F2F Minutes and Presentations

From IHE Wiki
Revision as of 09:48, 19 March 2013 by Eokoroi (talk | contribs) (→‎Minutes)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

IHE Lab Domain Face to Face Meeting Minutes – February 4-6, 2013


Ed Heierman, Abbott; Haridas Puthiyapurayil, Abbott; Riki Merrick, APHL; Christel Daniel, AP-HP; François Macary, ASIP Santé; Laurent Lardin, bioMerieux; Raj Dash MD, Duke University; Tim Polsinelli, Epic; Harry Solomon, GE; Hiroyoshi Okada, IHE Japan, ITEC Hankyu-Hamshin; Genichi Kato MD, JAHIS; Takashi Nakashima, JAHIS, Sakura fine tek Japan; Anna Orlova, Johns Hopkins; Anne-Gaëlle Bergé, Kereval / IHE Europe technical team; Filip Migom, MIPS; Colin Murphy, mTuitive; John Murphy, mTuitive; Jim Harrison MD PhD, University of Virginia, Kay Washington MD, Vanderbilt University; Andrea Pitkus, CAP; Beth Chmara, CAP; Carolyn Knapik, CAP; Edie Okoroigwe, CAP; Jeffery Karp, CAP; Jim Dvorak, CAP; Mark Whitsitt, CAP; Mary Kennedy, CAP; Maryrose Murphy, CAP; Rich Moldwin, CAP; Sam Spencer, CAP; Ted Carithers, CAP

Meeting Organization

Location: The College of American Pathologists – Deerfield, IL offices
Dates: February 4, 5 & 6, 2013
Host: Mary Kennedy – The College of American Pathologists
Chair: François Macary
Secretary: Mark Whitsitt

Presentations are available on the IHE ftp site, and accessible from this wiki page: http://wiki.ihe.net/index.php?title=F2F_Meeting_Presentations


  1. Greg Gleason, CIO and VP CAP Information Systems welcomed attendees and made introductory remarks, noting IHE activity is important to the CAP and it is happy to support the work.
  2. Joint meeting with AP domain to review microbiology and to harmonize AP and Lab with respect to genomic issues. Discussion included:
      • Need to clarify the message structure and vocabularies, as well as check consistency with other implementation guides, e.g., LRI (MU2) and Lab Results to Public Health.
      • Need to refine Lab Technical Framework volumes 2 & 3.
      • There are two organizing objects, the specimen and the isolate. This structure is specimen centric. The isolate is separate from the specimen as opposed to being in the specimen.
      • It is not always the case that tests on an isolate is a “reflex order” because isolate may not have been identified yet. Assumptions may not be true in all LAW situations – need to evaluate more situations to address issues with implementation in middleware.
      • Recapitulative mode: new observations to be reported with all previous observations. This is a global requirement for IHE LAB, and also for US Meaningful Use.
      • Most countries do not yet have SNOMED CT due to costs and license restrictions, so IHE cannot require use of SNOMED CT at an international level.
      • Microorganisms may be identified using one of these coded terminologies:
        • SNOMED CT for the countries members of IHTSDO.
        • JANIS for Japan
        • SNOMED 3.5 VF for France.
        • …others to be investigated.
      • Committee is comfortable with the OUL message structure, but Belgium protocol representatives have expressed the desire for an XML structure.
      • OBX-5 contains the result/measurement (e.g.; the MIC measured). OBX-8 contains the interpretation (e.g.; “S”, “I”, “R”, …). Reporting to Public Health allows either OBX-5 or OBX-8. Under OBX-8, HL7 is working to make table 0078 an HL7 defined table instead of user defined. Results can be complete with an empty OBX-5 value, however, if OBX-11 indicates the result is complete, one of OBX-5 or OBX-8 must be filled in.
  3. Filip Migom led a discussion on the usage of coded vocabularies for IHE LAB and the translation of working vocabularies to an international standard including LOINC, SNOMED CT, SNOMED 3.5, HL7 vocabularies, ICD-10, …
      • Assignment of codes is required at all steps starting with ordering, however, there is no agreed upon international terminology for test ordering.
      • The type of observations (OBX-3) is coded with LOINC, except in Japan where JLAC 10 is used.
      • Qualitative results can be coded in OBX-5 using the appropriate vocabulary (SNOMED CT, SNOMED 3.5, JANIS, …
      • IHE is seeking international cooperation to determine the best approach to pre-coordination and application configuration to federate terminologies.
  4. IHTSDO licensing – Steven Castanien, CAP Consulting
      • Any use of SNOMED CT requires at least a basic affiliate license from IHTSDO, either on a national basis via an IHTSDO member country’s release center, or purchased directly by a user outside of the IHTSDO membership, using an “affiliate license”.
      • Most member countries have guidelines for how SNOMED CT can be used. Vendors in member countries can sell in their country, but selling across national boundaries can be complex due to some vagueness in the licensing agreement.
      • There is an agreement between HL7 and IHTSDO where SNOMED CT codes and descriptions are available in certain HL7 products, however use of SNOMED CT in an HL7 product does not give vendor permission to use SNOMED CT in its entirety if not licensed through IHTSDO. Interoperability is enabled, but not the usage of full SNOMED CT functionality.
      • Once data is encoded with SNOMED CT, there are no restrictions on display and viewing of the data, but one cannot use that data to search the SNOMED hierarchy and construct new documents, without a license.
      • Vendors may obtain affiliate licenses for SNOMED CT and then sublicense to customers to the extent SNOMED is a part of the product, but this does not constitute a general-use license for full access to SNOMED CT if in a non-member country.
      • Annual fees for member countries are determined based on the country’s economy (GDP); fees for vendors vary based on anticipated usage and size of company.
  5. LOINC Challenges: CAP’s Experience – Andrea Pitkus, PhD – slides available on ftp://ftp.ihe.net. Discussion included:
      • It was noted that the vendor incentive to meet US Meaningful Use is purely a business decision.
      • Codes can be used for an Order, Observation, or Both, but the order encoding needs to be consistent with the expected observation code as determined by the orderer. Selection of codes by specific methodologies depends on expected results and the analyses to be performed.
      • Maintenance and versioning of LOINC (identification of deprecated and discouraged codes) is important to US Meaningful Use attestation.
      • Management of updated codes in historical data versus new data presents issues of synonyms and equivalents.
  6. LAW profile change proposal – Externally stored results - Harry Solomon (GE)
      • Some of the results by the analyzers do not fit very well in the “observation result” field of the database of middleware, LIS, EMR systems (e.g.; whole slide images, genetic sequencing, hematology plot diagrams, …). These results are conveniently placed in an external storage, accessible to the systems who need to consume them. A mechanism by which analyzers (or middlewares) can store data and pointers to that store can be passed between activities. – See slides for details.
      • To make this mechanism effective, usage of field OBR-49 was proposed: A new value “RE” indicates that a result of this order is stored in a persistent data repository, the link being provided in the OBX.
      • The group agreed this is a reasonable proposal to pursue.
  7. LAW profile change proposal – Observation identification – Ed Heierman (Abbott)
      • The representation of observation hierarchy encompassing the notion of analysis runs under a single OBR has several types of associated data. Although a unique AWOS is requested, multiple observations may be returned.
      • Alternative observations (alternate representations of the same analysis/result) can be sent as strings or numeric arrays, but the nature of the string or array cannot be determined from the results message.
      • Two alternatives were proposed:
        • Indication of data type in OBX-4 with string identifier
        • Indication of data type as ED in OBX-5.
      • Each vendor must assign definitions to alternative representation codes a priori. A way to clearly and immediately indicate that alternative replicates of data are option and may not need to be managed by the analyzer manager is required. Suggest:
        • If first triplet is empty, this is extra data that may be ignored.
        • Vendors to make this decision and not try to pre-define code sets.
        • Can limit OBX-3 (data type) options, but up to vendor to select a code from the list.
      • A case where an order for “albumin” is requested and the analyzer returns either/both “ALBP” and/or “ALBG” observations was discussed.
      • Mr. Heierman will organize the proposed changes and initiate the appropriate dialog; the proposal will be published on the wiki and on the ftp server in a dedicated folder.
      • A task force to stabilize proposals prior to publication consisting of Filip Migom, Laurent Lardin, Anne-Gaëlle Bergé, Hari Puthiyapurayil, Ed Heierman, and François Macary was nominated and will discuss via email and possibly teleconference.
  8. Gazelle testing environment overview – Anne-Gaëlle Bergé (Kereval/IHE Tehnical team) – See slides at ftp://ftp.ihe.net.
  9. Foundation for LAW validation and certification – Ed Heierman (Abbott)
      • Gazelle is becoming the standard tool for IHE certification.
      • It is unclear what the strategy for certifying LAW profile implementation will be in the future, and a collaboration between Gazelle and ICSA is suggested.
      • A traditional Connectathon would be used for certification and then separate testing could be done.
      • The goal from the analyzer manager perspective is to have a single driver for new instruments implementing LAW.
      • A potential project between IICC and IHE Services (Europe) may move this plan forward.
  10. Ancillary techniques and molecular biology reporting – Christel Daniel (ASIP santé) – see slides at ftp://ftp.ihe.net
      • A set of examples of how these techniques are reported in cancer checklists was presented.
      • Either a generalized approach or precise reporting is possible. Development of a common strategy and rules upon which to standardize is sought.
      • Parsimony of the important report information is still not included, i.e. values, reference ranges, conclusions, etc.
      • A workflow for sending specimens to bio-banking, send-out orders, etc. is also needed.
      • It will be important to differentiate between the clinically actionable data and the ancillary data required to document the decision rationale.
  11. Japanese IHE Activities – Takashi Nakashima (JAHIS) – see slides at ftp://ftp.ihe.net
      • Mr. Nakashima presented a report on Connectathon activity in Japan, where the Lab Domain is very active.
      • JLAC10, maintained by the Japanese Society for Laboratory Medicine, is a laboratory test code master list with multi-axis codes having 5 components.
      • JLAC10 is used in Japan because it is a government and professional society decision made more than 20 year ago, and there is currently no apparent reason to convert to LOINC.
      • JLAC 10 codes are required to pass a Japanese Connectathon, but when the analyzer is in a hospital, local or internal codes are typically used.
      • Japanese manufacturers are likely to be interested in increasing LOINC use to make their products more successful in the global marketplace.
  12. LCC profile development – James Harrison, MD Phd (Univ of Virginia)
      • Dr. Harrison reviewed the motivation and goals of the LCC profile development work. See slides on http://wiki.ihe.net/index.php?title=F2F_Meeting_Presentations.
      • Profile proposes a mechanism for order filler to send order change recommendations to order placer, adding a new proposed code describing “reason for recommendation”.
      • A number of changes would be necessary and are summarized in the slides.
      • A potential change to a TQ (timing/quantity) code may be necessary to introduce a recommendation time-out value as orders are placed in hold status while recommendation is reviewed, and cannot remain on hold indefinitely.
      • An assumption is made that the placer has initiated generation of a PID for an order so that a testing recommendation may be made.
      • The order modification transaction may now be completed and presented to HL7 O&O for their feedback.
  13. Microbiology reporting using CDA – see slides on ftp://ftp.ihe.net
      • Current HL7 structure was reviewed and a model of how this would be represented in CDA was discussed.
      • A Microbiology specialty section was proposed using LOINC code 18725-2 with one report item sub-section per specimen (specimen-centric), with both human readable test and machine readable content (structured data). The template is a set of constraints applied to the CDA/XML standard.
      • CDA segments would be properly associated by inclusion of appropriate codes in segments that refer to the same object.
  14. Presentation on MDHT tool
  15. Next steps:
      • Preparation of CPs to LAB TF vols 2 & 3 (corrections of typos) François
      • Preparation of CPs for LAW supplement Ed, + the task force dedicated to finalization of these CPs.
      • All CPs are passed to Sondra for registration and ballot management.
      • Guidance on how to structure microbiology results: A work group in France (involving Laurent, Filip and François) works on this subject, and will translate its deliverable into English for the LAB committee.
  16. Work plan for 2013:
      • Review IHE Lab wiki page
      • Continue development of LCC and review of LAW (joint with IICC). New LAW release will be published in July 2013. (LAW is expected to be tested at the 2014 North American Connectathon; Europe in May 2014; possibly in Japan in October 2014)
  17. Wrap up
      • Next face to face meeting will be in Tokyo, Japan. Three options of proposed dates are:
        • September 4-6, 2013
        • October 1-3, 2013
        • October 15-17, 2013
        • The choice will be made based on the results of a poll to be posted to the group.