Post-Processing Your Data

SMART Fetch strives to make as complete an export as possible for archival and research purposes.

There are situations where not all the data you might be clinically interested in gets exported, and SMART Fetch will try to get it for you after the fact.

SMART Fetch calls this kind of post-processing “hydration” and happens automatically for each export.

Below is a fuller explanation of each hydration task.

Inlining Clinical Notes

Both DiagnosticReport and DocumentReference resources refer to external clinical notes. The resources themselves are usually just a bit of metadata about the note and an external pointer.

That’s not very useful for a lot of clinical use cases, since the note is the most valuable bit. And since a note’s URL may change (or you may lose access to the EHR), it’s nice to have the note downloaded and ready to go.

So SMART Fetch will download all the clinical notes it finds (that are HTML or plain text) and stuff them back into your NDJSON as inline notes (using the data attachment field).

This way they can be processed with other FHIR tools that handle attachments.

Downloading Missing Observations

Vendors have a habit of treating certain Observations specially and not including them in bulk exports or searches.

Specifically, they may leave out Observations pointed to by DiagnosticReport.result or Observation.hasMember.

But since both kinds of Observations are clinically useful, SMART Fetch will download them all for you, if they weren’t already included in the export.

Downloading Useful-But-Not-Patient-Linked Resources

Resources like Medications, Locations, Organizations, Practitioner, and PractitionerRole might be useful for clinical analysis, but are not associated with patients directly. Thus, they aren’t bulk-exportable.

But since they are useful, SMART Fetch will download them for you when it encounters a reference in the exported data.