Assessing Artificial Intelligence-Powered Responses to Common Patient Questions on Radiofrequency Ablation and Cryoanalgesia for Chronic Pain

Background: Radiofrequency ablation (RFA) and cryoanalgesia are minimally invasive procedures used when conservative medical management fails and there are positive diagnostic blocks. Although both radiofrequency ablation (RFA) and cryoanalgesia are straightforward techniques, the increasing workload of physicians is leading to less time available for patient interaction, including addressing questions about indications, feasibility, long-term effectiveness, and potential complications. Generative artificial intelligence (AI) chatbots, such as ChatGPT, have the potential to reduce physician burden and enhance patient education. Methods: An expert panel compiled thirteen frequently asked questions about RFA and cryoanalgesia, which were subsequently submitted to ChatGPT-4.0. The AI-generated responses were evaluated by 41 participants, including pain physicians, healthcare professionals, and non-healthcare individuals. The Likert scale was used for evaluating the responses, focusing on reliability, accuracy, and comprehensibility using predefined acceptability thresholds. Results: Across all 13 questions, ChatGPT’s responses demonstrated high overall reliability, with a mean score of 4.9 ± 0.7. The mean accuracy score was 2.6 ± 0.3, suggesting alignment with evidence-based standards. Comprehensibility was rated at 2.7 ± 0.2 on average. Pre-procedural questions achieved the highest scores, while post-procedural questions posed more challenges for ChatGPT. Conclusions: ChatGPT demonstrated potential as an adjunct tool for patient education on RFA and cryoanalgesia. Improvements in procedural specificity and medical accuracy are needed before routine clinical implementation.