New approaches for improving cardiovascular risk assessment

Area of Science:

• Cardiology
• Medical Informatics

Background:

• Clinical guidelines advocate for cardiovascular risk assessment tools (risk scores) to predict cardiovascular death and aid decision-making.
• Current risk scores have limitations impacting their reliability in clinical settings.
• Cardiovascular disease (CVD) poses significant social and economic burdens.

Purpose of the Study:

• To present a novel framework designed to enhance existing cardiovascular risk assessment tools.
• To minimize the limitations of current risk scores and improve their clinical reliability.
• To aid in more accurate prediction of cardiovascular events.

Main Methods:

• Developed a framework combining data from multiple sources, including existing risk scores and physician input.
• Implemented a personalization scheme to group patients and identify the most suitable risk assessment tool for individual cases.
• Applied two distinct methodologies for data processing and patient stratification.

Main Results:

• Validation performed on a dataset of 460 patients with non-ST-segment elevation acute coronary syndrome.
• The proposed approaches demonstrated promising performance metrics, including sensitivity, specificity, and geometric mean.
• Specific performance metrics achieved were 78.79% sensitivity, 73.07% specificity, and 75.87% geometric mean for one approach, and 75.69%, 69.79%, and 72.71% for the other.

Conclusions:

• The developed approaches show superior performance compared to current cardiovascular disease risk scores.
• Further validation with additional datasets is recommended to substantiate these findings.
• The framework offers a potential improvement in predicting cardiovascular risk.