EBRO 2017

Sample size

• A clinician’s decision • meaningful difference

• risks of error to be accepted • type I error :  • type II error : β (1 – power) • To be calculated BEFORE +++++ • a non feasible trial ? • an underpowered trial ?