Table 2.
Median proportion of incompletely reported efficacy and harm outcomes per trial, stratified by trial characteristics, among 519 randomised trials published in December 2000 and cited in PubMed
Efficacy outcomes
|
Harm outcomes
|
|||
---|---|---|---|---|
Trial characteristic | No of trials | Median % of incompletely reported outcomes per trial* | No of trials | Median % of incompletely reported outcomes per trial* |
All trials | 505 | 42 | 308 | 50 |
Parallel group trials | 375 | 22 | 237 | 25 |
Crossover trials | 110 | 100 | 62 | 82 |
Other study designs | 20 | 59 | 9 | 80 |
General medical journal | 37 | 40 | 26 | 58 |
Specialty journal | 468 | 43 | 282 | 47 |
Full industry funding† | 163 | 46 | 133 | 56 |
Partial or non-industry funding† | 290 | 42 | 142 | 27 |
10-90th centile ranges were 0-100% for all median percentages except for efficacy outcomes in crossover trials (48-100%).
Trials with unknown funding sources were excluded.