Prognostic Impact of the Atrial Fibrillation Burden in Patients With Pulmonary Embolism

Background The prognostic implications of atrial fibrillation (AF) burden in pulmonary embolism (PE) remains unclear. This study aimed to investigate the relationship between AF burden and prognosis in PE patients. Methods In this retrospective cohort study using the Medical Information Mart for Intensive Care IV (MIMIC‐IV v2.0) database, patients with confirmed PE diagnosis were included. AF burden was calculated as the percentage of total AF duration relative to total records duration using nurse‐documented continuous electronic monitoring (CEM) records. Patients with AF were dichotomized into low‐burden and high‐burden groups using an optimal cut‐off value of 12.0%. Multivariable Cox proportional hazards models were employed to evaluate the 1‐year all‐cause mortality risk across sinus rhythm (SR), low‐burden, and high‐burden AF groups. Results Among 1,175 patients, 246 (20.9%) had AF, comprising 84 cases of low‐burden AF and 162 cases of high‐burden AF. The 1‐year all‐cause mortality rates were 22.1% (SR), 26.2% (low‐burden AF), and 40.7% (high‐burden AF), respectively. After full adjustment, high‐burden AF demonstrated a significant association with 1‐year all‐cause mortality (HR adjusted : 1.67, 95% CI: 1.22–2.29, p = 0.001), whereas low‐burden AF showed no statistically significant association (HR adjusted : 0.97, 95% CI: 0.62–1.53, p = 0.906). The robustness of these associations was maintained in subgroup and sensitivity analyses. Restricted cubic spline (RCS) analysis identified a nonlinear relationship between AF burden and 1‐year all‐cause mortality risk, demonstrating an inverted U‐shaped dose‐response pattern ( P for non‐linearity = 0.022). Conclusions High AF burden independently predicted 1‐year all‐cause mortality in PE patients, suggesting its clinical utility for risk stratification.