Which one? What’s the use case that forces us to record it that way?
Seriously, @Karthik? 
The entire OMOP CDM model relies on date relationships. The source data often don’t give us reliable cross-references. We don’t even know if procedures, drugs and devices belong to a visit explicitly in many cases. And now you want to have one special situation just for death, and only for one Visit (Inpatient)? How do you intend to make people familiar with that kind of a unique exception?
Look: I know where this is coming from. Folks in the EPIC EHR have a “discharge_disposition” field and want that field back with an entry “expired”, because they are used to it. That does not count as a use case. The clause “where death_date>=visit_start_date and death_date<=visit_end_date” is not too much to ask.
The Place of Service notion was retired a while ago. It is a crude mix of different things, and outside the USA with its quasi-standardization through the CMS it makes no sense. We now have Visit concept. They define the setting healthcare is delivered, who comes to whom in which environment, what is the level of service. Death is not a healthcare setting. It’s the opposite.