Impact of informational pollution against the use of fluorides on oral health behavior and dental caries development: An agent-based modeling

• Grant number: 24/12604-2
• Support Opportunities: Scholarships in Brazil - Doctorate
• Start date: December 01, 2024
• End date: November 30, 2027
• Field of knowledge: Health Sciences - Dentistry - Pediatric Dentistry
• Principal Investigator: Thiago Cruvinel da Silva
• Grantee: Olívia Santana Jorge
• Host Institution: Faculdade de Odontologia de Bauru (FOB). Universidade de São Paulo (USP). Bauru , SP, Brazil
• Associated research grant: 23/11709-2 - Dental infodemic: Investigating the consumption, propagation, and psychophysiological impact of false digital information about oral health, AP.R

Abstract

Despite the high potential for the spread of information pollution recommending the cessation of fluoride use for the prevention of dental caries on the Internet, there is still no evidence about the impact such information has on changing oral health behavior among users of digital platforms. Therefore, this study aims to analyze, through agent-based modeling (ABM) simulations, the impact of the consumption of information unfavorable to fluoride use on changing oral health behaviors, as well as the impact of behavior changes on the prevalence and incidence of dental caries. The project will be divided into four subprojects. In the first subproject, the trend of production and consumption of information related to fluoride use over time will be determined on Google, Facebook, and Instagram. The Google Trends will be used to obtain the search volume for fluoride in different countries to determine the interest and consumption of structured information on websites. Meta's Content Library and its API will be used to obtain the number of posts and interaction metrics with content about fluoride use to determine the production, interest, and consumption of unstructured information on social media, considering the polarity of the content (favorable or unfavorable to fluoride use). Social media content will be automatically classified based on an artificial intelligence model previously developed for this purpose (FAPESP #2021/10732-5). In the second subproject, a systematic literature review and possible meta-analyses will be conducted to determine the factors related to the consumption of misinformation and changes in oral health behavior, as well as their respective weights, using the RevMan 5.3 software. PRISMA recommendations and MOOSE guidelines will be followed. The third subproject will evaluate the transformation of information related to fluoride use found on websites, social media, and generative artificial intelligence platforms (AIGs) over time. Two independent investigators will analyze the content of websites and AIGs to determine the prevalence and incidence of unfavorable content towards total content in Portuguese and English, as well as the readability of the information (Flesch Kincaid Reading Ease and Brazilian Portuguese-Flesch Reading Ease) and the quality of the information (JAMA Benchmark and DISCERN). The analyses will consider four different time periods (baseline, 3 months, 6 months, and 12 months), including social media content classified in the first subproject. In the fourth subproject, ABM simulations will be developed considering three theories: (i) The Online Misinformation Engagement Framework, (ii) Information Processing Theory, and (iii) Social Cognitive Theory. The modeling will be developed and validated at two distinct levels and outcomes. At the first level, the outcome of the simulation will be the cessation of fluoride use by agents (yes or no). At the second level, the outcomes of the simulation will be the prevalence and incidence of dental caries in a specific population over time, influenced by the first-level outcomes.