Patient Education and AI-Assisted Complaint Summaries in Endodontic Patients

The goal of this randomized clinical trial is to compare the same-day effects of three standardized patient education methods in adults presenting to an endodontic clinic with toothache. The main questions it aims to answer are: * Does standardized verbal education improve patients' knowledge, reduce antibiotic-related misconceptions, and reduce anxiety? * Does leaflet-based education produce comparable changes in knowledge, antibiotic-related misconceptions, and anxiety? * Does video-based education produce comparable changes in knowledge, antibiotic-related misconceptions, and anxiety? * Does teach-back performance differ according to the method used to deliver patient education? * Can an AI-assisted three-sentence complaint summary be generated reliably and used as a feasible clinician-facing communication support tool? Participants will be randomly assigned to one of three patient education groups. In the verbal education group, participants will receive a standardized 90-second verbal explanation covering warning signs, appropriate home-care measures, avoidance of potentially harmful practices, and appropriate antibiotic use. In the leaflet-based education group, participants will review a written leaflet containing the same core educational messages for 90 seconds. In the video-based education group, participants will watch a short video presenting the same core educational messages for 90 seconds. Knowledge level, antibiotic-related misconceptions, and anxiety will be assessed immediately before and immediately after the educational intervention. After education, participants will also be asked to restate the main messages in their own words using a teach-back procedure. In addition, symptom information collected using standardized prompts will be entered into an artificial intelligence tool to generate a concise three-sentence complaint summary. The feasibility, completeness, and clinician-rated usefulness of these summaries will be evaluated as proc