Add ELEMENT.null_reason of type DV_TEXT.
(Somehow the commit is not linked from here, but the result can be seen at https://specifications.openehr.org/releases/RM/latest/data_structures.html#_element_class )
I think the null_reason would benefit from a bit more of a description along the lines of what Silje and Ian have stated in the comments in June/July 2016 at . I think this would be useful to have to understand why both null_flavour and null_reason are useful when reading the spec.
Regarding null_flavour, the examples in the description (Description: Flavour of null value, e.g. indeterminate, not asked etc.) are different to the ones in the terminology or am I reading that wrong?
Also, for readability, I would consider changing the order of the 3 attributes to “value”, “null_flavour”, “null_reason”.
I agree with Sebastian - I think only <concept id="272" rubric="masked"/> is missing from the docs and change the order as suggested.
Based on various comments on
Optional specific reason for null value; if set, null_flavour must be set. Null flavours only apply to a small sample of clinical data, but it's not uncommon to have to specify exactly why information is missing, particularly in reporting contexts. It may be further constrained to an internal or external codelist in particular archetypes or templates.
I've now made adjustments to the text as suggested by Ian.