GDPR - quick question on the new European regulation

[Macawac1]

Hi again,

This is not related to my earlier post which does not identify individuals, but instead is a general question.



If an individual can insist on complete deletion of all records held by an organisation regarding their personal details - how would this be managed in database design that still needs historic transnational records?

Just an ad hoc thought!

I am sure it has been discussed many times since the regulation's conception but just "thinking aloud" here without really doing any further looking into it.

Regards,

Karen.

[davegri]

Well, if all identifying information is to be deleted (replaced) you would need a means to replace the name, address, policy numbers, medical record number, etc with a generic set of fields.
If the records need be maintained and reported individually, you could replace the policy number (for example) with a new number that increases by one for each non-disclosure policy change.
The new number would have to be unique and occur outside of any possible real policy number series.
Another possibility is to assign a special policy number that identifies the record as GDPR.
There are limitless ways to change the identifying data so that the data is still useful but not pointing to an individual.

[NTC]

there would no choice but to replace/overwrite the data with pseudo info ... that way one could still do counts and have historical accuracy ....

typically the person's ID sits in a table - - and it links to other tables with a integer key - - not actually any of their personal info; so the overwrite of pseudo info maintains the record.

clearly the database designer must know the app - and such things as sorts or counts by last name versus postal code would/could be affected depending on how things are implemented.