[Ethical problems in randomized clinical studies from the aspect of statistics. Part II: when should patient randomization be stopped of modified?].

The ethical responsibility to the randomized study participants demands that the results be monitored during the course of the trial. The interim evaluation of the trial results may lead to the premature stopping. However, the repeated statistical testing inflates the probability of false positive error. Hence statistical adjustments are necessary. The probabilistic rules for the random allocation of patients may be changed during the course of the trial according to the observed responses.