Skip to main content

Table 2 Associations between limitations on life support on ICU admission and patient and hospital characteristics. Adjusted odds ratio (aOR) and 95% confidence interval (95% CI)

From: Limitation of life support techniques at admission to the intensive care unit: a multicenter prospective cohort study

 

Limitations, n = 238

No limitations, n = 2804

Model A aOR (95% CI)

Model B aOR (95% CI)

Patient characteristics

 Age

73.0 ± 13.5

61.6 ± 16.0

1.04 (1.03–1.06)

1.05 (1.03–1.06)

 Female sex

106 (44.5%)

1000 (35.6%)

1.30 (0.93–1.81)

1.30 (0.92–1.78)

 Reason for ICU admission

    

  Other

  Sepsis

  Coma or encephalopathy

  Worsening of chronic disease

77 (32.3%)

33 (13.9%)

61 (25.6%)

67 (28.1%)

1816 (64.8%)

377 (13.4%)

326 (11.6%)

285 (10.2%)

1

0.94 (0.57–1.57)

3.96 (2.50–6.30)

2.34 (1.50–3.66)

1

0.97 (0.58–1.62)

3.88 (2.45–6.13)

2.34 (1.50–3.66)

 Predicted risk of death (%)

46.3 [24.0–63.9]

12.0 [5.1–29.0]

1.03 (1.02–1.04)

1.03 (1.02–1.04)

Prior functional Knaus status:

 Class A

 Class B

 Class C

 Class D

37 (15.5%)

96 (40.3%)

76 (31.9%)

29 (12.2%)

1708 (60.9%)

848 (30.2%)

211 (7.5%)

34 (1.2%)

1

3.80 (2.44–5.92)

13.44 (8.00–22.58)

36.94 (17.34–78.71)

1

3.71 (2.39–5.77)

13.30 (7.93–22.32)

36.77 (17.29–78.20)

Hospital characteristics

 Intermediate care unit available

  Yes

  No

43 (18.1%)

195 (81.9%)

826 (29.4%)

1978 (70.5%)

 

1

1.85 (1.00–3.44)

 Patients with limitations on life support outside the ICU

  Yes

  No

143 (60.1%)

95 (39.9%)

1928 (68.7%)

876 (31.2%)

 

1

2.57 (1.45–4.57)

 Hospital variance (SE)

 LR test; p value

  

0.765 (0.271)

54.38; p < 0.001

0.453 (0.191)

24.69; p < 0.001

Intraclass correlation coefficient

 

0.189

0.121

Median odds ratio (95% CI)

 

2.30 (1.59–2.96)

1.90 (1.31–2.38)

  1. Mean ± standard deviation; n (row %); median [interquartile range]; SE standard error
  2. Model A—random effects multilevel logistic regression model with hospital as a second-level variable (random effect) and the patient characteristics as first-level variables
  3. Model B—random effects multilevel logistic regression model with hospital as a second-level variable (random effect) and the patient and center characteristics as first-level variables