EBRO 2017

• Lack of power

• small studies / small effects • repeated tests • Bias (≠ chance variability) • patient selection • flexibility (design, outcome, analysis) • selective reporting / reading

Ioannidis PLOS Med 2005