The randomization-induced risk of a trial failing to attain its target power: assessment and mitigation