Comparison of Framingham, PROCAM, SCORE, and Diamond Forrester to predict coronary atherosclerosis and cardiovascular events

Abstract

Background

Cardiologists are often confronted with patients presenting with chest pain, in whom clinical risk profiling is required. We studied four frequently used risk scores in their ability to predict for coronary artery disease (CAD) and major adverse cardiovascular events in patients presenting with stable chest pain at the cardiology outpatient clinic.

Methods and Results

We enrolled 1,296 stable chest pain patients, who underwent cardiac computed tomographic angiography (CCTA) to assess CAD (any, significant: stenosis ≥50%). Framingham (FRS), PROCAM, SCORE risk score, and Diamond Forrester pre-test probability were calculated. All patients were followed up for a mean 19 ± 9 months for all cardiovascular events (mortality, acute coronary syndrome, revascularization >90 days after CCTA). In ROC-analysis for prediction of significant CAD, the areas under the curve for FRS; 0.68 (95% confidence interval: 0.64-0.72) and for SCORE; 0.69 (95% confidence interval: 0.65-0.72) were significantly higher than for PROCAM; 0.64 (95% confidence interval: 0.61-0.68; P ≤ .001), as well as marginally higher than for Diamond Forrester; 0.65 (95% confidence interval: 0.61-0.68; P ≤ .05). Low FRS category showed the lowest number of patients with significant CAD, compared to patients with low risk using PROCAM, SCORE or Diamond Forrester ( P < .001). Also, low FRS category showed less events (compared to PROCAM and SCORE; P < .001, for Diamond Forrester; P = .14).

Conclusion

Our data show that in a stable chest pain population, the ability of FRS and SCORE to predict for CAD was similar and better compared to PROCAM and Diamond Forrester. The number of low risk patients showing significant CAD or events was lower using FRS. Consequently, risk categorization using FRS seems to be safest to stratify stable chest pain patients prior to CCTA.