Return to SDWG page.
Return to C-CDA: Enhancing Implementation (ONC Grant Project) page.
January 12-13, 2017
Trinity University
San Antonio, Texas
Attendees
- Joginder Madra (Facilitator)
- Jean Duteau (Facilitator)
- Dave Hamill (HL7)
- Dan Chaput, Annette DeVito Dabbs and Nagesh Bashyam aka Dragon(ONC)
- George Cole, Jon Anderson and Jennifer Grater (Allscripts)
- Brett Marquard (River Rock/IPO)
- Calvin Beebe (Mayo)
- Ed Donaldson (Greenway Health)
- Ben Flessner (Epic)
- Raychelle Fernandez and Ozlem Kurt (Dynamic Health IT)
- Lisa Nelson (Life over Time Solutions)
- Russ Leftwich and Alok Saldanha (Intersystems)
- Linda Michaelsen (Optum)
- Jiong Mao (NexGen)
- Joe Lamy (Aegis)
- Gay Dolin (IMO)
Agenda
Thursday, January 12, 2017
- 0900-0930 Introductions/Housekeeping
- 0930-1200 Scenario 1 ("Everything and the Kitchen Sink")
- 1200-1300 Lunch
- 1300-1400 Discussion
- 1400-1430 Ask the ONC
- 1430-1700 Scenario 2 (Care Plan)
Friday, January 13, 2017
- 0900-1100 Scenario 3
- 1100-1200 VSAC presentation (Rob McClure)
- 1200-1300 Lunch
- 1300-1330 Ask the ONC
- 1330-1600 Discussion
- 1600-1630 Wrap Up
Proposed Discussion Items
- Complex medication SIG
- Empty unit values in CCD or values with spaces (e.g. mg/dl creatinine vs. mg/dl_creat)
- Handling signatures in CCD
- Chart corrections (Ben)
- Goals and status of goals (Alok)
- Header level ID root (Linda)
- Text reference linking (Lisa)
- Combination codes/code translation (Ben)
- Date usage in conjunction with Problem, Medication, Allergy, etc.
- Care Plan certification test data from 170.315_b9_CP_Amb and _Inp
- Displaying a DI report as part of 170.315(e)(1)
- Partial refills for Medication - e.g. 90 capsules, but insurance only covers 30 days supply...refills = 1.5)
- Second observation to indicate number of items - e.g. this section contains 10 lab results
- Identify a medication as an inpatient or outpatient medication
- Sexual Identification and Gender Identity / Intersex birth certificate - http://www.cnn.com/2016/12/30/health/intersex-birth-certificate/index.html
Discussion
| Item
|
Notes
|
| Introductions and Housekeeping
|
- Dave Hamill welcomed the group. Brief introduction of facilitators and participants.
|
| Scenario 1 - CCD
|
Key items in this scenario:
- Care Team representation
- Section dates
- Section authors
Discussion:
- Are there any differences between inpatient and outpatient CCD for this scenario? No...should be treated the same.
- Item for Follow Up - How should authors of an imported document be represented if one creates a subsequent document that includes the imported information?
- Should not use informant as informant has a specific meaning - i.e. someone who was called to provide information.
- Item for Follow Up - How does on know when information on a problem list, medication list etc. was reconciled?
- Item for Follow Up - Causal observations - e.g. brother died of colon cancer in 2009
- Item for Follow Up - Amount of information presented in the narrative vs. amount of information presented in the entry - i.e. is it proper to have more information presented in the narrative than information presented in discrete data....or vice versa?
Scorecard questions:
- Is the scorecard validating units (e.g. ucum) correctly? It appears that the scorecard does not support ucum parsing and is looking at specific value set concepts.
- ID validation (for uniqueness) appears to be looking at the ID element instead of ID attributes. If an author is noted in multiple spots within the CDA, this error will be triggered. Participants should probably be excluded from this check.
- Effective time rules not appear to be applying correctly (e.g. effective time for all activities should be within the lifespan of the patient)
- These issues have been raised with Dragon and he will work with the C-CDA R2.1 scorecard team to address these.
|
| Scenario 2 - Care Plan
|
Key items in this scenario:
- Concerns, goals, and interventions
- Linkages and references between concerns, goals and interventions
Discussion:
- Open issue with scorecard where Care Plan documents are not being scored correctly
- Health concerns may be coded, but may need to be represented in narrative form as there are likely to be terminology system gaps
|
| Scenario 3 - CCD
|
Key items in this scenario:
- Documentation of concerns
- Communication of concepts in Patient History
- Documentation of future meds (with escalating dosage)
Discussion:
- Guidance around documentation of goals and health concerns is really geared towards the original document creator. Consuming systems should not be reorganizing information provided by a provider.
- How to deal with complex SIG?
- Current systems are not likely to process substance administrations with an organizer properly - suggest this approach not be used.
- It would be possible to use a single substance administration with multiple substance administrations under it. This would be the safest approach for backwards compatibility.
- Given that consuming systems are not likely trying to parse the SIG, use of a free-text SIG is also appropriate.
- Suggestion for score card - extra credit should be given to vendors who use the free-text SIG template
- What makes something a health concern?
- Something is a health concern if a member of the clinical team feels something is a health issue that should be tracked
|
| Discussion
|
- Chart corrections - How does one communicate something was sent in error? Outside of treating a new document as a complete replacement of a prior document, how should this be addressed?
- Immediate way to address this is to "pick up the phone"
- Item for Follow Up - Use case needs to be drafted in advance of further work.
- How to we differentiate between patient goals and provider goals?
- Patient goals have a patient author, provider goals have a provider author, and goals that have both a patient and provider author are shared/negotiated goals
- Can we apply meaning to the document ID root so we can determine who created a document?
- It is permissible to recognize your own document root, but there may not be a way to guarantee meaning for ID roots from documents created by other vendors. Use of authoring device may be better.
- How to we differentiate between patient goals and provider goals?
- Patient goals have a patient author, provider goals have a provider author, and goals that have both a patient and provider author are shared/negotiated goals
- Can we apply meaning to the document ID root so we can determine who created a document?
- It is permissible to recognize your own document root, but there may not be a way to guarantee meaning for ID roots from documents created by other vendors. Use of authoring device may be better.
- Handling signatures in CCD.
- Signature text has been added to authenticator and legal authenticator as an extension. This was work that came out of the attachments WG via CDP1. While this can be technically be done, workflow considerations still need to be worked out - e.g. who is the legal authenticator, what are they attesting to, how does this attestation persist, etc. We need to keep in mind that only a person may be a legal authenticator.
- Empty unit values in CCD or values with spaces (e.g. mg/dl creatinine vs. mg/dl_creat)
- Spaces are not allowed by either coded simple or UCUM. Proper UCUM representation i s mg{creat}/dL.
- CDA discrete entry concepts - see File:Discrete entry to narrative references.pptx
- text/reference - points to the entire human readable content for an act
- originalText/reference - always in reference to some (code, codeSystem) as a reference to the human readable text that led to the selection of the (code, codeSystem)
- displayName - if present, the name as defined by the code system, or at least that does not change the meaning as defined by the code system
- displayName is not for machine processing (it has no semantic meaning and should be considered unreliable) and should only be used for debugging. If it is displayed to the user, one has to consider that it is considered unreliable.
- Goals and status of goals
- how to track status or progress against a goal
- currently the status of a goal is fixed to "active"
- this could be handled by making an observation about a goal - perhaps via some sort of goal observation template.
- Action - Alok to submit DSTU comment asking how to track progress against a goal and to mark a goal as completed.
- Code/Code translation - see File:Combination Codes (BDF)(GD) V2.pptx
- for example, SNOMED-CT code a "best possible" and translation code is a "more specific" local code. The translation may have its own translation.
- Action - Gay and Ben to continue the discussion with SDWG at HL7 WGM
- Date usage in results, allergies, problems, etc.
- Results
- where do we capture order date, specimen collection date, date reported etc.?
- the specimen collection date is the most relevant date
- no consensus on how to do this
- Allergies
- This was addressed in the C-CDA R2.1 Companion Guide (section 4.3.1 - Allergies)
- However there appears to be an outstanding issue with problems and concerns.
- Action - Brett to follow up with SDWG
- Care Plan certification test data from 170.315_b9_CP_Amb and _Inp
- Should there be a medication section? The test data has a medications table (along with vital signs, etc.)...which suggests that a medications section (or vital signs section) should be present.
- While this was what was imagined by the ONC (per Dragon), the validator does not require these section be present for this document.
- It is acceptable to include these sections for a care plan, but it is not required
- Intersex birth certificate
- Partial refills for medication
- Partial fills (e.g. 1.5 refills) is an odd concept. What is seems to be the issue is the quantity of drug that remains to be dispensed. The number of fills should always be a whole number.
VSAC Presentation
|
| Ask the ONC
|
Q: When using the scorecard API - there are problems with retrieving/viewing the PDF. JSON is OK.
A: Dragon to follow up.
Q: Some scorecard issues that were raised at prior IATs have not yet been addressed. How would people know when changes are made to the scorecard?
A: Release notes are published.
Q: Should the scorecard be scenario or context sensitive?
A: This was not the original intent and there is belief that there is value in not being scenario based.
Q: The current release cycle has been very well received by the vendor community. Has it been good for the ONC as well?
A: Yes - it has allowed for regular release planning. 1.0 release is planned for March 2017.
Q: Care plan validation rules appear to be light. Is that the case?
A: Yes.
|
Files
Homework Scenarios
Scenario Samples
Pre-Published Patient Information
Pre-Published Practitioner Information
