August 20, 2015
9 MDT/10 CDT/11 EDT/16 BST
Attending: Matthew Cownie, Tom Creighton, Erick Emde, Jason Haag, Lucas Huang, Rosalyn Scott, Valerie Smothers, Michael Steele, David Topps, and Luke Woodham.
1 Review minutes
The minutes were accepted as submitted.
2 Note page of links: http://groups.medbiq.org/medbiq/display/XIG/Important+links
Valerie mentioned the links to external documents are found on the wiki page called important links. She encouraged the group to access, view and make comments. If anything is missing, please feel free to let us know.
3 Development updates from working group members
David thanked Ellen and Andrew for their contributions and refinements of the document. There wasn’t a lot of debate around most verbs. Opened, Terminated, and Experienced were topics of discussion. Jason commented that the CMI-5 group uses initialized, and the group agreed that initialized is preferable to opened. David agreed to make the change.
Matt asked what happens if you return to a case. David answered the verb would be resumed. David discussed Terminated. In open labyrinth, when a player leaves a case (stops), you can’t tell if they have left it or have been cut off. Matt asked about the ability to restart a case. David noted you can’t resume unless they bookmarked it, (suspended it). Valerie recommended better aligning the definition with that developed by CMI-5. Jason is currently working with a vocabulary working group taking terms from ADL and agreed to work with David.
The group discussed Experienced. He explained that learners may touch on a node but never see the content due to being redirected to another node. Valerie asked if a new verb be created for that so that we can maintain semantic consistency. Jason agreed and noted the need for a term similar to accessed. He added that we would need to develop IRIs for any new verbs. David recommended visited.
Luke asked if a learner looking back at a previous node would use the verb Experienced. He added that a pair of statement with 2 verbs could work – selected and visited, or visited and experienced. Valerie commented that experienced is narrower than visited. Jason recommended tweaking the definition of launched and using that. If they go to first node visited and can’t go there, they are launched to a different node and back to start. Valerie asked if you could launch a single page. Tom answered yes. They can add more meaning and build on to launch. If it doesn’t fit, you could add your own. David agreed they could tweak definition of launch.
Valerie asked if ignored was already in activity streams. David confirmed. Jason advised going forward to not use terms from activity streams, because they don’t know intended usage. He suggested finding definition sources for read, appended, ignored so they are not activity stream. David commented that several verbs in the Virtual Patient recipe come from activity streams. Jason will work with David on adding further metadata for those verbs.
David provided other updates on grant applications. He is reasonably confident on receiving the XAPI grant with Open labyrinth. They are going after a Canary (CANARIE?) grant, setting up the infrastructure for big data analytics beyond virtual patient use. They won’t know until February 2016. Funds from the CANARIE grant must be spent in Canada.
4 Discuss Gap analysis from Tom Creighton
Tom explained that the gap analysis is in support of the VA simulation and reporting analysis project, which is based on XAPI in conjunction with competency framework and performance framework standards. Valerie added Tom has thought through what work we have and haven’t done and what to consider when developing recipes for healthcare education. Tom explained that the VA project focuses on code simulation, Currently there is manual tracking of performance on paper. There is an opportunity to do that electronically and tie to other simulation data. Erick asked if there was a high level document about xAPI for those not familiar with it. They need to track outbreak training and would benefit from implementing xAPI. David asked what kind of information needs to be defined and if simulation companies are prepared to incorporate XAPI into their own from devices. Valerie replied this was a step towards getting there. We want to have guidance on simulation reporting.
Valerie commented that Tom’s gap analysis points to using IRI’s, internationalized resource identifiers, for activities and verbs. The list of things that would need IRI’s is located on page two. The recommendation is to use persistent URL’s (PURLS) to keep IRIs from changing. ADL is in the process of creating/registering PURLS for XAPI IRIs. Jason is the person doing this. Valerie asked what the process would look like if we wanted to have id for a new verb. Jason mentioned they would set up sub-directory, purl.org/xapi/medbiq, and make Valerie administrator of that URL for verbs, attachments or extensions. Valerie noted she has purl account but doesn’t have administrator access. Erick asked Jason if there is a government equivalent of purl.org. Jason answered not to his knowledge. Valerie asked what NLM uses for their Biomed central versions of pubmed (note – PubMedCommons references the doi and the PMCID).
Jason commented that the Actor format defines various ways to identify an Actor, such as through a learner’s email, OpenID or a user account. Use of a consistent Actor format will make it easier for systems to identify people. The recommendation is to use Actor Account format that allows the organization to use values with no PII, as opposed to an email or OpenID.
5 Open discussion
- Initialized will be used to indicate when a Virtual Patient is started (as opposed to Opened).
- The verb Launched will be used to indicate when a learner has been redirected to a different node based on a trigger.
- Jason will work with David on adding metadata to verbs originally from activity streams.
- Valerie will work with Jason to set up a MedBiquitous presence in the purl.org domain.