Patients were grouped according to Balanced Block Randomization method in four groups A and B, C and D. All four drugs were in boxes A, B, C and D, and only the doctor in charge of the project (supervisor) was aware of the drug content of each group. Then, for use for each patient, numbering was done by the supervisor and what was written in the office. For example, patient number 1 of drug A and patient number 2 of drug B or vice versa, how to blind it is described below.
The main idea of block randomization was to divide patients into M blocks of size 2N, so that in each block N patients were assigned A and N patients were assigned to B. The block was then randomly selected. This method ensures equal treatment allocation per block provided the block is fully utilized.
For example: two treatments A, B and block size 2 × 2 = 4
Assignment of treatment is possible within each block
(1) AABB (2) BBAA, (3) ABAB, (4) BABA, (5) ABBA, (6) BAAB
The size of the block, depending on the number of treatments, should be short enough to prevent imbalance, and large enough to prevent guessing treatment allocation in each group during the study. The block size should be at least twice the number of treatment nodes. The size of the block was stated in the study so that researchers would be blind to it.
If the blocks were expressed, the treatment series in each block could be guessed. For example, in block 2N = 4, A A B must be B and in A A as B B can be deduced.
This could lead to (selection bias). The solution to prevent this error was to: (1) not reveal the block mechanism. (2). Use random block size.