On 2016 Nov 12, Daniel Corcos commented:

It is surprising that the prevalent screen of the control group detects 14% more cancers than that of the screened group. Also, the incidence of cancer at the second screen is much too high as compared to cancer incidence in the control group. Both results strongly suggest that cancers ascribed to the second round of screening were actually detected at the prevalent screen of the screened group. After correction, it appears that all the differential increase between screened and control groups occurs between the second and third round of screening. Therefore we are not dealing with cancers that spontaneously regress, but with cancers that are induced by x-rays.

On 2016 Nov 12, Daniel Corcos commented:

It is surprising that the prevalent screen of the control group detects 14% more cancers than that of the screened group. Also, the incidence of cancer at the second screen is much too high as compared to cancer incidence in the control group. Both results strongly suggest that cancers ascribed to the second round of screening were actually detected at the prevalent screen of the screened group. After correction, it appears that all the differential increase between screened and control groups occurs between the second and third round of screening. Therefore we are not dealing with cancers that spontaneously regress, but with cancers that are induced by x-rays.