Table 2. Association between trial characteristics and treatment effect in 84 comparisons.
Potential effect modifier | Negative direction of effect | Small size | Moderate effect size | Large effect size |
OR<1.0 | OR≥1.0 and OR≤1.2 | OR>1.2 and OR≤1.5 | OR>1.5 | |
Implementation received by physician | 12 (16%) | 15 (19%) | 21 (27%) | 29 (38%) |
Implementation received by nurse | 4 (17%) | 4 (17%) | 6 (25%) | 10 (42%) |
Implementation received by other professionals | 1 (4%) | 6 (21%) | 5 (18%) | 16 (57%) |
Primary prevention | 2 (9%) | 5 (24%) | 4 (19%) | 10 (48%) |
Secondary prevention | 11 (18%) | 9 (15%) | 17 (27%) | 25 (40%) |
Tertiary prevention | 3 (14%) | 7 (33%) | 6 (29%) | 5 (24%) |
Design: c-RCT | 10 (17%) | 15 (25%) | 14 (23%) | 21 (35%) |
Design: RCT | 3 (12%) | 2 (8%) | 8 (33%) | 11 (46%) |
Adherence as primary outcome | 5 (12%) | 7 (17%) | 13 (32%) | 16 (39%) |
Adherence as secondary outcome | 8 (19%) | 10 (23%) | 9 (21%) | 16 (37%) |
Follow-up <12 months | 12 (18%) | 12 (18%) | 19 (29%) | 23 (35%) |
Follow-up ≥12 months | 1 (6%) | 5 (28%) | 3 (17%) | 9 (50%) |
Abbreviations: RCT: randomized controlled trial; c-RCT: cluster randomized controlled trial; OR: odds ratio.