"Randomisation bias" in the medical literature: A review