Artificial intelligence in airway management: A systematic review and meta-analysis
Authors' conclusion
Our study found that while some AI models have good discriminative ability (AUC ≥ 0.80) for difficult airways, most of them have just average discriminative ability AUC < 0.80.
Study context
- Population
- Patients in the ED or undergoing surgery under GA
- Participants
- 13 studies
- Comparison
- Artificial intelligence (AI) models for prediction of a difficult airway
- Outcome
- Discriminative ability for difficult airway (area under receiver operating curves (AUROC))
Methodological detailsSearched PubMed and Science Direct until March 2025; 2 studies in ED, 11 in surgery under GA
- Searched PubMed and Science Direct until March 2025; 2 studies in ED, 11 in surgery under GA
Pooled result
As reported in the abstract
Discriminative ability for difficult airways (AUC)
| Group | AUC | 95% CI | I2 |
|---|---|---|---|
| VGG | 0.84 | [0.83, 0.84] | 0% |
| SVM | 0.80 | [0.65, 0.96] | 99.7% |
| NB | 0.81 | [0.51, 1.10] | 99.3% |
AUC
VGG0.84
SVM0.80
NB0.81
95% CI
VGG[0.83, 0.84]
SVM[0.65, 0.96]
NB[0.51, 1.10]
I2
VGG0%
SVM99.7%
NB99.3%
Original publication
Artificial intelligence in airway management: A systematic review and meta-analysis. Anaesth Crit Care Pain Med. Published 9 July 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →