Association of Mental Health, Quality of Life, and Sport Specialization Using the Wisconsin Sport Specialization Questionnaire Among Adolescent Athletes.

Patients increasingly seek health information online before or after clinical encounters. In the United States, the proportion of individuals reporting the internet as their first source of health information increased from 61.2% in 2008 to 74.4% in 2017, with similarly elevated rates reported internationally []. Despite the widespread use of online resources, patients generally maintain trust in clinicians while seeking greater involvement in health decisions []. As digital information–seeking patterns have evolved, a growing proportion of patients are turning to large language models (LLM) and artificial intelligence (AI)‐driven chatbots for guidance on their health [, ]. Early evaluations of LLMs in otolaryngology (ENT) demonstrate their promising capabilities along with their prevailing limitations. Studies suggest that ChatGPT can provide accurate guideline‐based patient education for common ENT conditions [, , ]. However, its performance is less reliable when addressing complex issues, including pharmacologic recommendations, surgical risks, and postoperative management [, ]. LLM‐generated responses also tend to contain incomplete information or advice taken out of context []. A growing body of literature has focused on the accuracy of LLM‐generated medical information; however, less is known about how patients use or perceive these technologies. Evidence from other specialties (urology, internal medicine) suggests that patients view chatbot‐generated medical information as useful and accessible, although clinicians remain their preferred source of medical advice []. Within otolaryngology, patient interactions with LLMs remain poorly characterized. In this study, we surveyed patients presenting to an academic otolaryngology clinic to assess awareness and use of large language models for health information seeking, and to evaluate patient perceptions of their usefulness, accuracy, trustworthiness, and impact on communication with clinicians.