Ontology Mapper Call Minutes

7/1/08

Present on call:  Rob, Hillari, Brent, Ketty, John, Prakash, Paul, Liat, Marco, Mark, Maggie, Kevin  (Note: In the future we will handle the tasks for John Holmes at the beginning of each call to better accommodate his schedule.)

1)     The University of Texas Medical Center in Houston has joined the Ontology Mapper team and will be present on the call beginning next week.  UT is aware of our grant proposal and they do understand that they would not receive funding at this stage however they would still like to join our project.  Welcome UT!

2)     The new version of I2B2 will be available in beta within the next day or so.  If necessary, UCSF will make that new code available to this team on the Dynamsoft revision control system once available.

3)     We discussed Prakash’s recent achievements and next steps.  Further work on the i2b2 CRC cell and CRC related schema work will be held up until the new release of i2b2 is available.  Until then Prakash will be working on the handling of the aggregates and archetype rule types.

4)     We walked through Ketty’s lifecycle diagram.  This could be used to provide excellent high level structure to the expected “future state” of IDR implementations.

5)     We walked through Marks use case chart and how that would fit into the over-all paper design.  We now can see a complete connection from the highest conceptual level diagrams down to the use case definitions.  This should help us in developing the paper’s structure.  We also talked about the jargon used in the paper and how these use cases might be described regardless of the researchers’ domain of expertise.  Specifically regarding the use case:

1. This use case describes an adverse event associated with diabetes medication.
2. In this example we are tracking fluid retention and its effect on heart failure.  Also insulin usage has been associated with aggressive polyp formation.
3. How could the IDR have been used to explore these down stream ramifications of diabetes meds?
4. In the past to get the data necessary for this sort of study, data imports would need to be created from EMRs, pathology, labs, ambulatory care etc.
5. But in the IDR these data sources will already be available and preloaded.
6. Also prior to the IDR patient identifier information would need to be sent to all of the above mentioned data sources in order to obtain those records.  But in the IDR that initial hypothesis can be tested on a completely de-identified data set leading to a dramatically more secure information infrastructure.
7. With the development of tools based on i2b2 to assist in the testing of an hypothesis, we could test the robustness of a hypothesis based on how well that hypothesis is altered by inclusion and exclusion criteria on the fly .  This would lead to a dramatic decrease in the amount of time required.
8. An attempt will now be made to generate a diagram of abstract principles that can be derived from this use case, which would then be directly associated with Ketty’s high level diagram.

6)     We walked through Maggie’s diagram also and talked about how we would structure the paper to describe the current state of translational science succinctly.

Thanks!

Rob Wynden