RE: SBIR proposal: Prototype mediation service for Open-Source, Universal Healthcare Exchange Language: Pilot connecting VistA and OSCAR

Got it.  Thanks Michel.

charlie

________________________________
From: Michel Dumontier [michel.dumontier@gmail.com]
Sent: Thursday, June 21, 2012 12:26 PM
To: Mead, Charlie (NIH/NCI) [C]; HCLS
Cc: Eric Prud'hommeaux
Subject: Re: SBIR proposal: Prototype mediation service for Open-Source, Universal Healthcare Exchange Language: Pilot connecting VistA and OSCAR

Charlie,
  the public mailing list for the HCLS is - public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>

  the public calendar for the HCLS is the google calendar; if you login to the google calendar with w3c.hcls / hclshcls, you can click on your next date, add the agenda info to the box, save and send notice to participants.


m.

On Thu, Jun 21, 2012 at 10:59 AM, Mead, Charlie (NIH/NCI) [C] <meadch@mail.nih.gov<mailto:meadch@mail.nih.gov>> wrote:
SHould I send out a meeting annoucement for next TUesday's meeting to:  w3c.hcls@gmail.com<mailto:w3c.hcls@gmail.com><mailto:w3c.hcls@gmail.com<mailto:w3c.hcls@gmail.com>>?

________________________________
From: Michel Dumontier [michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>]
Sent: Thursday, June 21, 2012 10:40 AM
To: team-hcls-chairs@w3.org<mailto:team-hcls-chairs@w3.org>
Subject: Fwd: SBIR proposal: Prototype mediation service for Open-Source, Universal Healthcare Exchange Language: Pilot connecting VistA and OSCAR

Charlie / Eric
 here's a conversation i'm having that needs to happen on Tuesday as well.

m.

---------- Forwarded message ----------
From: Rafael Richards <rmrich5@gmail.com<mailto:rmrich5@gmail.com><mailto:rmrich5@gmail.com<mailto:rmrich5@gmail.com>>>
Date: Thu, Jun 21, 2012 at 10:37 AM
Subject: Re: SBIR proposal: Prototype mediation service for Open-Source, Universal Healthcare Exchange Language: Pilot connecting VistA and OSCAR
To: Michel Dumontier <michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com><mailto:michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>>>


Michel,

I would be delighted to discuss at next Tuesday's HCLS forum.

The grant is looking for a live, active mediation service that could broker patient data from any EHR, recognize what language the data is in  (VistA-ese, OSCAR-ese, etc.), and pass it to a polling recipient system. Ideally this will be a pull mechanism that allows one EHR to ask for specific data elements of a specific patient in another EHR, and obtain this without human intervention. How would we design such a system?


As the deadline for this SBIR proposal is at the end of the month, it leaves little time to apply.   Is there any small company within the HCLS community that would be a vehicle to apply for this grant, and engage the HCLS community in helping solve this very basic issue?

Rafael





On Tue, Jun 19, 2012 at 3:09 PM, Michel Dumontier <michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com><mailto:michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>>> wrote:
Hi Rafael,
 Well this is really cool :)

 As you may be aware, Eric, Charlie and I are working on providing feedback to HL7 and their FHIR initiative, which includes an RDF representation of the patient record.  Given that we have already done work [1] to integrate (synthetic, indivo-based) patient records with linked open data, there would be a prototype to follow, but this, i imagine, is significantly larger and has more potential impact. Would you care to discuss at next Tuesday's HCLS Health Care forum?

m.

[1]  http://www.ncbi.nlm.nih.gov/pubmed/21624155


On Mon, Jun 11, 2012 at 8:03 AM, Rafael Richards <rmrich5@gmail.com<mailto:rmrich5@gmail.com><mailto:rmrich5@gmail.com<mailto:rmrich5@gmail.com>>> wrote:
Michel,

The PCAST recommendation for a "Universal exchange language" ,  and  need for mediation services between RDF sources  might make a nice topic on HCLSIG.

Any thoughts on projecting OSCAR's  relational tables to RDF?  Mediation methods between EHR endpoints?

Please forward to the OSCAR development team and anyone in the HCLSIG community interested.

There is funding available for this  (see SBIR below).


Rafael


__________________

NOTES


>From the press releases below, it looks like VA and DoD are looking hard at 3M Health Data Dictionary as their semantic interoperability mediation.   The SBIR proposal states:

"Through their Health Data Dictionary (HDD) product, 3M has also supported some degree of semantic interoperability, using a knowledge representation scheme linked to a unique concept identifier, but it is not a true first-order predicate logic ontology".

=> How does  lack "true first-order predicate logic ontology" affect the HDD performance?
=> Woud an RDF mediation service have "true first-order predicate logic"?
=> Why would an RDF store and mediation offer better functionality than the 3M HDD API?

These are a few of the question we would need to answer in the SBIR.

As a pilot  why don't we also consider another open-source, community supported EHR that is well connected to the SemWeb community? (OSCAR).

Thoughts  on piloting an VistA -OSCAR - SMART   RDF mediation service?


Rafael

_________________________________________________

Details on 3M HDD

Press release on VA-DoD use of 3M HDD for interoperability:
http://news.3m.com/press-release/company/3m-opens-access-3m-healthcare-data-dictionary-under-agreement-us-departments-d

Timeline for incorporation of VA/DoD dictionaries into 3M HDD
http://www.hddaccess.com/timeline/

YouTube explanation of 3M HDD:
http://youtu.be/T_Qc93efnzI

HDD Overview:
http://www.hddaccess.com/intro/overview/

HDD  Demo:
http://search.hddaccess.com/
http://search.hddaccess.com/demo


This is the ideal opportunity for us get some traction for RDF as the PCAST report's  "Universal Exchange Language".

The SBIR grant is below, and link here:

http://www.dodsbir.net/sitis/display_topic.asp?Bookmark=42567


Proposals are due at the end of June.

This is a Three-stage SBIR, so potentially  large sums will be heading the RDF way if we can get this started.

Rafael





Proposals Accepted:
Program:        SBIR
Topic Number:   DHP12-004 (DHP)
Title:          Prototype, Open-Source, Universal Healthcare Exchange Language
Research & Technical Areas:     Biomedical
________________________________

Acquisition Program:    Office of the Principal Assistant for Acquisition – USAMRMC
 Objective:    Prototype architecture to execute an open source, universal health exchange language, as described in a recent President’s Council Of Advisors On Science And Technology (PCAST) Report.

 Description:          Most DOD, VHA, and civilian healthcare systems encounter significant challenges in exchanging health information due to the lack of a universal health exchange language. Without effective health information exchange, continuity of patient care is less than optimal, healthcare access and availability are hampered, and healthcare costs increase. This topic is of particular importance to DOD, which is a provider and payer of care for 9.6M beneficiaries. Two thirds of military healthcare is delivered in the civilian sector. As a matter of background, the PCAST Report On “Realizing The Full Potential Of Health Information Technology To Improve Healthcare For Americans: The Path Forward”, sets forth general recommendations regarding how healthcare information technology can be used to improve healthcare access, availability, acceptability, continuity, cost-effectiveness, and quality. The report cites the beneficial work of the HHS Office of the National Coordinator (ONC) for Healthcare Information Technology in developing standards and an initial nationwide healthcare information exchange to share data to improve healthcare delivery and support research for the public good. The report points out that “national decisions can and should be made soon to establish a “universal exchange language” that enables health IT data to be shared across institutions; and also to create the infrastructure that allows physicians and patients to assemble a patient’s data across institutional boundaries, subject to strong, persistent, privacy safeguards and consistent with applicable patient privacy preferences.” The PCAST report further states that, “creating the required capabilities is technically feasible, as demonstrated by technology frameworks with demonstrated success in other sectors of the economy. The best way to manage and store data for advanced health informatics is to break data down into the smallest individual pieces that make sense to exchange or aggregate. These individual pieces are called ‘tagged data elements, because each unit of data is accompanied by a mandatory “meta­data tag” that describes the attributes, provenance, and required security protections of the data.” Current state: Some research surrounding use of healthcare metadata is underway and moving from academic labs into some early practical uses, such as demonstrated by Dr. Parsa Mirhaji for use in public health surveillance. Much of this research has been conducted using semantic web technologies employing the Resource Descriptive Framework (RDF), and Web Ontology Language (OWL), and query technologies such as SPARQL, which utilize the concept of “tuples”, (subject, predicate, object), to relate data and achieve semantic interoperability. Other similar technologies exist, such as those from Metadata, Inc., although the semi-proprietary Metadata language, parts of which is available from Open Health Tools, is based on quintuplets vice tuples. Through their Health Data Dictionary (HDD) product, 3M has also supported some degree of semantic interoperability, using a knowledge representation scheme linked to a unique concept identifier, but it is not a true first-order predicate logic ontology. Language and Computing, now owned by Nuance, developed LinkBase, the world’s largest medical ontology, with limited commercial success. Adoption of these commercial technologies may have been limited by their proprietary, versus open nature, and associated licensing costs. It may be possible for some of these commercial technologies to be made open source, with companies then selling services around their technology, but this is a business decision that is up to the companies. Moving towards developing open source terminology mediation services may place these companies in a position wherein the Office of the National Coordinator for Healthcare IT could adopt those terminology mediation services for use in the Nationwide Health Information Network (NwHIN), and NwHIN Connect and Direct products (typically using a Berkeley Software Development License). In any event, the environment is ripe for continued academic and commercial collaboration under an STTR to advance the domain. Desired State: Research conducted under this topic will directly support the PCAST and ONC visions, but would be conducted primarily on behalf of the Military Health System, which provides an integrated healthcare delivery system for 9.6 million beneficiaries. This care is delivered through a combination of direct care Military Treatment Facilities and private healthcare delivery organizations under the TRICARE triple option health benefits program. Nearly two thirds of healthcare is delivered to military beneficiaries through the private sector. Given that military families are also highly mobile, moving on average every three years to new duty stations, finding a way to exchange data and create a longitudinal virtual electronic health record is an important objective of military medicine. Clearly the research would also be extensible to other national publically and privately funded healthcare delivery systems and information exchanges. The prototype to develop a universal exchange language for healthcare information and a digital infrastructure for locating patient records while strictly ensuring patient privacy may employ the U.S. Army Telemedicine and Advanced Technology Research Center’s (TATRC) Early Stage Platform (ESP) for Research and Development, which provides a fully replicate DOD Electronic Health Record and CHCS computerized physician order entry and results retrieval system for third party development, using virtual machine access. TATRC will coordinate this research closely with HHS, Office of the Nationwide Healthcare Information Coordinator, the Veterans Administration, and with the Center for Medicare Services (CMS). It is fully expected that the research will be extendable into the public good and will benefit the development of new electronic health records developed by the private sector that would utilize the universal exchange language.


       PHASE I: In Phase I, the awardee will outline a strategic, operational, and technical alternatives to creating a prototype, open source, universal health exchange language service that can operate as a service on the NwHIN, and support health exchange for military medicine. Phase I work should center on a limited number of use cases to be determined in conjunction with the government Contracting Officer Representative. Phase I will also provide opportunity for consultation with ONC for Healthcare IT and other subject matter experts. At the conclusion of Phase I, the awardee will recommend a technical reference implementation architecture, which will then be built as a prototype in Phase II of the SBIR.


       PHASE II: In Phase II, the awardee will build the universal health language prototype, as an open source service on the NwHIN Connect solution, and demonstrate the exchange of military, VA, and civilian health data with semantic interoperability in a laboratory setting, potentially using the TATRC early stage platform for research and development. At the conclusion of Phase II, the prototype will be demonstrated to U.S. Army TATRC, Military Health System, HHS ONC for Healthcare IT, HHS CMS, Veterans Administration, and other government officials.


       PHASE III: In Phase III, the universal health language service would be implemented on the NwHIN, or otherwise in local or regional health information exchanges to support scalable terminology mediation between electronic health records systems. Such work may also be commercialized and of interest to commercial electronic healthcare vendors. Ideally, the universal health language service would be an open source service, with the vendor choosing to sell services around the open technology.


 References:   1. President’s Council of Advisors on Science and Technology (PCAST) Report On “Realizing The Full Potential of Health Information Technology To Improve Healthcare for Americans: The Path Forward, “ December 2010, http://www.whitehouse.gov/sites/default/files/microsites/ostp/pcast-health-it-report.pdf 2. Marco Eichelberg, Thomas Aden And J Org Riesmeier, et al, “A Survey and Analysis of Electronic Healthcare Record Standards”, http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.107.5349&rep=rep1&type=pdf 3. Parsa Mirhaji, Dean Allemang, Robert Coyne, S. Ward Casscells: Improving the Public Health Information Network through Semantic Modeling. IEEE Intelligent Systems 22(3): 13-17 (2007), http://www.computer.org/portal/web/csdl/doi/10.1109/MIS.2007.53 4. Parsa Mirhaji, Min Zhu, Mattew Vagnoni, Elmer V. Bernstam, Jiajie Zhang, Jack W. Smith: Ontology driven integration platform for clinical and translational research. BMC Bioinformatics 10(S-2): (2009) http://www.biomedcentral.com/1471-2105/10/S2/S2 5. Marco Eichelberg, Thomas Aden And J Org Riesmeier, et al, “A Survey and Analysis of Electronic Healthcare Record Standards”, http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.107.5349&rep=rep1&type=pdf 6. Semantic Web Health Care and Life Sciences (HCLS) Interest Group http://www.w3.org/2001/sw/hcls/ 7. W3C Semantic Web in Health Care and Life Sciences Tutorial, http://videolectures.net/iswc08_prudhommeaux_swhcls/


Keywords:       PCAST Report, Healthcare Information Technology, Universal Health Exchange Language, Patient Record Locator, Patient Privacy
________________________________
Questions and Answers:
Q: Could you clarify some of the above requirements for the prototype:

1) Topic description refers (as current state) to exchange languages used by existing products, languages such as RDF, OWL and few proprietary ones. This followed by explanation: "...but it is not a true first-order predicate logic ontology"

Question: is inclusion of true first-order predicate logic ontology a requirement for successful exchange language?

2) It also mentioned among the obstacles "Adoption of these commercial technologies may have been limited by their proprietary, versus open nature, and associated licensing costs."

Question: In light of open source requirement is commercialization strategy for the prototype expected to be based on potential service fees from DoD and other potential governmental clients?

A: 1. No. However, the government advises that the proposer provide a clear description of the technologies, ontologies, and methods to be used in the project, along with a discussion of the advantages and disadvantages of those technologies, ontologies, and methods, and associated implementation risks, and maintenance costs.

2. The government prefers use of open, standards-based technologies, versus closed technologies. In addition, the government would prefer use of open source technologies, which are licensed under a variety of licensing models (i.e. BSD, GNU, and others), some of which are more business friendly than others. If the proposer bids such open source technologies, than it should clearly indicate in its commercialization plan how it would sell services around these open source technologies. With that said, some proposers may still choose to bid use of proprietary technologies with traditional end user licensing agreements or fee-based subscription services.

Q: Creating a full blown Open-Source, Universal Healthcare Exchange Language is a massive undertaking. We specialize in patient care in virtual worlds where patient data is collected and transmitted back to care givers. This transference necessarily demands an ontology. Are you interested in our creating a sub-standard for patient health data under the larger Open-Source, Universal Healthcare Exchange Language?

A: Perhaps. Note that there are many healthcare terminologies, lexicons, thesauri, controlled medical vocabularies, ontologies, and knowledge representation frameworks already. Our interest is primarily in how to automate mapping among many of these.


Q: So, is what you are really looking for is something like an online HealthCare Rosetta Stone?

A: Please define Rosetta Stone. If you are referring to Rosetta Stone™, a registered trademark for foreign language translation software, the answer is not really, but perhaps as related to healthcare. If you are referring to the terms Rosetta Stone in a more generic sense, and applied to healthcare, such as described in http://healthcarereformnow.blogspot.com/2007/12/health-care-rosetta-stone.html, and/or http://successfulworkplace.com/2012/03/16/translating-healthcares-rosetta-stone/, the answer is yes. There was also an HHS sponsored OASIS effort which developed a “Rosetta Stone” for Long Term Care terminology mediation, http://aspe.hhs.gov/daltcp/reports/2011/StratEng-E.htm; some of that thinking is applicable. Given all of this past history, though, the new 3M Open HDD may provide for universal health language exchange. Additional research is necessary on how this product would be implemented in electronic health records and health information exchanges.



Q: Successful execution of this project will require critical engagement with federal partners that can help guide the development of protocols and frameworks that meet the expectations of the proposed DoD/VA environment. Please comment on the anticipated accessibility of the health IT leadership at the DoD/VA to help the project team identify and execute the highest impact use cases across different phases of this project.

A: The Contracting Officer's Representative will make a best effort to integrate this work with a specific use case or more immediate research need which may present in the TRICARE Management Activity; HHS Office of the National Coordinator for Healthcare IT, Veteran's Administration; or DOD/VA Integrated Program Office. Key staff in these organizations are aware of the SBIR topic. With the announcement that certain aspects of the 3M HDD will be made open source, and considered for use in the DOD/VA joint integrated electronic health record; and a recent announcement that the VA has an RFI on the street to have a vendor assess the applicability of the Open 3M HDD to use with VA's VISTA system; this SBIR may be able to support that effort, and presents a possible route for technology transfer within government.

Q: A key validation of the PCAST framewrork is the extent to which it's able to address secondary and unexpected uses of healthcare data. Visionary and motivated sponsors within the federal government, specifically within MHS and VHA, are critical for helping define the PCAST framework. Thus, are candidate sponsors readily identifiable and available for applicants to interview prior to submission of proposals in response to this solicitation in order to better understand the MHS and/or VHA vision for the PCAST framework?

A: The Contracting Officer Representative will make a best attempt effort to have the recipient(s) of this SBIR Phase I funding work directly with representatives in the U.S. Army Telemedicine and Advanced Technology Research Center; Joint Program Committee-1; TRICARE Management Activity; HHS Office of the National Coordinator for Healthcare IT; VA; and DOD/VA Integrated Program Office (IPO), on this topic.

Q: Is the primary intent of this SBIR solicitation to demonstrate feasibility of the proposed UEL across a wide variety of use cases, or is it to address the technical scale and performance, as well as enterprise governance issues, for it to work in a realistic DoD/VA setting?

A: Given recent announcements at http://www.healthdatamanagement.com/news/3M-data-dictionary-open-source-interoperability-coding-44468-1.html, efforts are underway to determine if the VA and DOD/VA Integrated Program Office are interested in applying the effort of this SBIR to analyzing the use of the Open 3M HDD with the DOD/VA Integrated Electronic Health Record, which is under development.

Update from TPOC w/ additional information:
Open 3M HDD Access at: http://www.hddaccess.com/intro/faq/#answer3
Also: http://concepthub.org

Q: 1. Who would be the immediate consumers of the framework and system created through this SBIR initiative?

2. Additionally, are there relevant, high-priority projects already within MHS and/or VHA that can be described to the applicants prior to submission of the Phase I application?

A: There are several potential consumers of such a universal healthcare exchange language: the Nationwide Health Information Network, State HIEs, Regional HIEs, Local HIEs, individual medical centers who desire to exchange data internally for various use cases. The work may be of interest to commercial EHR vendors and HIE vendors. Given the recent announcement that the 3M HDD will become open, and potentially be used as a major component of the DOD/VA Joint EHR, the DOD/VA Integrated Program Office would be another likely consumer of the framework.

Q: Should a response to this solicitation focus on the life cycle, workflows, and tools for creation and management of UEL expressions implemented in the context of appropriate use cases, or, alternatively, should the response be centered around developing a minimal UEL framework that illustrates value and jump starts a community of practice around this?

A: The government is flexible to a proposal from either approach.

Q: PCAST makes ambitious recommendations around meta-data representation frameworks, UEL, identity management, encryption and protection, authentication and authorization, index/search/access/retrieval of distributed health information, to granular consent representation and patient centered consent management, context aware visualization and representation of health information. Given the ambitious nature of the report's recommendations, what specific area(s) is this SBIR solicitation targeting?

A: We concur that the PCAST report is ambitious. The government views the topic broadly and will accept proposals for any of the areas outlined in the PCAST report that rely on a universal exchange language. The government recognizes that an SBIR managed by the Army will not solve the complex challenge at hand, but the SBIR does provide an opportunity to jump start research in this area. Of particular concern is how the new 3M Open HDD may be implemented in the DOD/VA Joint EHR under development by the DOD/VA Integrated Program Office (IPO).












--
Michel Dumontier
Associate Professor of Bioinformatics, Carleton University
Chair, W3C Semantic Web for Health Care and the Life Sciences Interest Group
http://dumontierlab.com





--
Michel Dumontier
Associate Professor of Bioinformatics, Carleton University
Chair, W3C Semantic Web for Health Care and the Life Sciences Interest Group
http://dumontierlab.com




--
Michel Dumontier
Associate Professor of Bioinformatics, Carleton University
Chair, W3C Semantic Web for Health Care and the Life Sciences Interest Group
http://dumontierlab.com

Received on Thursday, 21 June 2012 16:30:56 UTC