In Reply Our study compared the end points of cancer-specific mortality and late-stage cancer in randomized clinical trials of cancer screening. Drs Kim and Gibbs write that the relationship between these could be affected by changes in cancer staging and treatment. We agree that this is true not only for our comparison of end points, but for cancer screening in general. For example, a screening test that is effective at reducing cancer mortality could become ineffective if new treatments decrease differences in survival between stages. For the cancer types in our study, Surveillance, Epidemiology, and End Results data generally show survival improvements over time across all stages.
