|
Abstract:
|
Always controversial, cancer screening has come under intense scrutiny in recent years. This is largely due to changes in the way risks and benefits of screening are being measured and prioritized by screening policy panels. Absolute, rather than relative, mortality reductions are now routinely cited to reflect benefit, and overdiagnosis has become the harm of most concern. The absolute mortality reduction tends to be modest even when screening confers a significant relative benefit, and overdiagnosis is frequently overestimated. Both absolute benefit and overdiagnosis estimates are highly sensitive to the follow-up period which can be limited in screening trials. However, screening trials are generally regarded as providing the most reliable evidence about the likely impact of screening tests. In this presentation I will explain why trials are often inadequate for assessing policy-relevant risks and benefits of cancer screening and will present some statistical and simulation modeling work that we have done in prostate cancer to bridge the gap between the data from trials and the evidence needed for developing population screening policies.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.