Death

The Death domain contains the clinical event for how and when a Person dies. A person must have exactly one record if the source system contains evidence about the Death. Existence in this table, even with no additional information, marks a patient as dead.

Field	Required	Type	FK Table	Description
person_id	Yes	bigint	PERSON	A foreign key identifier to the deceased person. The demographic details of that person are stored in the person table.
death_date	No	date		The UTC date of death (e.g. possibly from the NHS Spine).
death_datetime	No	datetime		The UTC date and time the person was deceased or was marked as discharged deceased.
death_type_concept_id	No	bigint	CONCEPT	Not used. Could be used to record how the EHR learnt about the death (e.g. NHS spine, died in hospital, etc)
cause_concept_id	No	bigint	CONCEPT	Cause of death.
cause_source_value	No	varchar(50)		Cause of death as stored in the source EHR.
cause_source_concept_id	No	bigint	CONCEPT	Cause of death concept stored in the source EHR (if present).
last_updated_datetime	Yes	datetime		Initially this is the time the row was written to OMOP. If the row is subsequently updated, it then becomes the most recent update time.
deleted_datetime	No	datetime		NULL initially. Set to the time when the row was marked for deletion. A value in this column sent to the central data source will result in this row being deleted (by person_id ) from the combined dataset. Deletions only need to be sent once.

Conventions

• Living patients should not contain any information in the DEATH table. Dead patients must have a record in the DEATH table.
• For any given patient, as many of death_date and death_datetime as are known should be submitted. Where the time is not known, death_datetime should not be provided. If the date of death is not known at all, both columns can be NULL (i.e. we know they are dead, but we don’t know when they died).