Table 2 Results of quality assessment of included studies

Randomisation sequence generation:

 - Adequate

3

 - Inadequate (quasi-RCT)

9

 - Unclear

76

 - Cluster design

6

Allocation concealment:

 - Unclear

94

Blinding (any):

 - Yes

5

 - No

89

Sample representativeness:

 - Very likely

22

 - Somewhat likely

47

 - Unlikely or can’t tell

25

Sample size calculated:

 - Yes

3

 - Unclear or No

91

Comparability between groups:

 - Reported no significant differences

87

 - Some difference or can’t tell

7

Contamination between arms:

 - Unlikely

18

 - Somewhat likely

68

 - Very likely

8

Loss to follow up:

 - Reported

22

 - No reported

72

Outcomes reported:

 - Blood pressure

48

 - Hypertension treatment

54

 - Adequate hypertension control

56