Abstract
BACKGROUND: Assessing individuals' risk of developing incident atrial fibrillation (AF) is important for making preventive and screening strategies. OBJECTIVES: The performance of the mC(2)HEST score for predicting incident AF has scarcely been evaluated, especially in a multi-ethnic population. METHODS: Participants from the MESA (Multi-Ethnic Study of Atherosclerosis were enrolled in the present study, which involved population of different ethnicities (Caucasian, African-American, Chinese-American, and Hispanic) aged between 45 and 84 from 6 communities in the United States. The discriminative and calibration performance of the mC(2)HEST score was compared with other risk models. RESULTS: A total of 4,524 subjects (mean age 60.2 ± 9.5 years; 53.0% female) were included; 565 (mean age 67.0 ± 7.9 years; 46.5% female) developed AF during 13.6 ± 2.5 years of follow-up, with an incidence of 0.93%/year. The mC(2)HEST score had good prediction at 10 years (C-index, 0.72; 95% CI: 0.701 to 0.753), and 15 years (0.773, 95% CI: 0.749 to 0.798). The risk of incident AF increased with higher mC(2)HEST score points and risk groups (log-rank P < 0.001). The mC(2)HEST score showed positive net reclassification indexes (0.057, 0.090, 0.128, and 0.143) and integrated discriminative improvement (3.2%, 3.9%, 5.7%, and 4.9%) compared with C(2)HEST, HAVOC, HATCH, and CHA(2)DS(2)-VASc scores, respectively. Optimal calibration was seen in the mC(2)HEST score (P = 0.41). CONCLUSIONS: The mC(2)HEST score is a practical model for predicting individuals' risk of incident AF that may be used for guiding AF surveillance, resource allocation, and screening strategies.