Evaluation of machine learning models as decision aids for anesthesiologists
Authors' conclusion
ML models can improve anesthesiologists' estimation of clinical parameters.
Study context
- Population
- Practicing anesthesiologists predicting outcomes for actual surgical cases100 surgical patients
- Participants
- 10 participants
- Comparison
- With Machine Learning (ML) assistance vs Without ML assistance
- Outcome
- Accuracy of anesthesiologists' estimates of peak intraoperative glucose and postoperative opioid requirements
Methodological detailsweb-based tool, questionnaire
- Web-based tool
- Questionnaire
Main finding
As reported in the abstract
Accuracy of peak glucose estimates
< 0.001
Further findings
Accuracy of opioid requirement estimates
P < 0.001
Anesthesiologists with improved predictions when ML estimates provided
- Peak glucose predictions improved
- 8 out of the 10 anesthesiologists
- Opioid requirement predictions improved
- 7 of the 10 anesthesiologists
Feedback questionnaire
- ML estimates primarily used as reference to modify clinical judgement
Original publication
Evaluation of machine learning models as decision aids for anesthesiologists. J Clin Monit Comput. Published 9 June 2022. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →