TABLE 3.

Statistical considerations to maximize information from rare disease cohorts

Stopping rules	Run study to completion, if no excess inhibitor incidence  Assess inhibitor rate after 20 subjects complete follow-up  If ≥10 subjects develop inhibitors, halt trial, as>30%
Dynamic randomization	Base sample size on number of events  After four inhibitors, randomization favours better arm 2:1, 3:1  Lessens inhibitor risk, preserves statistical integrity
Historic controls	Reduce sample size by baseline data from contemporary controls  Use surveillance registries: CDC, international registry for controls  Employ prospective cohorts, observational studies for controls
Master protocol	Follow all enrolled PUPs and screen for inhibitor development  Establish baseline pre-inhibitor data and specimens  Establish inhibitor natural history among screened subjects

CDC, Centers for Disease Control; PUP, previously untreated patient.