Skip to main content

Table 4 Factors associated with death during hospitalization: univariate and multivariate analysis using a Cox regression model (n = 78)

From: Undercorrection of hypernatremia is frequent and associated with mortality

 

Monovariate analysis

Multivariate analysis

 

p

HR

CI 95%

p

Adjusted HR

CI 95%

Age

0.04

1.05

1.00-1.10

   

Male gender

0.06

2.47

0.97-6.28

0.007

5.65

1.60-19.90

Place of origin and personal environment:

      

Home

Ref.

Ref.

Ref.

   

Institution or home medical care

0.65

1.24

0.50-3.08

   

Season of occurrence:

      

Winter

Ref.

Ref.

Ref.

   

Spring

0.36

0.54

0.15-2.00

   

Summer

0.79

0.86

0.29-2.59

   

Autumn

0.21

0.38

0.08-1.75

   

Dementia

0.54

1.40

0.48-4.04

   

Treatment modifying renal water excretion

0.95

1.04

0.32-3.33

   

Characteristics at admission:

      

Systolic blood pressure (mmHg) †

<0.001

0.97

0.95-0.98

   

Diastolic blood pressure (mmHg) †

0.001

0.95

0.92-0.98

   

Mean blood pressure

<0.001

0.95

0.92-0.97

<0.001

0.92

0.91-1.06

Neurologic symptoms

0.03

5.22

1.20-22.78

   

Extracellular dehydration

0.07

4.02

0.92-17.63

   

Natremia (mmol/L)

0.88

1.01

0.94-1.07

0.64

0.98

0.88-0.95

Calculated plasmatic tonicity (mOsm/L) †

0.06

1.02

1.00-1.04

   

Calculated plasmatic osmolarity (mOsm/L)

0.09

1.02

1.00-1.03

   

Creatinine (μmol/L)

0.004

1.01

1.00-1.01

   

eGFR (mL/min/1.73 m2) †

0.03

0.98

0.95-1.00

   

Main diagnosis in emergency department:

    

Extracellular dehydration

Ref.

Ref.

Ref.

 

Infection

0.04

4.21

1.09-16.29

 

Other

    

Perfused solute:

0.18

2.46

0.67-9.10

 

Hypotonic solute

Ref.

Ref.

Ref.

 

Isotonic solute, no perfusion or unknown

0.44

1.43

0.58-3.52

 

Natremia correction speed:

    

Mean speed between entry and last known natremia (mmol/L/h) †

0.10

18.79

0.59-597.13

   

No natremia improvement

0.02

3.12

1.22-7.96

<0.001

10.29

3.12-33.96

  1. Factors indicated by † were not included in the multivariate analysis because a linked factor was already included in the analysis.