January 2013 CBCC Working Group Meeting - Phoenix
Contents
28th Annual Plenary & Working Group Meeting - San Antonio, TX USA
Community Based Collaborative Care (CBCC) WORKING GROUP SESSIONS
Agenda
| Day | Date | Qtr | Time | Session Type | Event | Session Leader | Room | On-line Meeting Link and Call-in Information |
| Sunday | JAN 13 | Q1 | 9:00-10:30 | . | No Meeting | . | . | |
| Q2 | 11:00-12:30 | . | No Meeting | . | . | |||
| Q3 | 1:45 -3:00 | . | No Meeting | . | . | |||
| Q4 | 3:30 -5:00 | . | No Meeting | . | . | |||
| Monday | JAN 14 | Q1 | 9:00-10:30 | . | No Meeting | . | . | |
| Q2 | 11:00-12:30 | . | No Meeting | . | . | |||
| Q3 | 1:45 -3:00
Joint with SECURITY |
  |
CBCC-Security Joint Project Updates (5-10min each) current, on-going projects
Ballot Overview:
Continuing Work
FHIR
|
CBCC | Nueces | Join online meeting https://meet.RTC.VA.GOV/suzanne.gonzales-webb/RF3WDW15
Dial in: +1 770-657-9270 Participant Passcode: 994563 | ||
| Q4 | 3:30 -5:00
Joint with SECURITY |
|
NEW discussion items; NEW projects
Around the Room informal report out of Security and Privacy advancements since last WGM
|
Security | Nueces | Join online meeting https://meet.RTC.VA.GOV/suzanne.gonzales-webb/RF3WDW15
Dial in: +1 770-657-9270 Participant Passcode: 994563 | ||
| Tuesday | JAN 15 | Q1 | 9:00-10:30 | . | No Meeting | . | . | |
| Q2 | 11:00-12:30 | . |
|
. | Garden Terrace #129 | Join online meeting
https://meet.RTC.VA.GOV/suzanne.gonzales-webb/402NMQ77 770-657-9270, Participant Code: 994563 | ||
| Q3 | 1:45-3:00 | |
|
CBCC | Garden Terrace #137 | Join online meeting
https://meet.RTC.VA.GOV/suzanne.gonzales-webb/402NMQ77 770-657-9270, Participant Code: 994563 | ||
| Q4 | 3:30 - 5:00 | |
|
CBCC | Garden Terrace #137 | Join online meeting
https://meet.RTC.VA.GOV/suzanne.gonzales-webb/402NMQ77 770-657-9270, Participant Code: 994563 | ||
| Wednesday | JAN 16 | Q1 | 9:00-10:30 | |
TBD | CBCC | TBD | |
| Q2 | 11:00-12:30 | . | Joint w/EHR, Sec, SOA | . | Rio Grande West | |||
| Q3 | 1:45 -3:00 | |
|
. | Garden Terrace #135 | Phone: +1 770-657-9270, Participant Code: 994563 Join online meeting https://meet.RTC.VA.GOV/suzanne.gonzales-webb/ZDCH6TD4 | ||
| Q4 | 3:30 -5:00 | |
|
. | Garden Terrace #135 | Phone: +1 770-657-9270, Participant Code: 994563 Join online meeting https://meet.RTC.VA.GOV/suzanne.gonzales-webb/ZDCH6TD4 | ||
| Thursday | JAN 17 | Q1 | 9:00-10:30 | . | No Meeting | . | . | |
| Q2 | 11:00-12:30 | . | No Meeting | . | . | |||
| Q3 | 1:45 - 3:00 | . | No Meeting | . | . | |||
| Q4 | 3:30 - 5:00 | . | No Meeting | . | . | |||
| Friday | JAN 18 | Q1 | 9:00-10:30 | . | No Meeting | . | . | |
| Q2 | 11:00-12:30 | . | No Meeting | . | . | |||
| Q3 | 1:45 -3:00 | . | No Meeting | . | . | |||
| Q4 | 3:30 -5:00 | . | No Meeting | . | . |
Q1=9:00 – 10:30 am; Q2=11:00 – 12:30 pm; Q3=1:45 – 3:00 pm; Q4=3:30 – 5:00 pm
Meeting Minutes - DRAFT
2013 HL7 WGM - Cambridge
MONDAY Q3 Joint Security-CBCC
- Welcome and Introductions
- One of the use cases in interest of ONC which is applicable to others (Patient self-pay) in the US realm PSP gives the patient some right who sees records for which they are seeing who they pay out to. (First time shown) related to standards at their core (including ISO) HL7 has been the lead in this area to date. Along with FHIR and resources we have brought into play labeling concepts to include RBAC and others.
- Intent is to show this use case in the FHIR context.
- Agenda Review
- Duane DeCouteau HCS and SLS demonstration
Applying tags based on organizational policy (AUS is a 42CFR ) Patient has cancer, chosen to pay for treatments and has chosen not to share sensitive data Depending on how you tag the data Since patient consents changes, and organizational policy changes…everything is done dynamically. (example shown) based on values coming across (see Security Classification System, Figure 4 HCS Operational Model), 7.1 HCS General functional Model (Walter) when dealing with unstructured data (i.e. pdf file) This relies on structured data so this is an Achilles heel. Most data is human readable and this is a problem, we are anticipating more structured data or using “structured data: that is producing the unstructured narrative. Note that this also a process in time---we are currently in ballot; the next step is to have vendors to realize. (see DS4P Example: Dermatology Clinic, 1/10/2014 ONC HIT slide) The semantics of the data is one then, the accuracy of the law (is the second) The expression of the certain law into a computable way as well as implementable We are providing the policy outcome; See SPIFF The legal rule UML, is looking for a mechanism for a more mathematic… which requires non-rule based (Bernd) ‘fuzzy law’ (Kathleen) the PMAC is what is used is the filter used to encode the laws The legal representation. in order to preserve the fidelity of the law, we may need to look beyond the SAML assentation, etc., Is the contextual body of the all the information, that creates the degree of sensitivity of information (Mike) we that the way we do this is through the rules engine itself; it will have the clinical diagnosis for which it was prescribed…nature of underlying conditions, etc. provenance is an important...following the work in ts area. We have to have a holistic view incorporated into the drools engine, tagginment Drugs can be used off label On the encounter it will however list the diagnosis This goes back to design requirements of EHR, if you want to have persistent data across you have to have that every drug prescribe has the stated event. FHIR – Duane had difficulty with moving to latest version. Consumed most of his time. By moving to latest release, decreased ability to communicate with previously most available networks. However, many of the bugs have been removed. (David) – What level of authentication is supported by FHIR, whether the assurance level is adequate for the amount of information exchanged? (Duane) security came up on the last day. The policy should be coming out of the security rather than FHIR David – the LOA is capped out and you can use SAML, Q4 – Tony Mallia will give a brief overview of FHIR (not a security view) Commander Makarski – semantics questions Mike – relationship based access control; in ReBAC the RE of the patient to the different entities would determine the kind of information they would have access to. In or outside of the HCS, this is an area of which we are investigating. There is nothing formal yet, we have a couple of white papers, Kathleen – there is compartments in the security labels To a large degree we are content agnostic.
- How to get this type of technology to service?
Ballot Overview:
- Healthcare Classification Scheme (HCS) – on Wednesday
- Security Label Service (SLS) – on Wednesday
- DS4P – Johnathan, Serafina and Ioana not currently present
' Continuing Work'
- Composite Security and Privacy DAM/Information Model
- Security and Privacy Ontology, future directions (15-20 min)
' FHIR '
- FHIR Demonstration/Discussion of HCS Implementation - Duane DeCouteau
- FHIR Accounting Disclosures - John Moehrke
If had a structure for BP
MONDAY Q4 Joint Security-CBCC
' Presentation: Implementing FHIR resource profiles as SML schemas (Tony Mallia), 25 minutes '
- What is currently involved in interoperability, how does FHIR work in the exchange
- SLIDE 2 – Problem Statement
- Overall Proposed Process (Other paths are possible)
- CDA taken out of CDA can be used here – on the right hand side, which are currently not shown here (per Kathleen)
- <<semantype>> “semantic type; we have abstract type of observation
- Slide 7; top left is FHIR piece
- Restrict to bloodpressure.component > …> to LOINC code
- Slide 11: if you take the xi:type out, this becomes FHIR compliant
Kathleen: did you look at VMR at all? As part of decision support TonyM – I believe that Emory is working on that, because I’m not the one to declare the components.
- Slide 12 : eclipse; it knows when I’ve fat fingered stuff ;
- Slide 13: conclusion
Bernd and Tony M will be talking off line, regarding a project that began in 1989. You’ll have an observation to tag… IF YOU decompose the system, if you have the same composition Tony Weida – HL7 Security and Privacy Ontology Presentation: SPO - January 2014 HL7 WGM.pptx
- Ballot and is now a normative specification
- Outline – recap, future
Can you generate rules from OWL? In some sense we are already using OWL like a rules system, SWORL lets us write rules in terms of OWL and then use more and run them against the ontology and can use this to populate rules, rules engines
- Future directions (review)
Ontology browser software hosted http://216.47.173.3:8080/browser/
Ioana has developed style sheets in the CDA that are simplistic for the patient view and can be written up.
Physician response is not completely understood by patient, we need a mapping between the physicolican space of the physician to the feeling of patient description. Then it is not a translation in terms of working but a concept translation. The outcome was demonstration in scientific words and
Mike feels the need to formalize this better, we are proposing in the…
For natural language for patients… Suzanne will send out the draft PSS to the Security WG, CBCC already has
Security Tuesday Q1 – WGM
Serafina will be in Security Q2/Q3 for DS4P; Johnathan is available Q1, Q3
There is a formal request that is submitted for publication for Security and Privacy Ontology (per tony)
PASS SLS
Corrections to the Agenda made:
New items:
Scope Statement from CBCC – Natural Language Interface for patients (new start project); not sure how we’re going to do this; others have approached this before unsuccessfully (challenge) what make sense we have input which we need to study. Speaking to Dr. Hobbs – he is doing a similar thing on the clinical thing—patient expressing their symptoms to their MDs; conditions to their physician; with the goal of the patient able to express in their way their request on how their information for handling their healthcare data is handled. Patients are not able to express their exact desire accurately.
' Updates to Thursday:'
Updates Q2 – Red flag (notes from Kathleen… e-mail string?)
Security and Privacy DAM is going to expire in February; needs to go informative;
- Recommends (prefer) to move materiel as is, and go to R2 (to bring in current work)
Consent CDA is going to expire in February – need to figure out how to move forward or if already; ONC (per Johnathan) is interested in moving this forward; DS4P also references the CDA
Security SOA written up two years to pull all the security pieces together (SOA piece already incorporates the DAM)
- Bernd does not feel that the Security DAM is not an issue of SOA and we should not house this information in SOA.
- This is to the IM into compliance with SAIF
- Security still owns the Security projects (our DAM is still owned by Security)
- In order to move forward to normative the DAM will be named to Conceptual Information Model which is required for the SOA architecture….leaving DAM as is
Natural language style sheet – project may be redundant to the natural language… making a CDA form which populates the CDA
New Item: EHR for CR Thursday Q1
Change topic to the SLS ballot in place of John Moehrke’s FHIR talk on Wednesday
Beginning Q2 – reviews of DS4P; Johnathan not here but Serafina will be
Wed Q1 is EHR joint; Mike spoke to Gary and Steve Hufnagel about the EHR vocabulary that they’ve come up with; Tony W will be taking a look at that in respect to the ontology. Mike would like to be out of the owing vocabulary; rationale is that if we can move the stuff we develop into the EHR functional model and we integrate into their model; it will be more universal (but this becomes a Trojan horse situation)
- EHR (Kathleen) has not managed their vocabulary it would not pass a SMILE test
- Bernd – there was work to transfer out vocabulary work… LOINC to move IHTO, (?) ten years ago was creating their vocab independently and has now referred out to LOINC SNOMED CT, etc.; decision has made to refrain from maintaining vocabulary and we would refer and allow SNOMED CT or whomever to maintain the vocabulary; ITSIDO… should be management otherwise we will not have interoperability. This is related to Tony Weida…
- Mike does not want to own any clinical data as a security person
- There needs to be a transition path to harmonize the vocab with EHR, work with Tony W; working with an ontological approach with ultimate goal. EHR is willing to do this
- RBAC was a gab that needed to be filled
- Needs to have some semblance to some underlying system and EHR doesn’t; this also disagrees with the integrity concepts even though they come from established standards
- (discussion to continue on Wednesday)
CBCC TUESDAY Q2 – reconciliation with SOA
Q3, Q4 tutorials
Thursday Q1 – Red flag and other presentation
Q2 – future planning Q3 – FHIR if they show Suzanne/CDA consent Q4 –
TUESDAY Q2 CBCC
Agenda
Reviewed CIC DAMs (Lori) –, several items need rewriting, presenting proposed changed at Q3/Q4 at CIC on Wednesday.
The BH DAM
- Major Depression DAM (CIC)
- Schizophrenia DAM (CIC)
- It would makes sense for them to be rolled up in the BH DAM
- Lori spoke with CIC to figure out to make this a unified rather than individual disease states DAMs (they were supportive
- If this is the approach to take, since the BH Dam has already been balloted and now we have two disease state DAMs to roll them together rather than handle individually; how would one go about doing that?
- Both MD and Schizophrenia were balloted (?) unknown
- ACTION ITEM: (Lori) with Anita Walden to confirm whether or not the individual DAMs were balloted or not.
- The reason they were completing individual disease states in relation to their funding. It appears that where they received their funding may also be receptive to rolling up the individual DAMs to the BH DAMs (to be confirmed)
- Per e-mail: DAMs Balloted but they are still waiting on input from APA (Steve/Lori)
- (Lori) Two aspects to the CIC DAMs
- Flow chart – sequence of events when someone comes into initiate into MH system
- Two other diagrams (major) which spoke of the fields and which information needs to be stored.
In the activity portions (see flow chart diagrams); do a prelim psychological assessment; the flow is not correct; certain areas are very specific and others where they are not.
Clinicians need to be involved (per CIC) to assist with the flow chart
- Content fields (second aspect) – data fields needed to support both mental history and mental health
- Medical history
- Mental health episodes
- Content fields (second aspect) – data fields needed to support both mental history and mental health
- Diagnostic fields to support depression, schizophrenia
- Issue: what CIC is doing, is information that is unique to each individual disease state.
(Lori sending two diagrams to Suzanne); received sgw Lori has reviewed the BH DAM briefly but has not had time to review the details; if the Major Depressive/Medical History, etc. from CIC and insert in the BH model, then this would funnel to one DAM. If we want to expand the BH dam to include the CIC individual disease states then this would be the way to do it
EHR has a breakout session/regarding the EHR v1.0; with the 5-year expiration, would like ‘us’ (Steve Daviss) to take this on. Do we want to ask any of the EHR folks for them to come over to CBCC to start the discussion? Who needs to be involved? (HL7 and others…?) Do we need a project scope statement? V2 of the EHR functional model has major work to be done and is currently no workable. (In EHR meeting Q4)
- How do enter a BH piece if they do not have their baseline ready?
- Is V2 already balloted and they’ve started a V3
- If we have them over for Q3, we can have these questions answered
- What do we want to accomplish? – what is the ultimate purpose
- ACTION ITEM: Steve will invite EHR (via e-mail) to come to CBCC Q3 as part of their break out session.
- We want to make sure the vendor piece is also incorporated into the v2
- Don is currently working on cross-circulation APA,CCHIT,HL7 we might be able to have this cross-tab available next week; there is not a lot of overlap – which is some concern
- If there are too much academia the approach is not realistic (vs. real life) this may interfere
TUESDAY Q3 – CBCC
Should we be bothering with the EHR Functional model update because… (Profile can be done by anyone)
- Model is a given list at any point and time of all the capabilities that the EHR profile could run and all the conformance criteria associated with this material that is meant by the function.
- Limited to the EHR capabilities could be. There was never an expectation that any would create an EHR with all of those capabilities (these are subsets). There was a chapter 2 of the FModel and what it means to conform to the FModel. The reason it’s important is that it can claim—its subset, think what be added should conclude. One in theory could create a profile that would include 7 or 8 basic functions (which are subsets of the EHR.specfied in the model... or new stuff) the FModel has several wrinkles. If you accept a particular function into your profile then you obliged to accept all of its associated conformance criteria that are designated as SHALLs, you don’t have to accept conformance criteria which are SHOULDs or MAYs – if you do accept, you can change the verb… (Can go up but not down)…
- There are a number of overarching functions that you have to have—they also have implied various functions in their own text. ** i.e. EHR has to be able to provide capability to “blblblblb…” it requires function 36… as well as function 37… what that chain does in some context---is reasonable. But in other context it is virtually impossible to come up with an ‘EHR-lite’
- There is a tool to create profile; if you just had the minimum of functions absolutely required you ended up with a very large document. In most BH, you don’t need a complex document
- Clinical data is limited as compared to other specialties
- Lori, Steve do not agree with this statement but understand the implication
- Jim is referring to both V2 and V1
- V2 was just balloted (reconciliation in progress or just completed); and is much larger than V1…addition of detailed information, increased number of sections, etc., the concern is what is driving EHR to do a V3---who found missing/gaps in v2?
There is going to be a v2.1; there is a significant effort to get the functional model adopted by ISO (they go through their own balloting issues—which will drive some changes to the FModel) when those changes are balloted in ISO---this will become v2.1 [Current v2 will go through ISO and CEN balloting]
- (Lori) is the intent to incorporate all specialties
- (Steve) then this will make BH FM a part of v2
There are two stages to development for standard development
- DSTU – draft standard for trial use… 60% concurrent for ballot, after reconciliation so if negatives are raised…and then resolved (happy w/resolution) 60% … to make ANSI standard must have 90% concurrence
- # of clinicians involved with the writing of V2?
- Reaffirmation EHR BH Functional Model sent to Steve & Lori (from ballot site) by Suzanne
The intent of the model was to provide a shopping list of functions that any large BH provider, or state substance abuse treatment, would start with…for an acquisition document. (What do we want an EHR to do?) Conformance in this context is something that the EHR vendor asserts that they are in conformance with the specified profile.
The wording has changed and the organization has changed somewhat—there are statements that would not be included in a BH profile. At the margins of the v2 EHR FModel gets into items that you would expect. i.e. education for staff… etc., why that would be in an FModel is hard to understand—whether or not you want to accept those functions or not. Every function must have at least one SHALL to be included. If you have SHOULD you should have an assertion as to by what time it will be converted to a SHALL… i.e. in five years when xx becomes available…etc.
To move forward: does it make sense to convert the v1 profile to v2, and then take a loo The way you start (per Jim) start with v2 and become familiar with the conformance chapter—whomever you get to work on the profile needs to agree
- Skipping a functions may be tricky
- If a profile is a subset, you need to know ahead of time if it’s associated with something else. i.e. association with several SHALL—if removed, then all the associated SHALLs must be removed as well. (within the limits of the conformance chapter)
In order
Is there a function in v2 that talks about demographics?
The original profile took 2 years because the functional model kept changing.
- Showing CCHIT BH (Steve Davis Screen) excel spreadsheet.
- CCHIT Certified IFR Stage BEHAVIORAL Health Criteria 20100202 document shown
Start with a small group – then get to a level of detail then go to a broader group to flush out; make user we have everything… basically being through before
- Pull the information we want from v2
- Figure out what is missing or needs to be added to the existing functionality (part of the conformance
- Scope; are we strictly talking about psychiatry… but BH (needs to be defined…ONC is working on this next week)… social workers, etc.
The people who we want involved include:
- SAMHSA, psychologists, case managers, social workers, professional counselors, psychiatric nurses, case managers, software vendors/developers; insurance company (discussion tabled for now)
Documents sent to Suzanne from Lori ( )
TUESDAY Q4 - CBCC
Initiation of Project Scope Statement: Behavioral Health EHR Functional Profile v2
Wednesday Q1 – Joint CBCC, Security, SOA, EHR
Security Vocabulary Security Profile Behavioral Health – Functional Profile RMES vs. Security Profile (how they relate)
Updates: SOA CBCC Security EHR
We have a RBAC standard within Security that has a permission catalog and other supporting documentation… (Objects, verbs)
V2 standards… we have a table of verbs ‘permitted actions’ on objects found in an EHR.
CRUDEA – create, read update delete, execute, append
- Security does not want to own either the verbs or the healthcare objects – this was created prior to another list being found in HL7
- The verbs—we have been arguing with the hierarchy they need to be put into (we do have an ontology… there are distinctions between the verbs
- Steve and Gary have been working on this; challenge made to Tony to create a systematic approach to organize
- We have worn out two sets of ontologists
- Some of the verbs are similar to the ones found in EHR. We would like to harmonize the vocabulary so that the security functional model can deal with only one set of terminology/vocabulary.
- The security classification system is a different animal… labels… labels are attributes. The way the security works is ‘do you have a clearance’ to access that object; (that is equal to or dominates the access to that object)
- You apply the application to the object
- The general case in security for binding, is that binding is done at run-time... it is done by rules. The rules work on structural data; the policy… (patients can change rules day-by-day); based on the rules/done at run time.
- Security worked very with FHIR
- We demonstrated the ability to … self-pay…
- This is a normative set of verbs; we’re reluctant to continue this in an unstructured way—it will make it worse; can this be done in a repeatable process… outside of the original CRUDEA, security we don’t have an answer why we should continue to house this in security.
Note that security has one set of verbs
EHR space has two sets of verbs – (ISO21089 (?) … maybe a total set of 4 sets of verbs total
- SOA has a set of ontology which is a current project ; they are interested in also being involved---
Security Ontology Balloted: Normative v2 RBAC Verbs Balloted: Normative v2 EHR Verb Hierarchy Based Referenced Record infrastructure verbs Based on ISO 21089 (complementary) Normative SOA verb ontology For comment only ballot
A rich structure vocabulary (ontology) classifies in a taxonomy (concept hierarchy), connecting in other meaningful and useful way.
Tony Weida (presentation) : <<add>>
Currently Security and SOA are using the ontology.
SOA – Zoran (AUS)
CBCC – Suzanne
EHR – John Ritter
Security – Mike and Tony
First task: define the charter, start project scope
Lori gave presentation on recommendations for CIC BH DAMS (MD and Schizophrenia), general consensus was interested and supported discussion continued through lunch (Meredith Zozus, CIC primary contact from CIC on the MD and Schizophrenia project and is interested in ongoing collaboration); they have an addition 60 more individual
John Ritter and Don Mon met with Steve Daviss, Lori – processes review, (add CIC to PSS); Public Health Functional profile, based on v1 EHR functional model… funding from CDC. Without the funding it would have taken a lot longer…we’re doing the same thing in BH (the interest is SAMHSA, ONC)
Lori will be including a one slide BH model update in order to put the bug in on requesting funding; why would you want to develop a v2?
Virtual briefing Tuesday 1/29 from 9-1 ET – provider communities speakers from different provider organization; a series of webinars
There are very devices using MU2 (meaningful use) 1 out of 8 doctors who can actually use. There is a arege fall of products... many of the providers who are not able to use v2 are in big trouble for translating data which means the certifications of the software are being misled.
If this is the case (1 of 8) does this mean a lot of people will lose their funding?
- Unsure
- The first year you attest a particular stage, you only have to prove a year of data
- Valant EHR behavioral health Valant medical solution” their EHR is strictly for behavioral health
Richard will be heading to the annual ONC meeting to listen in on the MU2 certification (Meaningful Use) the Adoption of it… the delay of it
18 billion dollars spent on incentives to...
WEDNESDAY Q3 - CBCC
Review of
To be added
WEDNESDAY Q4 - CBCC
Teleconference
To be added
