Using Topic Modeling to Understand Patients'and Caregivers'Perspectives About Left Ventricular Assist Device:ThematicAnalysis

Background: Heart failure (HF) is a significant global clinical and public health challenge, impacting 64.3 million individualsworldwide. To address the scarcity of donor organs, left ventricular assist device (LVAD) implantation has become a crucialintervention for managing end-stage HF, serving as a bridge to heart transplantation or as a destination therapy. Web-based healthforums, such as MyLVAD.com, play a vital role as trusted sources of information for individuals with HF symptoms and theircaregivers.Objective: We aim to uncover the latent topics within the posts shared by users on the MyLVAD.com website.Methods: Using the latent Dirichlet allocation algorithm and a visualization tool, our objective was to uncover latent topicswithin the posts shared on the MyLVAD.com website. Through the application of topic modeling techniques, we analyzed 459posts authored by recipients of LVAD and their family members from 2015 to 2023.Results: This study unveiled 5 prominent themes of concern among patients with LVAD and their family members. Thesethemes included family support (39.5% weight value), encompassing subthemes such as family caregiving roles and emotionalor practical support; clothing (23.9% weight value), with subthemes related to comfort, normalcy, and functionality; infection(18.2% weight value), covering driveline infections, prevention, and care; power (12% weight value), involving challengesassociated with power dependency; and self-care maintenance, monitoring, and management (6.3% weight value), which includedsubthemes such as blood tests, monitoring, alarms, and device management.Conclusions: These findings contribute to a better understanding of the experiences and needs of patients implanted with LVAD,providing valuable insights for health care professionals to offer tailored support and care. By using latent Dirichlet allocation toanalyze posts from the MyLVAD.com forum, this study sheds light on key topics discussed by users, facilitating improved patientcare and enhanced patient-provider communication.