The vital status is whether a person is alive or deceased. I think the field is named status in our EHR. And it is always populated with either Alive or Deceased. No null values
We also retain records of multiple sources of death (e.g. EHR, SSA, data aggregator), and apply logic to compute the best known death from various sources and have that be the single data of death in OMOP. It isn’t always the earliest date recorded, as there are different levels of trust across the various sources.
I want to caution about the idea of using the below approach:
Within our EHR, if a person becomes an organ donor or has an autopsy, those clinical events are attached the the deceased person’s record - so can occur well after the date of death. One of the DQD logic options is to only flag death date as suspicious if there are data attached to a person more than 60 days after recorded date of death.
Please see the following thread for additional discussion: