assessing the Cardiovascular Risk.

Area of Science:

• Cardiology
• Medical Informatics
• Public Health

Background:

• Cardiovascular disease prevention requires accurate risk quantification.
• Existing European SCORE models lack comprehensive cardiovascular risk factors, such as diabetes.
• French national health authorities developed clinical practice guidelines to address these limitations.

Purpose of the Study:

• To formalize knowledge from French clinical practice guidelines for cardiovascular risk assessment.
• To implement these formalized rules into an automated cardiovascular risk self-assessment tool.
• To improve upon existing risk assessment tools by incorporating a wider range of risk factors.

Main Methods:

• Knowledge formalization from guidelines through a five-step process guided by medical experts.
• Development of a decision tree model utilizing eight key decision variables.
• Evaluation of the decision tree's accuracy against expert predictions in various clinical scenarios.

Main Results:

• A formalized decision tree model was created, incorporating eight decision variables.
• The decision tree demonstrated 80% agreement with medical informatics experts in predicting cardiovascular risk.
• Identified knowledge gaps within the Clinical Practice Guidelines contributed to discrepancies.

Conclusions:

• The developed decision tree effectively formalizes French clinical practice guidelines for cardiovascular risk assessment.
• The tool shows high accuracy, comparable to expert judgment, and can be extended for patient self-assessment.
• Future integration into a platform will provide personalized recommendations based on calculated cardiovascular risk.