From: Correcting for multiple-testing in multi-arm trials: is it necessary and is it done?
Exploratory trials (n = 20) | Confirmatory trials (n = 39) | All trials (n = 59) | |
---|---|---|---|
Number of study arms | |||
3 | 8 (40%) | 22 (56%) | 30 (51%) |
4 | 3 (15%) | 13 (33%) | 16 (27%) |
> 4 | 9 (45%) | 4 (10%) | 13 (22%) |
Trial design | |||
Parallel group | 19 (95%) | 30 (77%) | 49 (83%) |
a) Different doses or regimens of same treatment | 13 | 16 | 29 |
b) Different treatments | 3 | 8 | 11 |
c) Combined treatments | 3 | 6 | 9 |
Parallel group with factorial | 1 (5%) | 9 (23%) | 10 (17%) |
Comparisons reported (restricted to parallel group trials) | |||
All pairwise | 2/19 (11%) | 4/30 (13%) | 6/49 (12%) |
Multiple experimental versus control | 11/19 (58%) | 19/30 (63%) | 30/49 (61%) |
Experimental versus multiple controls | 1/19 (5%) | 2/30 (7%) | 3/49 (6%) |
Other | 5/19 (26%) | 5/30 (17%) | 10/49 (20%) |
ICD-10 Disease classification | |||
Certain infectious and parasitic diseases | 3 (15%) | 5 (13%) | 8 (14%) |
Neoplasms | 0 (0%) | 5 (13%) | 5 (8%) |
Endocrine, nutritional and metabolic diseases | 4 (20%) | 4 (10%) | 8 (14%) |
Diseases of the nervous system | 1 (5%) | 4 (10%) | 5 (8%) |
Diseases of the circulatory system | 5 (25%) | 4 (10%) | 9 (15%) |
Other | 7 (35%) | 17 (44%) | 24 (41%) |
Allocation ratio | |||
Equal | 18 (90%) | 31 (79%) | 49 (83%) |
Not Equal | 2 (10%) | 8 (21%) | 10 (17%) |
Adjustment procedure | |||
None | 9 (45%) | 21 (54%) | 30 (51%) |
Bonferroni | 2 (10%) | 6 (15%) | 8 (14%) |
Dunnett | 0 (0%) | 3 (8%) | 3 (5%) |
Hochberg | 1 (5%) | 3 (8%) | 4 (7%) |
Hierarchical (or other closed procedure) | 8 (40%) | 6 (15%) | 14 (24%) |
Sample size calculation reported | |||
In report | 16 (80%) | 30 (77%) | 46 (78%) |
Referenced or in supplementary material | 2 (10%) | 6 (15%) | 8 (14%) |
Not given | 2 (10%) | 3 (8%) | 5 (8%) |