Block randomization random permuted blocks

This is a method to ensure that after the entry of every x patients into the trial, the number of patients on each treatment will be equal; 'x' is referred to as the block size.

 ID no. TRT Patient initials Hospital 1001 3 1002 2 1003 3 1004 3 1005 1 1006 3 1007 3 1008 1 1009 2 1010 2 1011 3 1012 1 1013 1 1014 3 1015 2 1016 3 1017 3 1018 3

For example, in a trial of two treatments with a block size of four, we can ensure that of every four patients, two will receive treatment A and two will receive treatment B. The random element comes from considering all the possible arrangements of, in this example, two As and two Bs within a set of four, and then randomly choosing the order in which the sets are used to form a randomization list. Again using the example above, there are six possible permutations of two treatments (A and B) within a block of four:

To implement blocked randomization, allocate a digit (0-9) to each block, and use random numbers to choose the order of the blocks to be used in the randomization schedule.

For example, given the random number sequence 1843712 ...

^ AABB (ignore 8) BBAA ABBA (ignore 7) AABB ABAB

0 0