CancerRelatedMedication STU2 Updates

[dmendelowitz]

Summary

CancerRelatedMedicationExtractor has been converted to CancerRelatedMedicationAdministrationExtractor and a new CancerRelatedMedicationRequestExtractor has been added.

Profiles:
https://build.fhir.org/ig/HL7/fhir-mCODE-ig/branches/master/StructureDefinition-mcode-cancer-related-medication-request.html
http://hl7.org/fhir/us/mcode/2021May/StructureDefinition-mcode-cancer-related-medication-administration.html

New behavior

CancerRelatedMedicationAdministrationExtractor behaves the same as the old CancerRelatedMedicationExtractor. It's pretty much just a renaming. CancerRelatedMedicationRequestExtractor is a new extractor based on the profile from STU2 linked above.

Code changes

• All files associated with the previous CancerRelatedMedicationExtractor (the extractor, template and tests) have been renamed to CancerRelatedMedicationAdministration to distinguish them from CancerRelatedMedicationRequest
• New extractor, template and tests have been added for CancerRelatedMedicationRequest profile
• MCODEClient.js, index.js and ResourceGenerator.js updated to support the new and renamed extractors
• Updated sample client data to have new cancer-related-medication-request-information.csv and renamed cancer-related-medication-administration-information.csv

Testing guidance

• Ensure that in converting CancerRelatedMedicationStatement to CancerRelatedMedicationAdministration, I didn't miss any other changes made in STU2
• Check that my implementation of CancerRelatedMedicationRequest properly implements the profile (if I am missing any fields or implementing any fields that aren't necessary) and check that the new CSV format makes sense
• Ensure that the client runs and tests pass
• Here is an example config for the relevant extractors:

"extractors": [
    {
      "label": "patient",
      "type": "CSVPatientExtractor",
      "constructorArgs": {
        "filePath": "./test/sample-client-data/patient-information.csv"
      }
    },
    {
      "label": "cancerRelatedMedicationAdministration",
      "type": "CSVCancerRelatedMedicationAdministrationExtractor",
      "constructorArgs": {
        "filePath": "./test/sample-client-data/cancer-related-medication-administration-information.csv"
      }
    },
    {
      "label": "cancerRelatedMedicationRequest",
      "type": "CSVCancerRelatedMedicationRequestExtractor",
      "constructorArgs": {
        "filePath": "./test/sample-client-data/cancer-related-medication-request-information.csv"
      }
    }
  ]

[jafeltra]

This looks really great! All the changes in the code and the new fields in the extractor look correct to me! I ran the FHIR validator against the output and found a few things we might want to update in our sample data, so I left a few comments related to that, along with just a couple quick general comments.

JA1 - The src/index.js file should be updated similarly to the src/extractors/index.js file to include the new extractors.

JA2 - Can you update the csv.config.example.json file to use the two new extractors instead of the one old one?

JA3 - It looks like the status values for MedicationAdministration are different than they were for MedicationStatement. The options it can be are in this valueset. Do you think you could update the values we have in the sample-client-data CSV just so they reflect the expected values?

JA4 - Similar to above, the values for the reasonCode in the MedicationAdministration don't seem to match up with codes in the current Cancer Disorder Value Set. Could you update the values in our sample data fields to match codes in that value set?

JA5 - Pretty much the same as JA4, but the reasonCode for the MedicationRequest in the sample CSV also doesn't match a code in the current value set. Can you update that as well?

[Dtphelan1]

This looks great Dylan! In addition to the comments Julia made:
DP1 - In the same way that example data was added to the test/sample-client-data, we should update similar files in the docs folder. Specifically, let's get rid of the docs/cancer-related-medication.csv and add docs.cancer-related-medication-request.csv+ docs.cancer-related-medication-administration.csv documentation files.
DP2 - Similar updates, but in the CSV Templates file.
DP3 - Minor note: cancerRelatedMedicationAdministrationExtractor.test.js and cancerRelatedMedicationRequestExtractor.test.js use variables called csvCancerRelatedMedicationExtractor, which might make life easier from a shorthand perspective writing-wise but might be nice to replace with the xAdmin or xRequest equivalent.

[Dtphelan1]

DP4 - For some of the common template snippets – medicationTemplate, treatmentReasonTemplate and subjectTemplate come to mind – could we put those in separate Snippet files, then import them into the two main templates we're defining? Willing to push this into a separate task if it's sufficiently hairy/out of scope.

[jafeltra]

This looks good! All the JAs are resolved! The one last thing that I noticed is that I don't think you need to add the ~$CSV_Templates.xlsx file. I think the regular file is sufficient.

[jafeltra]

Looks great!

[Dtphelan1]

This looks good to me Dylan – addresses all the comments I made eariler. Great work Dylan!