Novel Risk Stratification Score For Hfpef And Afib: Had-Afib

Introduction Atrial fibrillation (AF) is highly prevalent in patients with heart failure with preserved ejection fraction (HFPEF). Both conditions are associated with shared risk factors including older age, hypertension and diastolic dysfunction. Given the significant phenotypical overlap between these diseases, it is difficult to disentangle outcomes. Therefore, a clinical tool for collective risk stratification is needed. Hypothesis We hypothesized that outcomes in patients with concomitant HFPEF and AF can be reliably stratified using a scoring system. To test this hypothesis, we developed HAD-AFIB, a comprehensive risk-stratification score based on risk factors for clinical use. Methods A random sample of 1,205 cases of concomitant HFPEF and AF was selected out of which 803 cases were used for the development of the predictive score and 402 were designated for validation. Using univariate Cox proportional hazards models, risk factors with significant contribution to mortality and heart failure (HF) hospitalization were identified. Forward stepwise selection along with clinical input was then used to identify multivariable models. Independent predictors for mortality and HF hospitalization were assigned scores based on hazard ratios. The score was then calculated and tested for association and fit in the validation cohort. Results Based on multivariate analyses, Hypertension, Age, Diastolic Dysfunction, Admission for HF, Filtration rate, Ischemic Heart Disease and BMI were selected as outcome predictors. The study cohort characteristics were: mean age at baseline of 74 ± 13 years, 601 (50%) males, 664 (55%) of patients were on two or fewer hypertension medications, 861 (71%) were not on an ACE-I, 439 (36%) had coronary artery disease (CAD), 184 (15%) had GFR less than 40 ml/min, 51 (4%) had BMI less than 20, and 271 (22%) had a prior hospitalization for HF. The median follow-up time was 41 months. Age, hypertension, GFR less than 40 ml/min, BMI less than 20 and prior hospitalization were found to be independent predictors for mortality and HF hospitalization. Diastolic dysfunction and CAD were independent predictors for HF hospitalization only. Based on these variables and tested in the validation cohort, the HAD-AFIB score had a C-statistic of 0.72 (0.67-0.78) for mortality (HR=1.28 (1.20-1.36), p<0.01) and 0.77 (0.70-0.83) for HF hospitalization (HR=1.30 (1.19-1.41), p<0.01) scores respectively. Conclusions Patients with HFPEF and AF can be risk stratified for mortality and heart failure hospitalization using the HAD-AFIB score.