Key result
A logistic regression model using eight-dimensional hand-crafted features predicted post-intubation tachycardia within 10 minutes with an accuracy of 80.5% and an AUROC of 0.85.
Why the study?
Tachycardia frequently occurs after endotracheal intubation and can cause serious complications in cardiovascular disease, making the ability to predict post-intubation tachycardia valuable for notifying clinicians to initiate pre-treatment.
Can machine learning models using pre-intubation electronic medical records and vital signs accurately predict post-intubation tachycardia?
Population
257 patient datasets remaining after filtering 1931 available datasets
Comparison
Five feature sets evaluated across multiple machine learning models
Design
Machine learning prediction and validation study using 10-fold cross validation
Follow-up
10 min
Authors
Loading...
May facilitate risk stratification during intubation; hypothesis-generating and requires prospective validation before clinical adoption.
Can machine learning models using pre-intubation electronic medical records and vital signs accurately predict post-intubation tachycardia?
Effect estimate: AUROC 0.85
A logistic regression model using pre-intubation clinical data can predict post-intubation tachycardia with good accuracy, potentially enabling clinicians to pre-treat at-risk patients.
Kim et al. (2020) studied Post-intubation tachycardia (n=257). Logistic regression model with eight-dimensional hand-crafted features vs. Other machine learning models was evaluated on Prediction of post-intubation tachycardia within 10 min (AUROC 0.85). A logistic regression model using eight-dimensional hand-crafted features predicted post-intubation tachycardia within 10 minutes with an accuracy of 80.5% and an AUROC of 0.85.
Synapse has enriched 3 closely related papers on similar clinical questions. Consider them for comparative context: