Cardiothoracic & vascular anesthesia
Machine Learning Models of Postoperative Atrial Fibrillation Prediction After Cardiac Surgery
Authors' conclusion
In the authors' study, the support-vector machine model was the best predictor; it may be an effective tool for predicting atrial fibrillation after cardiac surgery.
Study context
- Population
- Patients undergoing valve and/or CABG surgery with CPB, Second Affiliated Hospital of Zhejiang University School of Medicine, September 2013–December 2018
- Participants
- 1,400 patients
- Exposure
- Support-vector machine and gradient-boosting decision tree vs logistic regression for predicting postoperative AF
- Outcome
- Predictive performance of machine learning vs logistic regression models for AF after cardiac surgery
Methodological detailssingle-centre, machine learning vs logistic regression
- Single-centre
- Machine learning vs logistic regression
Main finding
As reported in the abstract
AUC for postoperative AF prediction
| Group | AUC |
|---|---|
| Support-vector machine | 0.777 (95% CI: 0.772-0.781) |
| Logistic regression | 0.767 (95% CI: 0.762-0.772) |
| Gradient-boosting decision tree | 0.765 (95% CI: 0.761-0.770) |
AUC
Support-vector machine0.777 (95% CI: 0.772-0.781)
Logistic regression0.767 (95% CI: 0.762-0.772)
Gradient-boosting decision tree0.765 (95% CI: 0.761-0.770)
Observational data: associations, not effects.
Further findings
Incidence of postoperative AF
- Postoperative AF occurred in 519 patients
- 37.1%
Clinical practicability (decision curve analysis)
- All models achieved appropriate net benefit
Original publication
Machine Learning Models of Postoperative Atrial Fibrillation Prediction After Cardiac Surgery. J Cardiothorac Vasc Anesth. Published 25 November 2022. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →