A computer-generated list was created by the blocks-of-12 randomization method.
We used randomization with blocks of random sizes, and stratified by type of disease (i.e. Crohn's disease or ulcerative colitis/indeterminate colitis).
Randomization by permuted blocks of random length, stratified by Workers Compensation claim (yes vs. no), and use of opioids.
The allocation schedule was generated by the Biometrics Department of ReSearch Pharmaceutical Services' coding system and contained blocks of six randomization numbers three assigned to each treatment arm in random order.
Blocks of four randomization numbers were allocated to each treatment centre.
One pharmacist dispensed the testosterone undecanoate or placebo according to the blocks of four randomization list.
The allocation ratio will be 1 1 in blocks of 2. Randomization will be performed at a site remote from trial location.
All patients were randomly allocated to receive parecoxib or ketorolac or placebo by block of 6 randomization using a computer generated random number (http://www.randomizer.org/).org/
Treatment was assigned randomly to each subject on a 1 1 basis in blocks of six and randomization was stratified within each clinical site.
A computer random number generator was used to select random permuted blocks of size 4. Randomization will be stratified by the diabetes treatment regimens of participants at the time of enrollment (diet only, or diet plus diabetes medications).
