BACKGROUND
Evidence-based decision-making (EBDM) is crucial in modern dentistry. Yet, navigating the vast and ever-evolving scientific literature can be daunting. Large language models (LLMs) like ChatGPT4 could revolutionize EBDM by analyzing vast data and extracting key information, significantly reducing time spent searching for evidence.
OBJECTIVE
This observational study aims to explore ChatGPT4's potential in dental EBDM, examining its capabilities, strengths, and limitations.
METHODS
Two independent dentists engaged in interactive sessions with ChatGPT4, simulating real-life clinical scenarios and seeking scientific information. ChatGPT4's responses were evaluated for accuracy, relevance, efficiency, actionability, and ethics using the ChatGPT4 Response Scoring System (CRSS) and ChatGPT4 Generative Ability Matrix (C-GAM) systems.
RESULTS
ChatGPT4 consistently performed across all five clinical scenarios, achieving a C-GAM score of 46.4% and a CRSS score of 12 out of 28. It successfully identified relevant evidence sources and provided concise summaries, potentially saving valuable time and improving access to information. However, a critical limitation was its inability to generate web links to relevant articles.
CONCLUSIONS
ChatGPT4 shows promise as an AI tool for EBDM in dentistry. Further development and training can address current limitations and enhance its effectiveness. However, clinicians retain ultimate responsibility for informed decisions, requiring expertise and critical evaluation of presented evidence.
CLINICALTRIAL
Nil