Evaluating the reliability of cardiovascular risk scales in patients with chronic inflammatory rheumatic diseases

Objective

To compare the performance of the QRESEARCH risk estimator version 3 (QRISK3), the Systematic COronary Risk Evaluation (SCORE) 2, and Predicting Risk of cardiovascular disease EVENTs (PREVENT) equationin a cohort of individuals with chronic inflammatory rheumatic diseases (CIRD) enrolled in the Spanish prospective CARdiovascular in RheuMAtology (CARMA) project.

Methods

Between July 2010 and January 2012, the study recruited CIRD patients from 67 hospitals across Spain. It included individuals diagnosed with rheumatoid arthritis, ankylosing spondylitis, and psoriatic arthritis. At the 10-year follow-up, data for all patients included in the initial cohort were assessed. We estimated four 10-year cardiovascular disease (CVD) incidence risk scores using data recorded at recruitment.

Results

2080 patients were included in this analysis. QRISK3 and PREVENT-CVD predicted an average of approximately 10 % CV events across the entire cohort, while SCORE2 and PREVENT-Atherosclerotic Cardiovascular Disease (ASCVD) predicted an average of only 6.3 %. The linear correlation coefficients between each pair of scales were consistently above 0.8, with an average of 0.9074. Notably, lower correlations were observed between QRISK3 and the other scales. When identifying patients with higher CV risk, the kappa index was higher between SCORE2, PREVENT-CVD, and PREVENT-ASCVD than between QRISK3 and any other scale. These findings suggest that most patients identified as high-risk by SCORE2 would also be classified as high-risk when using PREVENT-CVD or PREVENT-ASCVD.

Conclusions

The higher correlation and reliability observed between SCORE2, PREVENT-CVD, and PREVENT-ASCVD in our series of CIRD patients followed over a 10-year period suggest that these scales may be largely interchangeable for identifying high-risk CIRD patients.