Commentary: estimating cancer survival--which is the right approach?