| Makale Türü | Özgün Makale (Ulusal alan endekslerinde (TR Dizin, ULAKBİM) yayınlanan tam makale) | ||
| Dergi Adı | Gümüşhane Üniversitesi Sağlık Bilimleri Dergisi | ||
| Dergi ISSN | 2146-9954 | ||
| Dergi Tarandığı Indeksler | TR DİZİN | ||
| Makale Dili | Türkçe | Basım Tarihi | 12-2025 |
| Cilt / Sayı / Sayfa | 14 / 4 / 1372–1380 | DOI | 10.37989/gumussagbil.1791138 |
| Makale Linki | https://doi.org/10.37989/gumussagbil.1791138 | ||
| UAK Araştırma Alanları |
Protetik Diş Tedavisi
|
||
| Özet |
| Purpose This study aims to evaluate the information accuracy and reliability of ChatGPT's answers to frequently asked questions about prosthetic dentistry by patients and prosthodontists. Methods A total of 40 questions (20 patient-level and 20 professional-level) were submitted to ChatGPT-3.5. The responses were independently evaluated by two academic prosthodontists and two clinical prosthodontists using a 5-point Likert scale (1 = very poor to 5 = very good). Inter-rater agreement was assessed using Fleiss’s and Cohen’s Kappa statistics. Differences between evaluator groups and question types were analyzed using Mann–Whitney U and Wilcoxon signed-rank tests. Results According to the mean score of all specialists, the responses to patient-level questions were significantly higher than the responses to professional-level questions (p<0,001).There was no significant difference (p=0,876) between academics and clinicians in terms of patient-based questions. For prosthodontist-based questions, the mean scores evaluated by academics were significantly higher than those of clinicians. (p<0,01). Conclusions ChatGPT indicates promising reliability in answering common patient researches and prosthodontics questions. However, for advanced clinical questions, its performance is limited and should be supplemented by expert consultation. ChatGPT can be a useful tool for patient guidance but is not yet a substitute for professional expertise in prosthodontics. |
| Anahtar Kelimeler |