Long Term Follow Up

I am working with a protocol that includes survival follow-up conducted via phone calls or review of medical records.  We are struggling with how to approach the survival follow-up reviews which identified the subject to be deceased. At the moment we have:

• SE: SEENDTC for the SFUP element is derived as the date of the last Survival Follow-Up Review Date, regardless of whether the subject was identified as Dead or Alive
  • If we use the date of death instead then there are going to be entries in the SS domain which occur after the last element finished
• SV: All Survival Follow-Up Reviews are included, whether the subject was identified as Dead or Alive
  • If we don’t include reviews identifying the subject as Dead then the corresponding SS entry will  not have a VISIT/VISITNUM
• DM: RFENDTC currently derived as the latest SVENDTC from SV and RFPENDTC currently derived as the last SEENDTC from SE
  • This approach gets errors because RFENDTC/RFPENDTC is after the DTHDTC
  • Contemplating having RFENDTC derived as the last assessment date linked to a clinic visit
• DS: DSSTDTC for final disposition event is currently being linked to the End of Study/Death CRF
  • Therefore we currently have an SE element which goes beyond the last disposition date.
• SS: Includes all reviews, regardless of whether the subject was identified as Dead or Alive

This scenario is very common in oncology studies, so I'm thinking these questions were answered somewhere long long ago, but I can't find anything definitive.