Will ChatGPT-4 improve the quality of medical abstracts?

Background: ChatGPT received attention for medical writing. Our objective was to evaluate whether ChatGPT 4.0 could improve the quality of abstracts submitted to a medical conference by clinical researchers.

Methods: This was an experimental study involving 24 international researchers (the participants) who provided one original abstract intended for submission at the 2024 Pediatric Academic Society (PAS) conference. We asked ChatGPT-4 to improve the quality of the abstract while adhering to PAS submission guidelines. Participants received the revised version and were tasked with creating a final abstract. The quality of each version (original, ChatGPT and final) was evaluated by the participants themselves using a numeric scale (0-100). Additionally, three co-investigators assessed abstracts blinded to the version. The primary analysis focused on the mean difference in scores between the final and original abstracts.

Results: Abstract quality varied between the three versions with mean scores of 82, 65 and 90 for the original, ChatGPT and final versions, respectively. Overall, the final version displayed significantly improved quality compared to the original (mean difference 8.0 points; 95% CI: 5.6-10.3). Independent ratings by the co-investigators confirmed statistically significant improvements (mean difference 1.10 points; 95% CI: 0.54-1.66). Participants identified minor (n = 10) and major (n = 3) factual errors in ChatGPT's abstracts.

Conclusion: ChatGPT 4.0 does not produce abstracts of better quality than the one crafted by researchers but it offers suggestions to help them improve their abstracts. It may be more useful for researchers encountering challenges in abstract generation due to limited experience or language barriers.