Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Rc/test epic integration #35125

Open
wants to merge 34 commits into
base: master
Choose a base branch
from
Open

Conversation

Robert-Costello
Copy link
Contributor

@Robert-Costello Robert-Costello commented Sep 13, 2024

Product Description

Technical Summary

This is for integrating with Epic as part of SUDCare OCS study.
Details available in this document

Feature Flag

MGH_EPIC_STUDY

Safety Assurance

Safety story

Behind Feature Flag
Accesses specific case types, patient(host) and appointment(extension) , which limits the radius of influence.
Tested locally and on staging.

Automated test coverage

not covered by automated tests

QA Plan

no QA planned

Rollback instructions

  • This PR can be reverted after deploy with no further considerations

Labels & Review

  • Risk label is set correctly
  • The set of people pinged as reviewers is appropriate for the level of risk of the change

@Robert-Costello
Copy link
Contributor Author

More refactoring needs to be done here. This is in an initial test phase using sandbox data from Open Epic.

@Robert-Costello Robert-Costello added the product/feature-flag Change will only affect users who have a specific feature flag enabled label Sep 16, 2024
@Robert-Costello Robert-Costello marked this pull request as ready for review September 16, 2024 17:02
corehq/motech/fhir/tasks.py Outdated Show resolved Hide resolved
corehq/motech/fhir/tasks.py Outdated Show resolved Hide resolved
response = requests.post(url, data=data, headers=headers)
if response.status_code == 200:
return response.json().get('access_token')
elif response.status_code >= 400:
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

what happens for 3xx and 4xx status codes?

if response.status_code == 200:
response_json = response.json()
fhir_id = None
entry = response_json.get('entry')[0]
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It would be useful to have clearer docs available for what the response format is. It's hard to tell if this is correct. It seems like you expect it to be a list with at least 1 item in it but that item might be falsy (empty list, empty dict, null etc).

This code will error if entry is not in the response or if its an empty list.

Copy link
Contributor Author

@Robert-Costello Robert-Costello Sep 17, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

From the open epic docs entry is a required property, but may be an empty list. I added a check here 2409a2e

Name Description Is Optional Is Array
entry (Entry) An array of entries included in this bundle. false true

Name Description Is Optional Is Array
entry (Entry) An array of entries included in this bundle. false true

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this bundle

Bundles can be paginated. You can find code for iterating resources inside paginated bundles here.

corehq/motech/fhir/tasks.py Outdated Show resolved Hide resolved
return utc_iso_format


def convert_utc_timestamp_to_date_and_time(utc_timestamp):
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

similar comment to above re TZ conversion using utilities

corehq/motech/fhir/tasks.py Show resolved Hide resolved
corehq/motech/fhir/tasks.py Outdated Show resolved Hide resolved
corehq/motech/fhir/tasks.py Outdated Show resolved Hide resolved
return date, time


def sync_all_appointments_domain(domain):
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do you know the kinds of scale we expect (# patient / appointment cases). I'm just wondering if this should be done in batches.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I did give some thought to batching. I'm not sure about the number of patients, but I'm estimating that the number of appointments per patient will be relatively low for the timeframe of 12 weeks (still need to add the time constraint to the appointment query). One appointment per week seems like the upper end to me. If you double that it's still only 24 per patient. I'll talk to Chantal and see if she thinks those numbers are low and also get a sense of the number of patients.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

from Chantal:
the max number of patients enrolled in the study will be 55, with rolling enrollment over the next year
~5 patients per month and they are enrolled for 12 weeks so maybe ~15 open cases at a time in a given month

@Robert-Costello
Copy link
Contributor Author

Robert-Costello commented Sep 17, 2024

My reasoning for to this change is that when this case property is updated in CommCare it's formatted with seconds and milliseconds like so: 2016-03-10T17:00:00.000000Z
The value in Epic excludes milliseconds like so: 2016-03-10T17:00:00Z
These equate to the same appointment time, but will never match at the time of syncing appointments and so will always cause a form to be submitted to update that property.
open to other approaches

@@ -563,7 +562,7 @@ def sync_all_appointments_domain(domain):
for appointment in epic_appointment_records:
appointment_resource = appointment.get('resource')
appointment_id = appointment_resource.get('id')
if appointment_id and appointment_id not in appointment_fhir_ids:
if appointment_id and appointment_id not in appointment_map.keys():
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You don't actually need to use .keys() here since dicts support key lookup by default

Comment on lines +407 to +410
if response.status_code == 200:
return response.json()
elif response.status_code >= 400:
response.raise_for_status()
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The response of a POST request can often be a 201 or a 202. I'd return JSON for all success responses. You could also drop the check for error codes, because raise_for_status() won't raise an exception if there was no error.

Suggested change
if response.status_code == 200:
return response.json()
elif response.status_code >= 400:
response.raise_for_status()
if 200 <= response.status_code < 300:
return response.json()
response.raise_for_status()

return token


def request_epic_access_token():
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Handling authentication and authorization inside the corehq.motech.fhir.tasks module feels wrong. I think this either belongs in corehq.motech.auth, or maybe if this code is specific to only one project, it should go in custom.epic?



def generate_epic_jwt():
key = settings.EPIC_PRIVATE_KEY
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Did you consider using ConnectionSettings for storing (and encrypting) API auth secrets?

Comment on lines +450 to +456
base_url = "https://fhir.epic.com/interconnect-fhir-oauth/api/FHIR/R4/Patient"
params = {
'birthdate': birthdate,
'family': family_name,
'given': given_name,
'_format': 'json'
}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This looks super specific, not just to Epic, but to this particular project. Shouldn't this go in custom.PROJECT_NAME?

Comment on lines +458 to +461
headers = {
'authorization': f'Bearer {access_token}',
}
response = requests.get(url, headers=headers)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If you use the requests library directly, you won't get any remote API logging in HQ, and that will make monitoring and debugging a lot more difficult for you. Take a look at how the Requests class gets used. It will make your life much easier later.

You probably also want authentication to be handled for you, instead of handling it every time you make a request. You could subclass AuthManager for Epic, if an existing OAuth2.0 AuthManager subclass doesn't already do what you need.

Comment on lines +464 to +473
entry_list = response_json.get('entry')
if entry_list and len(entry_list) > 0:
entry = entry_list[0]
else:
entry = None
if entry:
resource = entry.get('resource')
if resource:
fhir_id = resource.get('id')
return fhir_id
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Using JSONPath could make this more readable, and the response that you're expecting would be more obvious.

from jsonpath_ng import parse as parse_jsonpath

# ...
    path = parse_jsonpath('entry[0].resource.id')
    matches = path.find(response_json)
    if matches:
        fhir_id = matches[0].value
    return fhir_id

Comment on lines +488 to +489
url = f'{base_url}?{urlencode(params)}'
response = requests.get(url, headers=headers)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
url = f'{base_url}?{urlencode(params)}'
response = requests.get(url, headers=headers)
response = requests.get(base_url, params=params, headers=headers)

... but again, don't use requests directly, use the Requests class to get built-in authentication, and logging that you can audit later.

@@ -2240,6 +2240,13 @@ def _commtrackify(domain_name, toggle_is_enabled):
help_link="https://confluence.dimagi.com/display/GS/FHIR+API+Documentation",
)

MGH_EPIC_STUDY = StaticToggle(
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

(Note to self: Read all the code before you start to comment.)

Instead of using a feature flag, I'd advocate moving this code into custom.mgh_epic, and enable the custom module in settings.py for the project spaces you want.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for taking a look, Norman. This has been in a holding pattern for a while, but I'm coming back to it to do some more work now. I can move this to a custom module.

epic_appointments_to_add = []
epic_appointments_to_update = []
# Get all appointments for patient from epic
epic_appointment_records = get_epic_appointments_for_patient(patient_fhir_id, access_token)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Did you look at using FHIRImportConfig (documentation)?

Would it be a lot of effort to extend it to allow syncing appointments for this project?

  • We might need to make some changes to allow for extension cases instead of child cases.
  • We would probably need to be smarter regarding how we set case ownership.
  • We would either need all cases to store their corresponding FHIR ID in their external_id case property, or use case search to fetch them.
  • We probably need to add a (custom?) serializer to handle timestamps for this project.

... and maybe it's a lot more complicated than that.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
product/feature-flag Change will only affect users who have a specific feature flag enabled
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants