Identification of Board-Certified Plastic Surgeons Using Artificial Intelligence: An Accuracy Assessment

Area of Science:

• Healthcare technology
• Medical informatics
• Artificial intelligence in medicine

Background:

• Growing interest in cosmetic procedures presents challenges for patients in identifying qualified plastic surgeons due to unclear credentials.
• Artificial intelligence (AI), including large language models, is increasingly utilized in healthcare for various applications.

Purpose of the Study:

• To evaluate the accuracy of AI tools in identifying board-certified plastic surgeons across all 50 US states.
• To address the gap in literature regarding AI's reliability for patient guidance in selecting cosmetic procedure providers.

Main Methods:

• Four AI tools (ChatGPT, Perplexity, Gemini, Claude) were prompted to list board-certified plastic surgeons nationwide.
• Results were cross-verified using independent internet searches and the American Board of Plastic Surgery database.
• Statistical analysis was performed to compare the accuracy of different AI models.

Main Results:

• Overall accuracy in identifying plastic surgeons was 94.1% across 1000 results.
• Gemini demonstrated the highest accuracy, followed by Perplexity, ChatGPT, and Claude, with statistically significant differences.
• The majority of misidentified professionals were otolaryngologists and general surgeons, with no significant difference in error types among AI models.

Conclusions:

• AI tools show promise in assisting patients to find board-certified plastic surgeons amidst a growing aesthetic market.
• Despite high accuracy, AI tools have potential for errors, necessitating independent verification by patients.
• Ensuring patients can make informed decisions requires reliable tools and patient diligence.