Lupus & antiphospholipid syndrome
ChatGPT4’s proficiency in addressing patients’ questions on systemic lupus erythematosus: a blinded comparative study with specialists
Authors' conclusion
ChatGPT4 demonstrated comparable, possibly better in certain domains, to address FAQs from patients with SLE, when compared with the answers provided by specialists.
Study context
- Population
- 95 frequently asked questions (FAQs) from patients with SLE (55 in Chinese, 40 in English); responses scored by a panel of rheumatologists and a group of patients with SLE
- Participants
- 95 FAQs
- Exposure
- ChatGPT4-generated responses vs responses from five rheumatologists to SLE patients' FAQs
- Outcome
- Scores in six domains (scientific validity, logical consistency, comprehensibility, completeness, satisfaction level, empathy) on a 0-10 scale
Methodological detailsblinded, comparative
- Blinded
- Comparative
Main finding
As reported in the abstract
Rheumatologists' scoring (0-10 scale)
ChatGPT4 vs RheumatologistsMean difference (95% CI)P value
Satisfaction level
0.537(0.252–0.823)
P < 0.01Empathy
0.460(0.227–0.693)
P < 0.01Observational data: associations, not effects.
Further findings
SLE patients' scoring, all six domains
- ChatGPT4 responses comparable to rheumatologist-provided answers in all six domains
Subgroup analysis by language
- ChatGPT4 more logically consistent and complete regardless of language
- ChatGPT4 greater comprehensibility, satisfaction and empathy in Chinese
- ChatGPT4 inferior in comprehensibility for English FAQs
Original publication
ChatGPT4’s proficiency in addressing patients’ questions on systemic lupus erythematosus: a blinded comparative study with specialists. Rheumatology (Oxford). Published 22 April 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →