Generative-AI as a source of caregiving guidance for medical tourists: A content and readability analysis

Medical tourism (MT) caregiving companions are often expected to navigate unfamiliar healthcare systems, manage high-stress situations, and bridge linguistic and cultural divides on behalf of medical tourists. Despite these responsibilities, the informational and educational needs of these surrogate caregivers remain largely underexplored.

Given the limited availability of clear and credible guidance for MT caregiving companions, this study examines generative-AI-produced caregiving guidance, evaluating its readability, evidentiary transparency, and topical coverage to assess its relevance for future human-curated educational resources.

Using a pre-scripted, sequential prompting design, researchers queried three widely used generative-AI platforms-ChatGPT-4, Gemini 1.0, and Copilot in March 2024. A thematic qualitative content analysis was conducted to evaluate generated caregiving guidance, including readability, evidentiary support, and description of key caregiving responsibilities.

GenAI MT caregiving guidance frequently exceeded recommended readability levels for general audiences and often lacked citations or clear evidentiary grounding. However, across platforms, responses consistently addressed a broader range of caregiving considerations with a high degree of convergence on the topics of logistics and planning, cultural navigation, patient advocacy, and continuity of care. CONCLUSIONS FINDINGS: indicate that GenAI frames medical tourism caregiving as a complex, multidimensional role, yet platforms differed in emphasis, provided inconsistent evidentiary support, and largely omitted relational aspects such as trust and surrogate decision-making. These results suggest that GenAI guidance should not be used as stand-alone health education but may inform resource development when critically evaluated and human-curated. Practice Implications RESULTS HIGHLIGHT: the need for targeted, plain-language educational resources for MT caregiving companions and greater transparency regarding the limitations, sourcing, and appropriate use of AI-generated health information. Findings also underscore practical and ethical considerations for integrating generative AI into future patient and caregiver education initiatives.