I realize that a healthy person could die due to a surprise heart attack in which there is no record in the past.
Should this death condition be also recorded in the condition_occurrence table? thanks
Hi @lychenus13, according to the documentation describing the structure of Clinical Data Tables in the OMOP CDM, the clinical event for how and when a person dies has to be stored in the Death table. Please see a fragment of it below:
Yes, it should be in the death table, but should it also be in the condition occurrence table?
I have not studied how most OHDSI packages are written in terms of code and protocol, but would it be possible that a study wish to include the event at death or just before death as prevalence but forgets to read the death table? Then the result would be diluted.
Hello @lychenus13.
Storing the cause of death in the CONDITION_OCCURENCE table is good practice. One of the reasons is not taking into account the DEATH table in the process of statistical analysis, as you already indicated. On the other hand, information about death may be presented various in different data sources and it maybe tricky to pick a correct place of store.
Possibly related: Please see the following thread for additional discussion
THEMIS Conventions for Death Table to Allow Multiple Records - Version 5.4 - CDM Builders - OHDSI Forumsemphasized text