Key result
ChatGPT-4 generates comprehensive responses to ~93% of achalasia questions and improves reading levels when simplified.
Why the study?
Achalasia has multiple diagnostic and therapeutic options, and the accuracy, comprehensiveness, and readability of ChatGPT-generated responses to patient questions about it were unknown.
ChatGPT-4 can generate accurate, comprehensive, and readable responses to patient questions about achalasia, serving as a potential adjunct to physician-led patient education.
Answers are grounded in this paper and cite their sources.
May support AI-assisted achalasia education; leaves open prospective validation before clinical adoption.
Background: Achalasia is a rare esophageal motility disorder characterized by impaired lower esophageal sphincter relaxation and aperistalsis, with multiple diagnostic and therapeutic options. ChatGPT, an artificial intelligence (AI)-powered large language model, has demonstrated utility as a patient resource across multiple medical and surgical domains. This study evaluated the accuracy, comprehensiveness, and readability of ChatGPT-generated responses to frequently asked patient questions about achalasia. Methods: Fifty-nine frequently asked achalasia questions were compiled from institutional patient education websites and the achalasia patient community. Each question was entered into ChatGPT (GPT-4; OpenAI) in April 2025 in a separate session, generating an initial response that was then simplified at a lower reading level. Responses were independently rated by two fellowship-trained foregut surgeons, with adjudication by a gastroenterologist with advanced training in esophagology. Initial responses were rated as comprehensive, correct but inadequate, mixed, or completely incorrect. Simplified responses were rated as less, equally, or more accurate and comprehensive than the originals. Readability was assessed using seven validated formulas. Results: Of 59 initial responses, 55 (93%) were rated as comprehensive. Simplified responses were equally accurate in 93% (n = 55) and equally comprehensive in 90% (n = 53) of cases. Initial responses corresponded to a college-to-graduate reading level, whereas simplified responses improved significantly to an eighth- to twelfth-grade reading level ( P < .001). Conclusion: ChatGPT-4 generated accurate and comprehensive responses to patient questions about achalasia and reliably simplified them while preserving content fidelity. ChatGPT may serve as a useful adjunct to physician-led patient information, though appropriate prompting strategies and clinical oversight remain necessary for responsible implementation.
No takes yet. Share an insight, caveat, or question.
Akhnoukh et al. (2026) studied Achalasia (n=59). ChatGPT-4 generated responses vs. Simplified responses was evaluated on Comprehensiveness of initial responses. ChatGPT-4 generated comprehensive initial responses to 93% of patient questions about achalasia, and simplified responses improved reading level.
Synapse has enriched 10 related papers. Consider them for comparative context: