In European heart journal. Quality of care & clinical outcomes
AIMS : Atrial fibrillation (AF) is a well-known risk factor for heart failure (HF). We sought to develop and externally validate a risk model for new-onset HF admission in patients with AF and those without a history of HF.
METHODS AND RESULTS : Using two multicenter, prospective, observational AF registries, RAFFINE (2,857 patients, derivation cohort) and SAKURA (2,516 patients without a history of HF, validation cohort), we developed a risk model by selecting variables with regularized regression and weighing coefficients by Cox regression with the derivation cohort. External validity testing was used for the validation cohort.Overall, 148 (5.2%) and 104 (4.1%) patients in the derivation and validation cohorts, respectively, developed HF during median follow-ups of 1,396 (interquartile range [IQR]: 1,078-1,820) and 1,168 (IQR: 844-1,309) days, respectively. In the derivation cohort, age, hemoglobin, serum creatinine, and log-transformed brain natriuretic peptide were identified as potential risk factors for HF development. The risk model showed good discrimination and calibration in both derivations (area under the curve [AUC]: 0.80 [95% confidence interval (CI) 0.76-0.84]; Hosmer-Lemeshow, P = 0.257) and validation cohorts (AUC: 0.78 [95%CI 0.74-0.83]; Hosmer-Lemeshow, P = 0.475).
CONCLUSION : The novel risk model with four readily available clinical characteristics and biomarkers performed well in predicting new-onset HF admission in patients with AF.
Ishii Kai, Matsue Yuya, Miyauchi Katsumi, Miyazaki Sakiko, Hayashi Hidemori, Nishizaki Yuji, Nojiri Shuko, Saito Yuki, Nagashima Koichi, Okumura Yasuo, Daida Hiroyuki, Minamino Tohru
2022-Dec-20
Atrial fibrillation, heart failure, machine learning, prediction model, risk score
