Abstract
Despite the promise of the PSA test for early detection of prostate cancer, it is now clear that PSA screening can be costly, generating harms alongside benefits. Overdiagnosis, or the detection of cancers that would never have been detected clinically, is perhaps the harm of greatest concern in PSA screening. However, different studies provide widely varying assessments of the magnitude of the overdiagnosis problem. A report from the European prostate screening trial estimated that 48 men would have to be overdiagnosed to prevent one prostate cancer death, and a report based on US incidence trends estimated that more than one million men had been overdiagnosed and overtreated by 2005. These statistics are alarming but dramatically overstate the problem. In this presentation I will demonstrate how different definitions, methods and even populations can produce very different, and sometimes inaccurate, assessments of the frequency of overdiagnosis associated with PSA screening. I will present results from two studies of the US population that showed that about one in four men over 50 detected by PSA screening prior to 2000 was overdiagnosed. I will also present personalized estimates of the likelihood of overdiagnosis that vary from from 3 to 80% depending on age, grade and PSA. These estimates can be used in designing population screening programs that reduce costs and harms while preserving benefit.
Citation Format: Ruth Etzioni. Estimating how many prostate cancers are overdiagnosed: Overcoming challenges and avoiding mistakes. [abstract]. In: Proceedings of the Eleventh Annual AACR International Conference on Frontiers in Cancer Prevention Research; 2012 Oct 16-19; Anaheim, CA. Philadelphia (PA): AACR; Cancer Prev Res 2012;5(11 Suppl):Abstract nr FO01-01.
RSS Feeds