Comprehensive Summary
This study compared ChatGPT-4o and Google Gemini 2.5 to see how well they explain urologic telesurgery, a form of robotic surgery performed remotely. Researchers designed 19 patient-style questions grouped into four categories: prospective (before surgery), technical, recovery, and other. Then, they assessed each chatbot’s answers using two validated tools: DISCERN (which measures quality and reliability, scored out of 80) and PEMAT-P (which measures how understandable and actionable information is, reported as percentages). Gemini outperformed ChatGPT in nearly every area. Its overall DISCERN score was 52.4 / 80 versus ChatGPT’s 45.8 / 80, meaning its responses were thorough (completer) and more trustworthy. In the “prospective,” “technical,” and “recovery” categories, Gemini scored 49.2 vs 39.1, 52.3 vs 44.3, and 53.7 vs 45.4, respectively. They are differences large enough to be statistically significant. Both chatbots met the readability benchmark of ≥ 70%, showing they explained information clearly (Gemini 79.7%, ChatGPT 78.1%). However, their actionability, or how well answers helped patients take next steps, was weak. Only Gemini passed the 70% mark once (73.4% vs ChatGPT’s 65.4%), and both lacked consistent guidance or visual aids. Gemini also cited credible sources like the NIH, Mayo Clinic, and Johns Hopkins in 16 of 19 answers, compared to ChatGPT’s 4, showing greater transparency.
Outcomes and Implications
Overall, both chatbots delivered accurate, intelligible explanations of telesurgery, but neither reliably provided clear, actionable advice. For patients, these AI tools can serve as an accessible way to understand complex procedures. For healthcare professionals, they may complement patient education but should not replace clinician guidance. From a public health perspective, the findings highlight AI’s growing potential to improve health literacy, provided that future models are refined for accuracy, ethical use, and practical application in real world care.