The E value method deals with unmeasured confounding, a key source of bias in observational studies. The E value method is described and its use is shown in a worked example of a meta-analysis examining the association between the use of antidepressants in pregnancy and the risk of miscarriage.
