Alternatively, we can incorporate the excess cohort-mortality into the clinic-based population to create a new (simulated) analysis dataset for which the assumption of comprehensive mortality information is appropriate. No statistical adjustment for late-entry bias is therefore necessary. It is assumed that JCS mortality will not exceed clinic-based mortality beyond age 25 (the age of the oldest cohort members in 1999).