Skip to main content

Table 6 Feature importance ranking of logistic regression

From: A machine learning model to predict the risk of 30-day readmissions in patients with heart failure: a retrospective analysis of electronic medical records data

Rank

Feature description

1

Cumulative number of 30-day readmissions

2

Presence of acute respiratory failure ICD-9 code

3

Number of abnormal sodium level laboratory tests

4

Presence of pneumonia (organism unspecified) ICD-9 code

5

Number of abnormal chloride level laboratory tests

6

Presence of acute kidney failure (unspecified) ICD-9 code

7

Presence of diabetes (with other specified manifestations) ICD-9 code

8

Presence of acute kidney failure (any) ICD-9 code

9

Presence of respiratory failure, WKF

10

Presence of disorders of fluid electrolyte and acid-base balance (any) ICD-9 code

11

Presence of disorders of fluid/electrolyte/acid-base, WKF

12

Number of abnormal albumin level laboratory tests

13

Presence of asphyxia and hypoxemia ICD-9 code

14

Presence of anemia of chronic illness ICD-9 code

15

Presence of hypokalemia ICD-9 code