Paediatric anaesthesia & critical care
Development of an Arterial Carbon Dioxide Estimation Model Using End-tidal Carbon Dioxide Levels during Surgery in the Pediatric Population
Authors' conclusion
The authors' machine learning model accurately estimates intraoperative arterial carbon dioxide partial pressure using noninvasive parameters.
Study context
- Population
- Pediatric patients undergoing surgery, VitalDB registry8,853 paired measurements
- Participants
- 3,586 patients
- Outcome
- Accuracy of estimated arterial carbon dioxide partial pressure (mean absolute error, root mean squared error)
Methodological detailsmachine learning, internal temporal validation, external validation
- Machine learning
- Internal temporal validation
- External validation
Performance
As reported in the abstract
Gradient boosting model (best of 4 algorithms)
Mean absolute error
2.73 mmHg
Root mean squared error
4.13 mmHg
Further findings
Validation of the model
mmHg
- Mean absolute error, external (institutional) validation, 92 pairs
- 3.65
- Mean absolute error, temporal validation, 2,138 pairs
- 3.67
Feature importance (Shapley additive explanations)
- Most impactful features: end-tidal carbon dioxide, body temperature, FiO2, preoperative hemoglobin
Original publication
Development of an Arterial Carbon Dioxide Estimation Model Using End-tidal Carbon Dioxide Levels during Surgery in the Pediatric Population. Anesthesiology. Published 17 June 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →