Impact of the Management and Proportion of Lost to Follow-Up Cases on Cancer Survival Estimates for Small Population-Based Cancer Registries

Figure 1

Schematic representation of the steps of the simulation process regarding simulating loss to follow-up. Time until death for four hypothetical cancer patients in this study; horizontal lines represent time from diagnosis with dots representing the date of death; the labels projected on top of the lines convey information regarding time under observation or to event in a 5-year survival analysis depending on the exact simulation step. (1) Complete cases, no simulation applied; (2) censured case scenario: randomly patients B and C were censored to be lost (simulating date of death was not reported by vital statistics, and therefore, patient erroneously considered alive at analysis point); time of patient B changed from 2.5 to 5.0 years, time of patient C unchanged by simulation as patient really did not decease; (3) excluded cases scenario: patients B and C in previous step changed to be lost to follow-up were excluded from analysis.