The Effects of including a replacement questionnaire in the first follow-up to an industrial population