Selective screening for atrial fibrillation using multivariable risk models.

Area of Science:

• Cardiology
• Preventive Medicine
• Health Informatics

Background:

• Atrial fibrillation (AF) poses a significant stroke risk if untreated.
• Identifying individuals at high risk is crucial for efficient screening of asymptomatic, paroxysmal AF.
• Current screening often relies on age, but risk models may offer superior discrimination.

Purpose of the Study:

• To compare the effectiveness of existing AF risk models against age for screening purposes.
• To develop and evaluate a novel risk model for AF screening.

Main Methods:

• Nine established AF risk models were evaluated using data from the Atherosclerosis Risk in Communities study (n=11,373).
• A new model, Screening for Asymptomatic Atrial Fibrillation Events (SAAFE), was developed using data from the Monitoring Disparities in Chronic Conditions study (n=3790).
• Performance was assessed by the proportion of patients receiving treatment due to a high CHA₂DS₂-VASc score identified during screening, alongside C-statistic and net benefit.

Main Results:

• Five risk models demonstrated significantly better performance than age alone in identifying individuals who developed AF.
• Age identified 61% of at-risk individuals, while the best-performing risk model identified 82%.
• The SAAFE model identified 81% of at-risk individuals, utilizing age, congestive heart failure, and coronary artery disease.

Conclusions:

• Employing risk models enhances the identification of individuals susceptible to atrial fibrillation.
• The SAAFE model, a high-performing option, does not necessitate an ECG, enabling its use as a cost-effective screening criterion in place of age.