Skip to main content

Table 4 Logistic regression models on the subgroup admitted after elective cardiac surgery

From: Impact of chloride and strong ion difference on ICU and hospital mortality in a mixed intensive care population

 

Outcome: 30-day mortalitya (n = 2350)

Outcome: hospital mortalityb (n = 2156)

n

% outcome (%)

Odds ratio (95 % CI) versus normochloremia

p value

n

% outcome (%)

Odds ratio (95 % CI) versus normochloremia

p value

Chloride category

 Normochloremia

287

5.6

  

246

13.4

  

 Hypochloremia

21

23.8

3.49 (0.62–19.62)

0.16

15

40.0

1.45 (0.30–7.18)

0.64

 Moderate hyperchloremia

596

3.0

0.63 (0.26–1.54)

0.31

550

5.1

0.43 (0.22–0.83)

0.01

 Severe hyperchloremia

1446

2.7

0.57 (0.22–1.44)

0.23

1345

4.5

0.37 (0.18–0.73)

0.004

 

n

% outcome (%)

Odds ratio (95 % CI) versus mean SID

p value

n

% outcome (%)

Odds ratio (95 % CI) versus mean SID

p value

SIDa category

 Intermediate SIDa

941

2.6

  

873

5.2

  

 Low SIDa

1110

2.0

1.46 (0.68–3.11)

0.34

1015

4.1

1.13 (0.63–2.03)

0.68

 High SIDa

229

10.7

0.85 (0.39–1.83)

0.67

268

14.9

0.71 (0.38–1.31)

0.27

 

Area under ROC 88.3 %

  

Area under ROC 83.9 %

Maximal VIF 3.53

 

Maximal VIF 3.58

Tolerance 0.66

 

Tolerance 0.66

  1. SIDa apparent strong ion difference (excl. lactate), ROC Receiver under Operating Characteristics Curve, VIF variance inflation factor
  2. aConfounders adjusted for in the model: sodium (high vs. normal, low vs. normal*), SAPS-3, RIFLEcrea, lactate, heart failure*, COPD*, pCO2* (low, normal, high), albumin*, SIG, potentiometry type (*p > 0.05). RIFLEurine omitted because of collinearity
  3. bConfounders adjusted for in the model: sodium (high vs. normal, low vs. normal*), SAPS-3, RIFLEcrea, lactate, heart failure*, COPD, pCO2* (low, normal, high), albumin*, SIG, potentiometry type (*p > 0.05). RIFLEurine omitted because of collinearity