Skip to main content

Table 4 Subgroup and sensitivity analyses for hospital mortality

From: Association between time of discharge from ICU and hospital mortality: a systematic review and meta-analysis

Analysis

Studies, n

Odds ratio (95% CI)

P heterogeneity

I 2

Study reference

Subgroup analysis

 The definition of night

  18:00–06:00

3

1.36 (1.29, 1.43)

0.198

38.30%

[9, 15, 30]

  18:00–07:59

3

1.30 (1.15, 1.46)

0.432

0.00%

[10, 12, 29]

  19:00–07:59

1

1.29 (1.14, 1.46)

  

[31]

  19:00–06:59

1

1.05 (0.64, 1.70)

  

[13]

  22:00–07:30

1

1.7(1.03, 2.9)

  

[7]

  21:00–06:59

1

1.22 (1.10, 1.36)

  

[11]

  16:00–08:00

1

1.11(0.93, 1.31)

  

[14]

  20:00–07:59

1

1.70 (1.28, 2.25)

  

[6]

  22:00–06:59

2

1.30 (0.96, 1.76)

0.209

36.80%

[5, 8]

 Geographic region

  Oceania

6

1.35 (1.31, 1.39)

0.457

0.00%

[7–10, 15, 30]

  Europe

4

1.33 (1.08, 1.63)

0.039

64.00%

[5, 6, 14, 29]

  North America

4

1.24 (1.16, 1.33)

0.82

0.00%

[11–13, 31]

 Total discharge number

   ≤ 10000

4

1.56 (1.32, 1.84)

0.785

0.00%

[6, 7, 10, 29]

   > 10000

10

1.29 (1.23, 1.36)

0.091

39.90%

[5, 8, 9, 11–15, 30, 31]

 Study design

  Multicenter studies

8

1.30 (1.23, 1.38)

0.067

47.00%

[5, 9, 11, 14, 15, 29–31]

  Single-center studies

6

1.38 (1.20, 1.59)

0.268

22.10%

[6–8, 10, 12, 13]

 Whether or not adjusted for severity of illness at the time of ICU discharge

  YES

3

1.26 (1.02, 1.57)

0.296

17.80%

[13, 15, 29]

  NO

11

1.31 (1.25, 1.38)

0.076

40.90%

[5–12, 14, 30, 31]

 Whether or not adjusted for treatment limitation orders

  Yes

6

1.28 (1.15, 1.43)

0.152

38.10%

[7, 13–15, 29, 30]

  No

8

1.32 (1.23, 1.42)

0.116

39.40%

[5, 6, 8–12, 31]

 whether or not adjusted for premature discharge

  Yes

2

1.31 (0.94, 1.84)

0.2

39.20%

[5, 7]

  No

12

1.31 (1.25, 1.38)

0.09

37.70%

[6, 8–15, 29–31]

 Sensitivity analysis

  Fixed-effects model

14

1.33 (1.30, 1.37)

0.105

33.80%

[5–15, 29–31]

  Random-effects model

14

1.31 (1.25, 1.38)

0.105

33.80%

[5–15, 29–31]

 One-study-out method

  Santamaria et al./2015

1

1.31 (1.24, 1.39)

  

[15]

  Azevedo et al./2015

1

1.31 (1.25, 1.37)

  

[31]

  Gantner et al./2014

1

1.31 (1.24, 1.38)

  

[30]

  Laupland et al./2011

1

1.30 (1.21, 1.39)

  

[29]

  Singh et al./2010

1

1.32 (1.25, 1.39)

  

[10]

  Hanane et al./2008

1

1.31 (1.25, 1.38)

  

[13]

  Laupland et al./2008

1

1.32 (1.25, 1.39)

  

[12]

  Pilcher et al./2007

1

1.31 (1.24, 1.37)

  

[9]

  Tobin et al./2006

1

1.29 (1.22, 1.36)

  

[8]

  Priestap et al./2006

1

1.33 (1.26, 1.40)

  

[11]

  Duke et al./2004

1

1.32 (1.25, 1.38)

  

[7]

  Uusaro et al./2003

1

1.31 (1.24, 1.38)

  

[14]

  Beck et al./2002

1

1.31 (1.24, 1.38)

  

[6]

  Goldfrad et al./2000

1

1.33 (1.27, 1.39)

  

[5]

Back to article page

Critical Care

ISSN: 1364-8535