HL7 delivers healthcare interoperability standards

HL7 May 2010 Working Group Meeting

Around 11 HL7 UK Members went to the Working Group Meeting (WGM), 16-20 May, in Rio de Janeiro, Brazil. You can read their feedback below. For more details about the forthcoming WGM...>>

To view details of HL7 UK members with leadership positions in HL7.org please follow this link.

HL7 UK Summer 2010 eZine includes Working Group Meeting Report from HL7 TSC Chair Charlie McCay

Rene Spronk

Report from the WGM in Rio de Janeiro, May 2010

The WGM in Rio had just over 200 attendees. The number of attendees from South-America was very low, which means that (this time around) the use of a WGM for marketing purposes (to attract local attendees) has basically failed. The same was true for the meeting in Kyoto. HL7 will reexamine the viability, both in terms of productivity as well as revenue, of WGMs outside of North-America. The only ones that have been both productive and break-even from a revenue perspective were Noordwijkerhout (May 2005) and Cologne (May 2006).

The role of an "HL7 Europe" (currently just a postal address in Brussels, see http://www.hl7.eu/) is still unclear. The intent is for it to be the point of contact between the European Union (and its projects: e.g. epSOS, mandate 403) and the HL7 organization. The relationship between the various European affiliates and HL7 Europe is still up for discussion, as is the mandate of HL7 Europe. HL7 Norway became an official affiliate, they finalized the paperwork. (see video at http://www.vimeo.com/12013066). It still has to create TCs, organize meetings etc. - but the paperwork is done.

In the RIMBAA WG (the "HL7 v3 implementers group") the group decided to update its mission statement (see http://wiki.hl7.org/index.php?title=RIMBAA_Mission_and_Charter). The issue of Hl7 version 3 serialization/deserialization was added to its scope; support for the Java SIG software implementation was removed from its mission - RIMBAA will support third parties in their efforts to create HL7 version 3 toolkits. Grahame Grieve (the main author of the Data Types specification) presented the key differences between Datatypes R1 and R2 from an implementers perspective. He urges (amongst other hints and tips) all implementers of Datatypes R1 to pre-adopt the II.scope attribute. GTS.code (using a code to specify timing instead of having to provide fully structured time) also looks like something that projects may want to pre-adopt. Two whitepapers on "code generation" (a process used by nearly all v3 implementers) were regarded as "good to go": http://wiki.hl7.org/index.php?title=MIF_based_code_generation and http://wiki.hl7.org/index.php?title=Schema_based_code_generation  .

The InM WG discussed the HL7 version 3 "Batch Wrapper" ballot. This wrapper is being used in production by three different projects (one of them is AORTA, the Dutch NHIN). Although there were some ballot comments that will lead to changes there were no contentious issues. An updated version of the material will be included in the next ballot round. We expect the material to be sufficiently mature during the ballot cycle and therefore expect it to pass as normative material.

The Marketing WG (a volunteer based group, with its leadership appointed by the HL7 board) has started to work on a Marketing Plan. A draft version is available (see http://www.hl7.org/Library/Committees/marketing/HL7%20Marketing%20Plan%202010%20FINAL.doc), which is more of a communications plan than a marketing plan. A marketing plan speaks to the positioning of HL7's products and HL7's product strategy. There were a number of discussions during the week regarding the overlap in activities between the Education WG (tutorials, educational sessions) and Marketing (ambassador presentations). The Education WG has taken upon itself the task of creating an organization wide Education Plan which includes the work done by Marketing as well as the affiliate organizations. The creation of such plans is fraught with issues related to the internationalization of the organization - they can't be written as if HL7 is one global organization; and they can't mention the affiliates as "competitors" to HL7 International either.

There are ongoing discussions about templates and template based validation. Discussions range from clinicians (how to express clinical models; and the metadata associated with those models), to CDA authors (how to express and test CDA templates, currently those are mostly expressed in textual form), a template registry project (how to store and retrieve templates and their associated metadata), the toolsmiths (how to test for template conformance), to the core methodology group within HL7 (how do we express templates as MIFs, as 'incomplete models'). It all points to the fact that we need templates for validation and conformance testing. Progress is being made, but there's not one single overarching definition yet.

The Patient Administration group had a lot of discussions about Provider Registries. The HL7 v3 standard contains a domain on this subject; the aim is to merge the existing v3 artefacts with new service specifications as defined by SOA/HSSP as well as by an Australian project. The focus is on Provider Registries, whilst acknowledging that in terms of patterns there will be similarities between this registry and other registry types (e.g. Patient, Location).

The Orders and Observations (OO) group discussed a proposal for the creation of a "Order to create a document". (See http://wiki.hl7.org/index.php?title=Order_interaction_for_the_creation_of_a_document). The proposed interactions would allow an Order Placer to "order the creation of a summary document for patient 123". This proposal has now been effectively finalized; the material will be included in the common order domain. [Note that IHE just has created a proposal for a related profile: S3D]. OO also discussed the behavior model (a business-process level activity diagram) for ordering - a somewhat frustrating process given that one should neither include too much detail, nor stay too much at the abstract level. Progress, albeit slowly, is being made in this area.

During the Sunday HL7 UK meeting Richard Kavanagh (NHS) discussed some of the issues they encountered whilst attempting to support the HL7 UK A.2 profile in the existing conformance tools (such as MWB). The HL7 UK profile is a mixture of HL7 v 2.4 and 2.5, with (ehrm..) 'creative' data type extensions. It was suggested that HL7 UK may want to base its new v2 profile on version 2.6. This seemed like a logical thing to do to most attendees; it will need discussion in the UK TSC.

All the regular attendees (the "usual suspects") were present in Rio. Having a small group of attendees has the advantage that one can focus on some core issues and that one has the ability to sort these out with some of the key people; on the other hand it means that some meetings don't have quorum, or that the persons with the expertise required for some discussions aren't present. As such having a meeting like this once a year probably doesn't impact productivity, but a larger audience is required to make progress on all issues on the table. Let's see how we'll do in Cambridge USA..

Australian Delegation Report from May 2010 WGM


Last modified 19/10/10