Generative artificial intelligence (AI) in preanesthetic consultations: Effects on efficiency, documentation workload, quality, and physician–patient interaction: A simulation trial
Primary result · Consultation durationResult — go to the resultDifference 252 sAuthors' conclusion
LLM-based generative AI-supported documentation significantly improved efficiency and user experience in simulated preanesthetic consultations.
Study context
- Population
- Anesthesiologists at University Hospital Zurich conducting standardized preanesthetic consultations with a simulated patienteach did both consultations
- Participants
- 30 participants
- Comparison
- AI-assisted documentation vs conventional manual documentation
- AIn = 30
- Manualn = 30reference
- Primary endpoint
- Consultation duration
Methodological detailscrossover, randomised case order, single-centre, simulated patient
- Crossover
- Randomised case order
- Single-centre
- Simulated patient
Primary result
As reported in the abstract
Difference
252 s
P < 0.0001
Secondary outcomes
| Outcome | Estimate | P value |
|---|---|---|
| Screen fixation (eye-tracking) | relative change −78% | P = 0.0002 |
| Refixations | relative change −73% | P < 0.0001 |
| Keyboard input | relative change −87% | P < 0.0001 |
| Mouse clicks | relative change −19% | P = 0.01 |
| Documentation quality (PDQI-9, external raters)Higher for manual reports | difference +4 PDQI-9 points | P = 0.004 |
Subjective workload (NASA-TLX)
- Clinicians reported a trend toward lower workload (−16%, p = 0.07)
Patient engagement (self-assessed)
- Clinicians reported better patient engagement (median rating 87 vs. 69)
Confidence in formatting
- Clinicians expressed less confidence in AI-generated formatting
Workflow preference
- Preferred AI assistance overall
- 60%
Original publication
Generative AI in preanesthetic consultations: Effects on efficiency, documentation workload, quality, and physician–patient interaction: A simulation trial. J Clin Anesth. Published 19 February 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →