Table 3 Risk of Bias assessment of the studies by domain
Study type | NON-RANDOMIZED (125 studies) | Risk of bias, n (%) | ||
|---|---|---|---|---|
CRITICAL | MODERATE | LOW | ||
Domain | Confounding bias | 119 (95) | 6 (0.5) | 0 |
Bias in selection of participants | 110 (88) | 15 (1.2) | 0 | |
Bias in classification of interventions | 107 (86) | 18 (14) | 0 | |
Deviation from intended interventions | 92 (74) | 32 (26) | 1 (0.8) | |
Missing data | 0 | 15 (12) | 110 (88) | |
Bias in selection of reported results | 0 | 10 (8) | 115 (92) | |
Study type | RANDOMIZED (9 studies) | CRITICAL | MODERATE | LOW |
Domain | Selection bias | 2 (22) | 5 (55) | 2 (22) |
Performance bias | 2 (22) | 6 (67) | 1 (11) | |
Detection bias | 3 (33) | 5 (55) | 1 (11) | |
Attrition bias | 1 (11) | 6 (67) | 2 (22) | |
Reporting bias | 0 | 2 (22) | 7 (77) | |
Other bias | 0 | 1 (1) | 8 (6) | |