Artificial intelligence in rehabilitation: comparing OpenAI ChatGPT recommendations to evidence-based meta-analyses

Abstract background
ChatGPT, short for Chat Generative Pre-Trained Transformer, is a language model developed by OpenAI that is capable of generating human-like text responses to various prompts. In the field of rehabilitation, ChatGPT has the potential to be a valuable resource for patients and clinicians alike. In light of the open, easy access to this tool, its application in rehabilitation should be considered and evaluated

Aims
This study aimed to test the feasibility of using the OpenAI ChatGPT tool to generate clinical recommendations in rehabilitation and compare its findings with recent meta-analyses.

Material and methods
The selection of rehabilitation areas was based on a report on the global estimates of the need for rehabilitation, with the top three disorders being musculoskeletal disorders, neurological disorders, and chronic respiratory diseases. The authors evaluated the responses of ChatGPT to clinical questions in these areas against the existing evidence on rehabilitation interventions from systematic reviews and meta-analyses in the Cochrane Library database

Results
The findings of the study indicate that the AI's recommendations align with the current scientific knowledge on rehabilitation interventions, as demonstrated through a comparison with the Cochrane Library database.

Conclusions
Using AI language models like ChatGPT can help identify effective rehabilitation methods for common conditions.