A group sequential adaptive treatment assignment design for proof of concept and dose selection in headache trials.

摘要: The trial objective was to test whether a new mechanism of action would effectively treat migraine headaches and select dose range for further investigation. motivation group sequential, adaptive, placebo-controlled design (1) limited information about where across the seven doses focus attention, (2) need limit sample size complicated inpatient treatment (3) desire reduce exposure patients ineffective treatment. A based on sequential up down designs developed operational characteristics were explored by simulation. primary outcome headache response at 2 h after Groups four treated two placebo assigned one dose. Adaptive selection rates 60% seen with other treatments. If more than responded, then next lower dose; otherwise, increased. stopping rule least five groups target that ensure selected be statistically significantly (p=0.05) superior placebo. Simulations indicated good in terms control type 1 error, sufficient power, modest expected bias estimation. is attractive phase clinical trials when acute simple, ideally binary, comparator required, patient accrual relatively slow allowing collection processing results as basis adaptive assignment groups. this successful both proof concept selection.