Authors’ Reply: Interpreting epidemiologic studies of colorectal cancer prevention

Excerpt

We would like to thank Brenner and Hoffmeister for their comments on our commentary. We agree that including potential prevalent Colorectal cancer (CRC) cases in clinical trials of endoscopic screening may dilute the effect estimate of polypectomy, which is the predominant mechanism of action for the preventive effect of endoscopy against the incidence of Colorectal cancer. However, as stated in our commentary,(1) there are both conceptual and methodological reasons why there is no simple fix for the prevalent case issue. Conceptually, prevalent cases are challenging to define because [1] participants may be at different stage of the natural history of CRC and [2] while present in both the intervention and control groups prevalent cases cannot be identified from the control group who did not receive an endoscopy. Methodologically, there is no established method to effectively and rigorously identify and quantify prevalent cases. The method that Brenner et al. proposed (2) is rather crude and based on several uncertain assumptions. (1) therefore, we believe that, instead of conducting crude analysis that can lead to more problems than fixes, effort can be better spent on fundamental research to address the conceptual and methodological issues related to prevalent CRC cases that occur in both clinical trials and observational studies of screening. …