Letters Section Editor: Robert M. Golub, MD, Senior Editor.
In Reply: I concur with Dr Singh and colleagues that cancer rates among those selected by a single negative colonoscopy result should not converge completely to those in the general population. However, they should come closer together as time passes, due to the diminishing selection effect as new polyps are developed in the subset of the those with negative screen results still at risk. For a simple numericalexample, suppose 90% of those with a negative screen result are polyp-free and 10% have missed polyps. Then the immediate effect of the selection is to reduce the rate to less than 10% of that in the general population. As time passes, however, more and more (though certainly a lower percentage than in the general population) of the 90% who are polyp free will develop polyps and subsequent cancer, reducing the difference between the average incidence and the incidence in those with negative screen results. However, this effect is not evident in their data. The difference actually grows over time. This negates the wholly expected finding of a large and sustained difference between those with negative screen results and the general population.
Church TR. Colonoscopic Screening for Colorectal Cancer—Reply. JAMA. 2006;296(20):2438-2439. doi:10.1001/jama.296.20.2438-b