Hi! I am doing a survival analysis on retrospective registry data (cancer diagnosed 1995-2018).

Last follow-up is 31.12.2018, so obviously patients will have different follow-up depending on the date of diagnosis.

Can anyone explain the difference between doing survival analysis with K-M method OR actuarial method? Is one or the other more reliable when doing survival analysis on registry data. I understand that actuaril method uses fixed time intervals and calculates relative survival, and that eg SEER uses that merhod (Ederer 2 for relative survival analysis). Is there any downfall of using K-M method over actuarial?

Is relsurv package in R ”best” fitted for actuarial survival analysis?

Thx in advance