In preparing this review, we followed the published methodologic guidelines
for review papers (2). A comprehensive literature search was performed to
identify papers and books on both methodologic theory and application of cancer
screening case–control studies published in the English language before January
1997. Sources of information were identified through personal knowledge contributed
by the co-authors, conferences, Medline searches, the National Cancer
Institute Physicians Database Query (PDQ), and references from published papers
and books. Unpublished work was excluded unless it appeared in conference
proceedings. Narrative techniques are used to summarize the methodologic
theory and application papers reviewed.
We restrict our review to study designs that use mortality from the cancer
under study as the end point. For example, we excluded published studies of
cervical cancer and prostate cancer screening that use an end point of cancer
incidence or advanced disease. The end point of mortality was chosen for simplicity,
since it is the most objective end point, and because it is the primary goal
of screening. Papers included in this review are listed by publication year in
Table 1. The table is organized into 5-year time periods by year of publication
so that methodologic theory and application papers published in the same time
frame are easily identified. For the purposes of classification in Table 1, we list
a paper under application if it includes a case–control study with mortality as the
study end point.
1. Selection of