Prediction of congestive heart failure in patients with non valvular atrial fibrillation.

OBJECTIVE: Patients with atrial fibrillation (AF) have an increased risk of congestive heart failure (CHF) as well as ischemic stroke. The aim of this study was to investigate the clinical predictors of CHF in patients with non-valvular AF (NVAF).

METHODS: Three hundred and forty-seven patients (derivation cohort) with NVAF were retrospectively evaluated between 2004 and 2005. The associations between potential risk factors and CHF were tested using a Cox proportional hazards analysis, and a risk score for predicting CHF was created. The model was then validated in 161 patients (validation cohort) enrolled between 2008 and 2010. During the follow-up period, 41 patients in the derivation cohort developed CHF requiring hospitalization due to New York Heart Association (NYHA) class III or IV disease. Four independent risk factors were identified, each of which was assigned a number of points as follows: Age ≥72 years (1 point), heart rate ≥80 bpm (1 point), hypertension (1 point), and a previous history of congestive heart failure (2 points). The patients were grouped into one of three risk categories according to the calculated risk score (ARC2H score): low risk (0 points), intermediate risk (1-3 points) and high risk (4-5 points).

RESULT: In the derivation cohort, the annual rates of CHF in these risk categories were 0%, 2.5% and 18% per year respectively. In the validation cohort, the corresponding rates were 0.8%, 8% and 35% per year respectively.

CONCLUSION: A simple clinical risk score, the ARC2H score, was developed to predict CHF in patients with NVAF and validated in an independent cohort.