Evaluating ChatGPT’s ability to simplify scientific abstracts for clinicians and the public

Abstract

This study evaluated ChatGPT's ability to simplify scientific abstracts for both public and clinician use. Ten questions were developed to assess ChatGPT's ability to simplify scientific abstracts and improve their readability for both the public and clinicians. These questions were applied to 43 abstracts. The abstracts were selected through a convenience sample from Google Scholar by four interdisciplinary reviewers from physiotherapy, occupational therapy, and nursing backgrounds. Each abstract was summarized by ChatGPT on two separate occasions. These summaries were then reviewed independently by two different reviewers. Flesch Reading Ease scores were calculated for each summary and original abstract. A subgroup analysis explored differences in accuracy, clarity, and consistency across various study designs. ChatGPT's summaries scored higher on the Flesch Reading Ease test than the original abstracts in 31 out of 43 papers, showing a significant improvement in readability (p = 0.005). Systematic reviews and meta-analyses consistently received higher scores for accuracy, clarity, and consistency, while clinical trials scored lower across these parameters. Despite its strengths, ChatGPT showed limitations in "Hallucination presence" and "Technical terms usage," scoring below 7 out of 10. Hallucination rates varied by study type, with case reports having the lowest scores. Reviewer agreement across parameters demonstrated consistency in evaluations. ChatGPT shows promise for translating knowledge in clinical settings, helping to make scientific research more accessible to non-experts. However, its tendency toward hallucinations and technical jargon requires careful review by clinicians, patients, and caregivers. Further research is needed to assess its reliability and safety for broader use in healthcare communication.

Ähnliche Arbeiten