The performance of ChatGPT in day surgery and pre-anesthesia risk assessment: a case-control study of 150 simulated patient presentations
Authors' conclusion
ChatGPT can assist anesthesia practitioners and surgeons by alerting them to the ASA-PS classification and assessing perioperative risk in day surgery patients.
Study context
- Population
- 150 simulated day surgery patient profiles with complications of varying acuity and severity
- Participants
- 150 simulated patient profiles
- Comparison
- ChatGPT vs expert anesthesiologists
- Outcome
- ASA-PS classification and whether day surgery was recommended
Methodological detailssimulated patient presentations, ChatGPT answers graded independently by two experienced anesthesiologists
- Simulated patient presentations
- ChatGPT answers graded independently by two experienced anesthesiologists
Main finding
As reported in the abstract
ASA-PS classification and anesthesia risk assessment
- No difference between ChatGPT and experts for ASA-PS classification and anesthesia risk assessment
- ChatGPT vs Expert: P > 0.05
Observational data: associations, not effects.
Further findings
Recommendation for day surgery (ASA ≥ II) · per_group events/value = day surgery can be performed, n (%)
P < 0.05
Needs further treatment and evaluation ChatGPT 56 (37.3) · Expert 66 (44.0)
Day surgery is not recommended ChatGPT 18 (12.9) · Expert 29 (9.3)
| Outcome | ChatGPT | Expert |
|---|---|---|
| Knowledge of day surgery anesthesia (131 questions)Correct answers | 94.0% | |
| Points considered comprehensive | 70% | — not stated in the abstract |
— not stated in the abstract
Recommendation for day surgery (ASA ≥ II)
- Experts inclined to require further examination or treatment
Knowledge of day surgery anesthesia (131 questions)
- Better performance in peri-anesthesia concerns, lifestyle, and emotional support domains
Original publication
The performance of ChatGPT in day surgery and pre-anesthesia risk assessment: a case-control study of 150 simulated patient presentations. Perioper Med (Lond). Published 21 November 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →