Level of response in epidemiologic studies using the card-back system to contact subjects.

Response levels obtained from two methods of patient contact to obtain subject consent for participation in epidemiologic studies were compared. When a signed, return card was required prior to further contact, 70 to 83 per cent of patients consented to participate, compared with consistently higher levels (89 to 93 per cent) obtained with telephone contact. These data suggest that the likelihood of obtaining the high response level required in epidemiologic studies is increased by methods that do not rely on a return card.