Underestimation of Cardiovascular Risk by the SCORE2 Model in Primary Care: A Call for Recalibration.

Area of Science:

• Cardiovascular epidemiology
• Risk prediction modeling
• Primary care research

Background:

• The SCORE2 model estimates cardiovascular event (CVE) risk, intended for primary care use.
• It guides treatment decisions for first-onset CVE in routine practice.

Purpose of the Study:

• To evaluate the SCORE2 model's accuracy in predicting first-onset CVE in Dutch primary care.
• To compare observed CVE risk with SCORE2-predicted risk during patient follow-up.

Main Methods:

• Retrospective cohort study of 205,548 adult patients.
• Analysis included age, sex, smoking, diabetes, cholesterol, and blood pressure.
• Model performance assessed using O/E ratios, calibration plots, and C-statistics, stratified by demographics and risk categories.

Main Results:

• Observed 10-year CVE risk was 10.1%, exceeding the predicted 6.2% (O/E ratios: 1.54-1.68).
• Underestimation was observed across age groups (<50 and ≥50 years) and sexes.
• An estimated 35% of patients may have missed preventive treatments due to SCORE2's underestimation.

Conclusions:

• The SCORE2 model appears to underestimate CVE risk in primary care settings.
• Calibration of the SCORE2 model is recommended for accurate risk assessment in primary care.