Cardiac arrest & resuscitation
Machine Learning Identifies Higher Survival Profile In Extracorporeal Cardiopulmonary Resuscitation*
Authors' conclusion
ML can develop a tiered risk model to guide ECPR patient selection with tailored arrest profiles.
Study context
- Population
- Adults 18–75 years with refractory out-of-hospital cardiac arrest (OHCA) caused by a shockable rhythm, cardiac ICU in a quaternary care center301 underwent extracorporeal cardiopulmonary resuscitation (ECPR)
- Participants
- 376 adults
- Outcome
- Neurologically favorable survival after ECPR (prediction model)
Methodological detailsmachine learning prediction model, internal validation with holdout test set, single-centre
- Machine learning prediction model
- Internal validation with holdout test set
- Single-centre
Main finding
As reported in the abstract
Machine learning (ML) model prediction of neurologically favorable survival
- Neurologically favorable survival with ECPR
- 119 of 301 (40%)
- In-sample AUC
- 0.89
- In-sample misclassification rate
- 0.19
- Out-of-sample validation AUC
- 0.80
- Out-of-sample misclassification rate
- 0.23
Observational data: associations, not effects.
Original publication
Machine Learning Identifies Higher Survival Profile In Extracorporeal Cardiopulmonary Resuscitation*. Crit Care Med. Published 27 March 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →