Look out for intervention comparisons where what happened was not measured in all of the original subjects. All subjects should be followed up.
Subjects in comparisons that are not followed up to the end of the study may show different effects than those that were followed up. For example, they may have been excluded because the action was not working or because of side effects. If those subjects are excluded, the findings of the study may be misleading. This may apply when subjects are marked or tagged and a subset are subsequently undetected or their fate unknown.
REMEMBER: Think about whether lots of subjects who started out in a study comparing treatments were missing in the results.