These two factors come together in working out the sample size of a trial.

It's crucial that a trial includes enough participants to be able both to show a significant result and have enough power to detect any difference between the groups. Under-powered research is a significant source of bias in published studies.

Trials should report a power calculation, done in advance of the research to work out how many patients are needed in the study to show significant results. This calculation is a combination of:

Rare events and/or small differences in event rates require a larger sample size.