Query deleted resources for p2p sync

[ekigamba]

Is your feature request related to a problem? Please describe.

Describe the solution you'd like
There are several proposed solutions:

1. Default to soft-deletes and have an API for hard deletes
2. Soft-deletes only. Never purge data unless the database is deleted
3. Keep a separate table for tracking deleted resources on the implementing application (FHIR Core/P2P)
4. We can partially open up the local changes API
5. We can have a log table on FHIR Engine to keep track of all resource changes

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context
Add any other context or screenshots about the feature request here.

Would you like to work on the issue?
Please state if this issue should be assigned to you or who you think could help to solve this issue.

[ekigamba]

Solution 1

• Search and other APIs need to default to undeleted records
• We need to create APIs to provide access to soft-delete records
• How does the developer/application know when to hard-delete/purge the records

Solution 2

• Generally, we don't have a lot of cases where data will require to be deleted. However, if data is erroneously created many times, the data can become large or an issue.
• We don't want private and misplaced medical data on a device

Solution 3

Solution 4

• The local changes table will become too big if the device never syncs to the server
• The delete-type local changes will be deleted if the device syncs to the server and cannot be shared back to the other devices through p2p

Solution 5 (Log table on FHIR Engine tracking resource changes)

• When do we delete these logs? Does the device share these records with multiple p2p devices

Based on the discussion

• Have a quick call in the next week
• List out all the use-cases so that we can come with a robust solution
• Update notes and p2p-based solutions in this doc

Use cases

1. Erroneously created tasks and care plans that need to be hard deleted
2. Erroneously created data needs to be purged
3. Devices that never connect to the server after initial login. These devices uses p2p sync to share data to the supervisors device which then connects to the internet
4. Devices that are offline at times and use p2p sync but are connected to the internet once in a while

[ekigamba]

Feb 24, 2022

We have two options for p2p currently:

1. Mark resources as soft-deleted and provide a flag in the Search API that provides either soft-deleted records
2. Provide an API that returns a list of Resource IDs for deleted resources filtered by date

Generally, we agreed that:

• that we need a purge API
• Solution 3 would need each implementer to separately develop this

Other notes

• We can use the Patient.active to indicate deleted resources. This solves a lot of the issues with syncing deleted resources. The transient resources/entities have a field like active. Other resources like libraries do not need to have this field
• Does the server provide a list of deleted resources using the normal search API? How do apps get updates on deleted resources