2a) Not clear if/how provenance is captured, represented and persisted (over time) in this specification. Knowing who did what when were and why is essential. Capture points include: i) point of origination (at the source, for newly captured content), ii) point of update (for newly updated content whilst preserving previous content and its provenance).

Who = subject of care/information (typically the patient)
Who = participant in action taken, including role (e.g., performer, assistant, observer) and credentials (e.g., MD, RN, PharmD, therapist, MA...)
Who = author of information captured or updated
Who = organization
What = action taken
When = date/time of action taken
When = date/time of information capture or update
Where = physical location of action taken
Where = physical location of information capture or update
Where = network address and/or device ID where information captured or updated
Why = rationale, purpose of action taken

2b) Provenance elements, if not captured at the point of origination or point of update, are often forever lost beyond that moment.

2c) Granularity of provenance (e.g., binding authorship to content) may be at the dataset or element level, as appropriate.

2d) Provenance elements are intrinsic to what the source system or device already “knows” at the point of origination or update and thus should not increase burden by requiring extra input on the part of the entering author (clinician or other end user).

(Comment 2 - imported by: Robert Dieterle)