When using CaTS to find the optimal two-stage design, make sure the disease model is specified in such a way that the target power for the two-stage design is less than or equal to the power in the one-stage design. If a one-stage design is selected that provides 80% power, optimizing the two-stage design may produce a message indicating that ‘The requested power cannot be achieved for your sample size and disease model’. This is because the one-stage power has been rounded to 80% but is in fact slightly lower. Relaxing your model parameters slightly whilst maintaining a one-stage rounded 80% power value should solve the problem.
