Skip to main content

Table 1 Data characteristics: The first column shows the number of affirmative cases for binary fields and the number of unique values for multivalue categorical fields. The second column shows the number of admissions with a valid value (e.g., if height is missing, it is deemed invalid). Empty cells denote N/A

From: Development and validation of ‘Patient Optimizer’ (POP) algorithms for predicting surgical risk with machine learning

 

# Affirmative / # Categories

# Valid values

Median

Mean (SD)

Demographics

    

M

6,234 (54.33%)

11,475 (100.00%)

  

F

5,241 (45.67%)

11,475 (100.00%)

  

Age

 

11,475 (100.00%)

62.00

59.24 (17.81)

Height

 

4,958 (43.21%)

167.00

166.01 (14.63)

Weight

 

6,920 (60.31%)

81.00

84.13 (21.46)

Derived features

    

BMI

 

4,931 (42.97%)

29.38

32.10 (14.09)

Procedure-risk

 

11,475 (100.00%)

1.00

1.63 (0.71)

Patient-risk

 

11,475 (100.00%)

3.00

4.06 (3.13)

Other

    

Emergency procedure

1,416 (12.34%)

11,475 (100.00%)

  

Categorical fields

    

Procedures

911

   

Medication

923

   

Pathology

75