ChatGPT and Gemini as Tools for Parental and/or Caregiver Guidance about Baby Oral Health

• Grant number: 24/16912-3
• Support Opportunities: Scholarships in Brazil - Master
• Start date: July 01, 2025
• End date: October 31, 2026
• Field of knowledge: Health Sciences - Dentistry - Pediatric Dentistry
• Principal Investigator: Tamara Kerber Tedesco
• Grantee: Bruna Bergamo Campoi
• Host Institution: Faculdade de Odontologia (FO). Universidade de São Paulo (USP). São Paulo , SP, Brazil

Abstract

The baby's first visit to the pediatric dentist should take place within the first year of life. During this visit, parents or caregivers should receive guidance to ensure the healthy development of the baby's oral cavity, as well as clarifying any doubts the family may have. With the advance of technology and the emergence of artificial intelligence (AI) programs, many parents and caregivers have sought information through these tools, such as ChatGPT and Gemini. Although the answers are instantaneous, there are concerns about the origin of the information and whether it is in fact based on the best scientific evidence. Therefore, the aim of this study is to analyze and compare the information provided by ChatGPT and Gemini on the main questions parents and/or caregivers have about their baby's oral health, as well as to evaluate the source of the data provided by the programs. The questions will be obtained through a website that generates frequently asked questions. In addition, three panels bringing together participants from different countries - Brazil, the United States and the Dominican Republic - will be formed to collect the main questions from parents or caregivers, and dentists will be able to add questions based on their daily experiences. The questions will be entered into the ChatGPT and Gemini programs, and the answers generated will be analyzed for readability, usefulness, quality, reliability, comprehensibility and action of the information. Finally, for each answer generated, the source of the data will be questioned. Continuous quantitative variables will be subjected to Linear Regression, while ordinal categorical variables will be analyzed using Ordinal Logistic Regression (¿=5%). (AU)