Leaving nothing to chance: how to randomize a clinical trial

L. Toltl,D. Arnold
DOI: https://doi.org/10.1111/j.1537-2995.2012.03730.x
2012-12-01
Transfusion
Abstract:W hen you toss a coin, the chances of landing on heads or tails are theoretically equal with each toss. After 500 tosses, you would expect to have landed an equal number of each. However, this might not always be the case; for example, if you toss the coin only 20 times, you may end up with a disproportionate number of heads or tails. Similar to tossing a coin, in a randomized controlled trial (RCT) the assignment of patients to one group or the other is random; and since this is such an integral part of any RCT, ensuring the integrity of the randomization process must not be left to chance.
What problem does this paper attempt to address?