Table 3.
Subgroup analysis for association between reporting of trial registration and positive study findings
| Funding source | Any registration v not registered | Prospective registration v not registered | Retrospective registration v not registered | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| RCTs with positive findings/all RCTs (%) | Adjusted risk ratio (95% CI) | P value interaction | RCTs with positive findings/all RCTs (%) | Adjusted risk ratio (95% CI) | P value interaction | RCTs with positive findings/all RCTs (%) | Adjusted risk ratio (95% CI) | P value interaction | |||
| Non-industry funded | Registered: 122/264 (46.2) not registered§: 155/249 (62.2) | 0.75 (0.63 to 0.89) | 0.046 | Registered: 37/79 (46.8) not registered: 155/249 (62.2) | 0.75 (0.58 to 0.97) | 0.110 | Registered: 85/185 (45.9) not registered: 155/249 (62.2) | 0.75 (0.62 to 0.90) | 0.074 | ||
| Solely or partly industry funded | Registered: 127/237 (53.5) not registered: 35/65 (53.8) | 1.03 (0.79 to 1.36) | Registered: 63/117 (53.8) not registered: 35/65 (53.8) | 1.06 (0.78 to 1.45) | Registered: 64/120 (53.3) not registered: 35/65 (53.8) | 1.02 (0.76 to 1.35) | |||||
RCT=randomised controlled trial.