When assessing the risk of bias in studies included in your review, you have to address the issue of incomplete outcome data. Loss to follow up is a key factor that may lead to incomplete outcome data. But how do we know when the loss to follow up is too high?

In this brief training video, Dr Gordon Guyatt from the McMaster University explains how to judge the risk of incomplete outcome data in dichotomous outcomes by looking at the loss to follow up in relation to the number of events.