Cardiovascular risk algorithms in primary care: Results from the DETECT study

Area of Science:

• Cardiology
• Preventive Medicine
• Epidemiology

Background:

• Cardiovascular disease (CVD) prevention guidelines rely on risk scores to tailor treatment intensity.
• A comprehensive comparison of commonly used CVD risk scores in a German primary care setting was lacking.

Purpose of the Study:

• To evaluate the correlation, discrimination, and calibration of ten prevalent risk equations in a German population.
• To identify the most accurate risk prediction tool for primary prevention of CVD in Germany.

Main Methods:

• Analysis of 4044 participants from the DETECT study (Diabetes and Cardiovascular Risk Evaluation: Targets and Essential Data for Commitment of Treatment).
• Comparison of ten distinct risk equations assessing their correlation, discrimination, and calibration.
• Evaluation of sensitivity, specificity, and positive predictive value at a 10-year CVD risk threshold of 10%.

Main Results:

• All ten risk equations demonstrated good correlation and similar discriminatory power.
• Significant variation in absolute risk predictions was observed across equations, influenced by endpoint definitions.
• The ACC/AHA atherosclerotic cardiovascular disease (ASCVD) equation exhibited the highest specificity (69%) and positive predictive value (17%) at the 10% risk threshold.
• Three risk scores showed good concordance between calculated and observed risks.

Conclusions:

• The ACC/AHA ASCVD equation offers the most valid risk prediction for primary CVD prevention in Germany.
• Its precise calibration, broad age coverage, and inclusion of fatal and non-fatal events support its utility.
• Further validation of risk scores is crucial for effective CVD prevention strategies.