Akademik Veri Yönetim Sistemi
Cumhuriyet Dental Journal, cilt.28, sa.4, ss.586-593, 2025 (Scopus, TRDizin)
Objectives: Chatbots powered by artificial intelligence are increasingly used as tools for obtaining medical and dental knowledge. This study aimed to assess and compare the performance of four AI chatbots in providing evidence-based information on pediatric oral surgery topics, with reference to the American Academy of Pediatric Dentistry (AAPD) clinical guideline. Materials and Methods: This descriptive observational study evaluated four AI chatbots (ChatGPT-5, Gemini, Copilot, and DeepSeek) by posing 20 questions derived from the AAPD Guideline on Management Considerations for Pediatric Oral Surgery. Responses were assessed for accuracy using the grading system, for quality using the 16-item DISCERN instrument and for readability using the Flesch-Kincaid Grade Level (FKGL) formula. Non-parametric Kruskal-Wallis and Mann-Whitney U tests with Holm-Bonferroni adjustment were employed for statistical comparisons (p < 0.05). Results: Significant differences were observed among chatbots in all outcome measures. Gemini and ChatGPT-5 achieved the highest accuracy scores (1.30 ± 0.47 and 1.40 ± 0.60, respectively; p = 0.001), whereas DeepSeek and Copilot showed lower accuracy. In terms of information quality, DeepSeek produced the highest DISCERN scores (52.90 ± 3.73; p < 0.001), followed by Copilot. ChatGPT-5 and Gemini yielded more readable outputs (10.73 ± 1.98 and 11.68 ± 1.91, respectively), though readability differences were not statistically significant (p > 0.05). Conclusions: Of the models evaluated, Gemini and ChatGPT-5 produced the most accurate responses, while DeepSeek generated the highest-quality content. While AI chatbots show promise as supplementary tools for patient education and clinical learning in pediatric oral surgery, their reliability varies considerably across platforms. Continuous validation and guideline-based evaluation are essential prior to clinical integration.