Comments on the Use of Time-Specific and Cohort Life Tables