FI-590: Add input for Device type

[Jammjammjamm]

This branch adds an input field for Device type. This is needed because the US Core Implantable Device profile only applies to implantable devices, but without specifying a Device type, there is no way for Inferno to differentiate between implantable and non-implantable devices.

Submitter:

• This pull request describes why these changes were made
• Internal ticket for this PR: https://oncprojectracking.healthit.gov/support/browse/FI-590
• Internal ticket links to this PR
• Internal ticket is properly labeled (Community/Program)
• Internal ticket has a justification for its Community/Program label
• Code diff has been reviewed for extraneous/missing code
• Tests are included and test edge cases
• Tests/code quality metrics have been run locally and pass

Reviewer 1:

Name:

• Code is maintainable and reusable, reuses existing code and infrastructure
  where appropriate, and accomplishes the task's purpose
• The tests appropriately test the new code, including edge cases
• You have tried to break the code

Reviewer 2:

Name:

• Code is maintainable and reusable, reuses existing code and infrastructure
  where appropriate, and accomplishes the task's purpose
• The tests appropriately test the new code, including edge cases
• You have tried to break the code

[arscan]

To keep the UI simpler, can we just ask for the code and not bother with the system? We are really just looking for a way to filter down the results to known implantable device types, and it seems unlikely that they would use the same code repeated in two separate codesystems in a small handful of patient records to make this technically needed.

And if that sounds reasonable, I think it would be worthwhile to accept a comma separated list of codes to filter by instead of a single code, so you could paste in a long list of allowable device types and just validate any that are in that set.

[arscan]

If you run it without the device code filter, then rerun it with a bad filter (so none should come back), it still seems to know about the device it saw previously. Probably because we save references in the db and don't clear them? I don't know exactly how we want to handle this... might be a real pain.

[Jammjammjamm]

Making the Device sequence destroy all of the Device references when it starts wouldn't be difficult, just another special case to add.

[arscan]

I think a special case to clear device references is fine. Just don't do it on all the us core profile tests (unless we really analyze the impact globally).

[Jammjammjamm]

@arscan The issue you raised should be addressed now.

[arscan]

looks great, thanks!