[bhc3]

A metric that may be useful: fragility index. How many positive results would need to be reversed to have the p-value exceed 0.05? Seems applicable for low-powered studies such as this one.

Details: https://emcrit.org/pulmcrit/fragility-index-ninds/