Abstract
Reliable prediction of atrial fibrillation (AF) progression from paroxysmal to non-paroxysmal form of AF could help in optimizing follow-up and decision-making regarding the rate and rhythm control management. The HATCH score and others have been proposed to identify AF patients likely to progress to sustained forms of AF, but external validation is limited. We aimed at evaluating these scores in a large series of AF patients and to identify possible factors leading to persistent/permanent AF.
Methods
All patients with AF seen over a period of 10 years were identified in a database and followed up for AF progression and mortality. Predictors of outcomes were identified using Cox regression model. The values of HATCH, CHA2DS2-VASc, ALARMEc, APPLE, MB-LATER scores and CHARLSON comorbidity index were evaluated with C statistics for prediction of AF progression.
Results
Among 8962 patients (71±14 years), 4991, 476 and 3495 had paroxysmal, persistent AF, and permanent AF, respectively. During a follow-up of 927±1084 days, 404 paroxysmal AF patients progressed to persistent or permanent AF (yearly rate of 3.0%). Progression was associated with a trend toward increased cardiovascular mortality. Independent predictors of AF progression were heart failure (hazard ratio (HR) 2.07; 95% CI 1.50–2.85, p<0.0001), valvular disease (HR=1.87, 95% CI=1.35–2.58, p=0.0002) and the use of digoxin (HR=2.39, 95% CI=1.75–3.29, p<0.001). Conversely, a history of stroke was associated with a lower rate of progression (HR=0.50, 95% CI=0.28–0.88, p=0.02). Overall, most of the score, particularly the HATCH score, were modest predictors of progression (table). The best score was actually the CHARLSON comorbidity index. The predictive values of all scores were better on patients not treated with antiarrhythmic agent.
C-statistics for AF progression All patients (n=4,991) p value* HATCH 0.576 (0.562–0.590) <0.0001 CHA2DS2-VASc 0.532 (0.503–0.560) <0.0001 ALARMEc 0.634 (0.607–0.660) 0.03 APPLE 0.631 (0.605–0.658) 0.03 MB-LATER 0.612 (0.587–0.638) 0.001 CHARLSON 0.667 (0.640–0.693) – *DeLong test vs CHARLSON comorbidity index.
Conclusion
Most of scores have a modest predictive value to identify the risk of evolution to permanent AF. Among them, the CHARLSON comorbidity index had the best predictive value and outperformed other tools. This suggests that AF progression may actually reflect global aging both at the individual and local atrial level.
Age-adjusted Charlson-comorbidity Index Predicts Short-term and Long-term Outcomes after Hepatic Resection of Hepatocellular Carcinoma