November 19, 2013
1 PM EDT/6 PM GMT/7 PM CET
Call in Number
Please note: the conferencing service will ask you to enter the pound sign. Press # for pound.
To mute, press *6.
Attending: JB McGee and Nabil Zary, co-chairs; Susan Albright, Matt Cownie, Andrzej Kononowicz, Michael Steele, Valerie Smothers
1 Review minutes of last meeting
JB commented he is supportive of the direction things were going: looking more broadly at the definition of virtual patient and at the use cases. More focus on activity surrounding the virtual patient and possibly a new standard. He asked Nabil if that was consistent with the direction of the last meeting.
Nabil replied that the last two meetings were an exploratory phase. It was a brainstorming discussion about possible directions. one was idea of virtual patient as an activity; that could be experience API, deciding what is educationally sound to track. Susan was looking at was was going on at AMEE.
JB commented he wanted Ankur Doshi, a faculty member at Pitt and user of virtual patients to comment. he is not able to be with us today, we will make sure he is on the next call. We need to hear form those focusing on th educational uses of VPs. hopefully we can outline some concrete steps we can take.
Susan commented that the GIR survey supposedly has gone out this week, but she has not seen it. It was a 10 questions survey, to be back by next week. JB will ask his colleague who is on GIR. (See: https://www.surveymonkey.com/s/VirtualPatientSurvey)
JB asked how the survey came about. Susan replied it was her suggestion but for the benefit of the working group to help us make decisions. We should have data by next meeting if people respond.
The minutes were approved.
2 Discuss use cases
JB commented use cases would be appropriate end result of our explorations.
a. Ankur’s perspective on types of VBPs and what is of value to educators (Ankur)
b. Matt’s use cases
He did a quick brain storm and thought up some use cases for different contexts. These are cases they have come into trying to use VPs with partners. the first two are mirrors of each other. First, clinical staff want to share a case. they want students to learn form it. But the trust don't want to collect those results. XAPI could post results to a system. You may have different expectations depending on the level of the learner. Second, educational institution wants to share with clinical practice. trust might be interested but not care how it was played, where as institution does care. two audiences - HR department and author/educator.Third, authors have created cases that do not use the standard but would like to capture results data. You would not have to adopt the whole standard. Fourth, do a VP, physical simulation, then do the VP again. HSVO has run into difficulties. He went to a talk with a guy form ARM in Cambridge. They are into the "Internet of things," sensors everywhere. that would provide a deluge of data. You could measure a learner's heart rate while they do a case. St George's has people play cases collaboratively. How might you do that online. Many options.; connect different streams of activity and synchronize them again. Six, could you link where a learner is in a game. Seven, keeps popping up, playing a case over an extended period of time. use reporting mechanism to link incidents to same patient. in their nursing curriculum, they see the same patient in years 1-3. the reporting API could link those together. Last, getting into the big data thing, everything a learner might do. You could see what patterns emerge. Links were to map words we use with the ADL profile. Original suggestion - could we use the ADL profile, or do we need a virtual patient community profile. They are not too far apart.
JB thanked Matt for his efforts. It's a great starting point. Does the Experience API cover close enough to what we need, or do we need to work on our own version? The other bog question, what else would we do to make sure we have all the use cases our community needs.
Matt commented that the verbs are fairly close. there is an advantage staying as close as possible. the whole thing is written for communities of practice to have profiles. Some of it is a case of wading through the spec. he is trying to do it; that shows where the holes are. Activities may be a richer place to think about. Not just in a virtual patient, but for other kinds of simulation activities. They are having good results with 3D and virtual reality. How do you fit that with a virtual patient. If we had a structure we could add to, that would be ideal, for instance, adding an activity.
Nabil commented this is very good work. He summarized where we are. What we have is a standard that describes the activity within the virtual patient, the virtual patient itself, and packaging the data. It's for exchanging the VP as a package. If we look as the virtual patient as an activity, we add two layers. One is the experience around the VP. The VP becomes a black box. The next layer is the educational aspects. You want to make sense of the data in terms of competency, etc. Are we discussing a way to standardize experience around the virtual patient, or has he missed something.
JB replied a lot of this is up in the air for the group to decide. He is thinking about whether we need to change the existing standard or work with the XAPI standard and not worry about the content. Or we could track activity in a virtual patient. Matt commented we could construct a profile - two files. Almost an appendum or guidance. JB commented that would require work on our part. Most people care about what a student does with a whole series of activities, not just virtual patients. Matt commented that anything that lets them integrate with other people's systems is a win.
Nabil commented that if we do this piece of work, people that have not adopted the VP standard, they don't need to worry about it. they will just go for this one; the value is in exchanging the data about experience. This is a new type of adopters that would just go for the API. Matt and JB agreed. Matt commented they may need to look at the original standard to better understand the concepts.
JB asked if anyone else had experience with the Experience API. JB commented it would be good to get someone else on the working group working with it. He would like to know more about what the community is doing. Valerie commented University of Michigan had been working with it. She offered to contact them.
JB added Susan's survey may spawn some use cases. Susan commented she could report on how Tufts is using it She has to ask Ethan who is receiving the survey.
Nabil added that XAPI is a good thing. in the context of the working group, we can become implementers. To increase the value, we can use it as glue between the standard. Is this the type of work the working group would like to engage in.
Valerie commented we could develop guidelines for virtual patient implementers or a profile. But she can't say which is appropriate.
JB commented that looking at creating a formal profile would be in our area of interest. The question is if we extend it outside of what we call virtual patients. extending to healthcare and patients would be more interesting. Matt added that OSCE's would be good. JB commented that when you bring those different types of data together is what is valuable. They are working on data warehousing and answering multiple questions. that becomes worth the trouble.
Nabil commented this work would have very high value. JB added that multiple sources of data may be necessary to determine competency.
c. Other ideas
3. Priorities for the group
4. Next steps
JB summarized we could extend the XAPI, allowing the community of educators to get the relevant data around a simulation activity. Or do we need to create a new standard form scratch. A third option, extend the existing virtual patient standard in a way that feeds into the XAPI standard.
Susan commented that we need to build on Matt's work before deciding the answer. Matt commented he needs to finish mapping what he's got, taking the data, and putting it into an LRS, and seeing if it makes sense.
JB commented it will be up to the vendors to create usable tools. It is extremely hard to get actionable insight out of your data. The software is very expensive for commercial data warehousing software. JB offered to research the XAPI community further. If you define which decisions relate to competency, mapping those to student actions reveals answers clearly.
Michael commented that in a study with anesthesia, it is a challenge to correlate, but you can see exactly what the user's experience was and start to understand why people didn;t perform well. If you look at Khan academy in simple quizzing constructs, they allow the educator to figure out where to intervene. The educator can see who needs a review. That is hopefully where we will go.
JB commented that was nice use case. JB asked if we want to restrict ourselves to virtual patients or branch to other types of simulation and education. Michael commented we should include mannequin based as well as virtual simulations. It would be interesting to that community as well. Valerie commented it would be a great opportunity to get others involved.
JB added there was an effort with the ABMS. Not sure what end result was. If we have energy and skills to expand and we pull in the right experts. it will be valuable to the educational community.
Nabil added that the approach of contributing use cases, creating a proposal, and adding new participants was a good approach. There are no clear borders anymore.
- Valerie will contact University of Michigan contacts re API use
- Matt will put data into a Learning Record Store and provide feedback to the group.
- Susan will provide the results of the GIR virtual patient survey.
- JB will look into the XAPI community and open source solutions.
- Working group members will review existing use cases and add additional use cases.