Skip to main content

Table 1 Characteristics of included variables

From: Development and validation of inpatient mortality prediction models for patients with hyperglycemic crisis using machine learning approaches

Type

Variable

Missing

Complete rate

Median

P25

P75

Hist

Factor

Sex

0

1

-

-

-

▇▁▁▁▇

Numeric

Age(years)

0

1

57

46

69

▂▅▇▆▁

Numeric

Length of stay(days)

0

1

9

6

13

▇▁▁▁▁

Factor

Type of diabetes

0

1

-

-

-

▁▁▁▁▇

Numeric

Course of diabetes

0

1

4

1

10

▇▂▁▁▁

Factor

Infection

0

1

-

-

-

▇▁▁▁▃

Factor

Mechanical ventilation

0

1

-

-

-

▇▁▁▁▁

Factor

Hypoglycemia

0

1

-

-

-

▇▁▁▁▁

Factor

Hypokalemia

0

1

-

-

-

▇▁▁▁▃

Factor

MSOF

0

1

-

-

-

▇▁▁▁▁

Factor

AKI

0

1

-

-

-

▇▁▁▁▁

Numeric

First 24 h insulin dosage

0

1

42

30

58

▃▇▂▁▁

Numeric

First 24 h infusion volume

0

1

3300

2100

5200

▇▇▃▂▁

Numeric

BMI

0

1

22.90

21

24.60

▂▇▂▁▁

Numeric

CCI

0

1

3

2

4

▅▇▂▁▁

Factor

Hyperglycemic crisis type

0

1

-

-

-

▇▁▁▁▁

Numeric

Glucose level before treatment

0

1

23

18.10

30.50

▇▅▁▁▁

Numeric

HbA1c

228

0.863

11.30

9.80

13.10

▂▇▇▃▁

Numeric

pH

0

1

7.29

7.21

7.30

▁▁▁▁▇

Numeric

ABE

424

0.746

-8.85

-16.60

-3.10

▂▆▇▃▁

Numeric

AB

478

0.713

15.20

10

20.30

▂▇▁▁▁

Numeric

AG

480

0.712

17

13.80

21.20

▆▇▃▁▁

Numeric

Serum creatinine

132

0.921

66.10

51

94.10

▇▁▁▁▁

Numeric

Serum sodium

0

1

138

135

141

▁▇▆▁▁

Numeric

Serum potassium

0

1

4.04

3.70

4.50

▁▇▅▁▁

Numeric

EPOP

1

0.999

306

299

315

▁▇▂▁▁

Factor

Inpatient death

0

1

-

-

-

▇▁▁▁▁

  1. MSOF: multiple system organ failure; AKI: acute kidney injury; BMI: body mass index; CCI: charlson comorbidity index; ABE: actual base excess; AB: actual bicarbonate; AG: anion gap; EPOP: Effective plasma osmotic pressure