Please fix logical IDs

[reeager]

Please fix logical ids to continue to be unique but must remove the following:

• spaces
• tildes
• be less than 64 character

This is to ensure the IDs are meeting R4.

[alan-francis]

We are creating the logical Ids for the following resources in the pattern described below:

1. Bundle- Custom uuid--
2. Patient - -
3. Consent - consentFor-
4. Observation - observationResponseQuestion--md5()-

We will be generating the above ids without spaces, tildesand perionds. But the bundleId& observationIds tend to be more than 64 characters.

@reeager @raphaelvp @ratheesh-kr @shah

[reeager]

@reeager periods can be used- update made.

[reeager]

For Consent: mrn, assigning auth, Organization (custodian).

[ratheesh-kr]

The following are the generic id-primitive and dateTime-primitive rules that FHIR follows.

id-primitive

• The id primitive in FHIR must adhere to specific constraints to ensure uniqueness and consistency:
  • Type: String
  • Length: 1 to 64 characters
  • Allowed Characters:
    • Lowercase alphabetic characters (a-z)
    • Numeric characters (0-9)
    • Hyphen (-)
    • Dot (.)
    • Underscore (_)
  • Regular Expression Pattern: ^[A-Za-z0-9\-\.]{1,64}$

dateTime-primitive

• The dateTime primitive in FHIR follows a pattern derived from ISO 8601. It can represent:
  • Full Date and Time: YYYY-MM-DDThh:mm:ss[Z|(+|-)hh:mm]
  • Date Only: YYYY-MM-DD
  • Year and Month: YYYY-MM
  • Year Only: YYYY
• Components:
  • YYYY: Four-digit year
  • MM: Two-digit month (01-12)
  • DD: Two-digit day of the month (01-31)
  • T: Literal character 'T' used as a separator between the date and time parts
  • hh: Two-digit hour (00-23)
  • mm: Two-digit minute (00-59)
  • ss: Two-digit second (00-59)
  • Z: Indicates UTC (Coordinated Universal Time) if present
  • (+|-)hh: Time zone offset from UTC

@reeager @shah @raphaelvp

[alan-francis]

@reeager ,

• we understand from the above document that we need to hash all the created logical Ids with the sha256 method.
• Also we need to use the ~ as the concatenation string for each ID.
• Can we have a more clarity on where to find the value for visit_number to be used for creating the IDs in Encounter and Observation?
• Also the only Date-time field that we currently have is the RECORDED_TIME inside the screening CSV. So should we use that as the period:start/message_date ?
• If so, as there is no Rejection for an incorrect format of the RECORDED_TIME we may not have control over the format which will be then used in the ID generation process.

@shah @raphaelvp @ratheesh-kr

[reeager]

@alan-francis

• Tildes are not an allowed character for IDs.
• I am not sure we are going to get visit_number, this might have to be derived.
• Yes to your question about recorded time. The team should also start formatting recorded time correctly. We should discuss this with Ready.

[raphaelvp]

@reeager,

Please provide clarifications for the following:

• For visit_number, if it needs to derived, specify the logic and source data to derive from.
• There are specific format requirements in the Logical ID design document from the Data Lake team. If the RECORDED_TIME is incorrectly formatted, it will fail for other fields also.

[reeager]

• We can use encounter # for visit_number.
• Let's talk to Ready about recorded_time.

[alan-francis]

@reeager : So we can use the ENCOUNTER_ID from the SCREENING csv for visit_number. Also as this ENCOUNTER_ID is not a mandatory one and we are generating this if not available from FACILITY_ID and PAT_MRN_ID, we can do the same for visit_number as well.

@shah @raphaelvp @ratheesh-kr