This made me wonder how often reviewers actually complete a withdrawal during testing. Do they explain when they could not?
