Retrospective and Prospective Methods of Studying Association in Medicine

Publisher Website
Google Scholar
Abstract
The classification of epidemiologic surveys into retrospective and prospective types is criticized because (a) the word "prospective" is unsuitable on etymological grounds; (b) the classification is incomplete, since it does not include surveys confined to the here and now; and (c) the classification is not based on the method of sampling, which is the critical feature in determining what conclusions may validly be drawn from the data. An account is given of 3 types of sampling commonly used in surveys of the association between a disease and a putative cause of the disease. The uses and limitations of each method are discussed and illustrated. The conclusion is drawn that the various methods are complementary.
Keywords