If a Element can be null, seems reasonable to allow that to a Cluster for the exact same reasons (e.g. there could be reasons to not being able to get a series of measurements and then an empty Cluster with a reason would be created)
We can also discuss if it should go to ITEM or ITEM_STRUCTURE (or leave it until we redo the ITEM_STRUCTURE classes)
I remember hearing that CIMI has also gone this way, but probably Thomas knows better.
Hi Diego, the assertion about the ability to identify the person event if each element is empty assumes a lot, like: needs access to the demographic database, knowledge about how it is implemented / structured, a little knowledge about openEHR RM maybe, etc. I still don't think that's a good reason to add null_flavour to CLUSTER (or ITEM), but if others see value on this, is ok with me.
Reading all the comments above I understand that:
1.) we will not do this
2.) related things are done (or in progress) in Release 1.1.0
I will add here also Release 1.1.0, and next SEC will do further handling of this issue.
I don’t exactly remember which was the original discussion prior to raising this issue, so I’m fine with 1)
I agree with (1) - ‘non-persuasive’ as our friends say
I’d agree that the use case for this doesn’t seem to be sufficient considering the increased complexity and likely confusion.