Two randomized screening trials with prostate cancer mortality, two interim results: a consideration of usual-care window in screening trials.

Currently, random screening trials with prostate cancer mortality focus on testing the hypothesis that screening reduces the mortality. When the trials are carried out in places where diagnostic practices differ, this hypothesis test can be misleading. The divergence of the interim results from two large random screening trials for prostate cancer mortality, one in Europe and the other in the United States, confuses practitioners and raises questions about the trials' credibility. I proposed a theoretical framework for random screening trials and concluded that the application of the results from an individual trial should be limited to the regions where similar diagnostic practices are observed. I also suggested that a synthesis of individual trials could lead to the largest usual-care window with which screening has the designed effect.