Table 4 Comparison between the new model constructed by LR algorithm and other risk scoring models.

From: Development and validation of a machine learning-based readmission risk prediction model for non-ST elevation myocardial infarction patients after percutaneous coronary intervention

Model

AUC

(95%CI)

Z

P value

ML-LR

0.767

(0.743–0.789)

–

–

Adjusted GRACE

0.645

(0.619–0.670)

4.486

 < 0.001

KAMIR

0.678

(0.653–0.703)

3.321

 < 0.001

ACEF

0.591

(0.564–0.617)

6.341

 < 0.001

  1. Adjusted GRACE: adjusted global registry of acute coronary events risk score; KAMIR: The Korea Acute Myocardial Infarction Registry risk score; ACEF: The Age, creatinine, and ejection fraction score.
Back to article page