Table 2 Characteristics of the included trials

N = 230

N = 384

Year of publication, median (IQR)

2003 (1994–2010)

2005 (1996–2011)

Sample size, median (IQR)

276 (146–561.5)

139 (166–284.3)

Number of centers, median (IQR)

13 (3–40.5)

1 (1–3)

Medical fields

 Psychiatry/Addictology

109 (47.4)

184 (47.9)

 Pneumology

15 (6.5)

66 (17.2)

 Ear nose and throat

12 (5.2)

12 (3.1)

 Cardiology

9 (3.9)

15 (3.9)

 Dermatology

6 (2.6)

11 (2.9)

 Pediatrics

11 (4.8)

21 (5.5)

 Infectious diseases

8 (3.5)

4 (1.1)

 Others

60 (26.1)

71 (18.5)

Intervention in experimental group, n (%)

 Pharmacological

79 (34.3)

171 (44.5)

 Non-pharmacological

151 (65.7)

213 (55.5)

Intervention in control group, n (%)

 Active comparator

35 (15.2)

78 (20.3)

 Placebo

29 (12.6)

39 (10.2)

 No added intervention (i.e., usual care)

115 (50.0)

188 (49.0)

 Sham comparator

2 (0.9)

2 (0.5)

 Other

49 (21.3)

77 (20.0)

Outcome objectivity, n (%)

 All-cause mortality

3 (1.3)

32 (8.3)

 Objectively assessed

55 (23.9)

48 (12.5)

 Objectively assessed but influenced by clinician or patient

95 (41.3)

170 (44.3)

 Subjectively assessed

75 (32.6)

132 (34.4)

 Unclear

2 (0.9)

2 (0.5)

Cochrane Risk of Bias tool, n (%)

Random sequence generation

 Low

120 (57.4)

197 (57.4)

 High

11 (5.3)

11 (3.2)

 Unclear

78 (37.3)

135 (39.4)

Allocation concealment

 Low

116 (50.4)

165 (43.0)

 High

21 (9.1)

26 (6.8)

 Unclear

93 (40.4)

193 (50.3)

Blinding of participants and personnel

 Low

68 (45.6)

74 (30.1)

 High

33 (22.2)

62 (25.2)

 Unclear

48 (32,2)

110 (44.7)

Blinding of outcome assessors

 Low

61 (62.9)

83 (43.4)

 High

14 (14.4)

50 (26.2)

 Unclear

22 (22.7)

58 (30.4)

Incomplete data outcome

 Low

135 (64.3)

215 (62.7)

 High

34 (16.2)

53 (15.4)

 Unclear

41 (19.5)

75 (21.9)

Selective reporting

 Low

85 (63.9)

144 (58.5)

 High

9 (6.8)

23 (9.4)

 Unclear

39 (29.3)

79 (32.1)