Symptom Network and Clusters of the Multidimensional Symptom Experience in Patients With Atrial Fibrillation

Abstract

Background The symptom network can provide a visual insight into the symptom mechanisms. However, few study authors have explored the multidimensional symptom network of patients with atrial fibrillation (AF). Objectives We aimed to identify the core symptom and symptom clusters of patients with AF by generating a symptom network. Furthermore, we wanted to identify multiple characteristics related to symptom clusters. Methods This is a cross-sectional study. A total of 384 patients with AF at Tianjin Medical University General Hospital were enrolled. The University of Toronto Atrial Fibrillation Severity Scale was used to assess AF symptoms. Network analysis was used to explore the core symptom and symptom cluster. Results Shortness of breath at rest (r s = 1.189, r c = 0.024), exercise intolerance (r s = 1.116), shortness of breath during physical activity (r s = 1.055, r c = 0.022), and fatigue at rest (r c = 0.020) have the top centrality for strength and closeness. The top 3 symptoms of bridge strength were shortness of breath at rest (r s = 0.264), dizziness (r s = 0.208), and palpitations (r s = 0.207). Atrial fibrillation symptoms could be clustered into the breathless cluster and the cardiac cluster. We have identified multiple factors such as mental health status, left ventricular ejection fraction, heart failure, sex, B-type natriuretic peptide, and chronic obstructive pulmonary disease as significant contributors within the breathless cluster, whereas sex, mental health status, and history of radiofrequency ablation were strongly associated with the cardiac cluster, holding promise in elucidating the underlying mechanisms of these symptoms. Conclusion Special attention should be given to shortness of breath at rest as its core and bridging role in patients' symptoms. Furthermore, both the breathless and cardiac clusters are common among patients. Network analysis reveals direct connections between symptoms, symptom clusters, and their influencing factors, providing a foundation for clinicians to effectively manage patients' symptoms.