Keywords: G-methods; bias due to coarsening time interval; colorectal cancer; screening effectiveness.