Thanks for taking notes. I prefer to avoid proprietary formats, so I hope everybody doesn't mind another copy...
Meeting Notes summary GPC –DEV 02/04/13 Attendees: Aaron Miller, Ashok Mudgapali, Bonnie Westra (UMN), Dan Connelly (DC), Debbie Yoshihara (UW-Madison), Finamore, Joe D, Geroge Kowalski, Glenn Bushee, Hubert Hickman (UNMC), James Campbell (JC), Justin Dale (UMN), Supreet Kathpalia (UMN), Keith Wanta, Nadkarni Prakash, Nathan Graham, Nathan Wilson, Phillip Reeder (UTSoutwestern), Russ Waitman (RW), Simon Lin, Tom Mish (UW- Madison). 1. Accept Meeting notes from 01-28-2014: DC: Second the meeting notes, 02/11 DC is OOO and hence the meeting next week is on Thursday, 02/13/14. Scribe: from UW Madison 2. Plan next meeting: On 02/18/13 – DC is unavailable to chair the meeting all week hence we could cancel the meeting or someone else can chair the meeting. It would be the last meeting before the Hackathon. The plan was to schedule it in case of last minute organizational issues for the Hackathon. 3. Connecting with PCORI on PopMedNet: Update from Aaron: Aaron connected with PCORI and early March go-live, PCORI to provide more information at that time. There were questions around where the nodes should be and RW is inclined to have the node set up at KUMC. Per Russ, we don’t need everyone to implement Pop Med Net. We would just need one node for the network. 4. Hackathon Agenda: Available on the wiki site. Someone from Madison had questions on it. Per DC there are 2 hours for data standardization on first day. There is a possibility on the second day we might be able to write some code for babel. JC may be sending out some homework in terms of preparation for the Hackathon. Regarding the PI agenda, RW reported the plan was to be together from 9:00- 11:00am to discuss data standardization. There are other topics such as patient engagement, areas of challenge, etc. that would be scheduled based on Lauren’s availability and then after lunch on the first day. Get everyone together at the end of the first day. The entire team would be together at 11:30. The PI discussion breakout sessions to be held based on Lauren and Rick’s availability. Also clarified that it is a day and a half and people can leave on Tuesday at 11:00am. However, if others are staying later then we could meet for lunch. 5. i2b2 Demo by UTSW (Phillip): Key data elements discussed were: Diagnoses: have modifiers such as billing principal diagnoses, billing secondary diagnoses, encounter diagnosis, and problem list diagnosis. Beacon cancer information: Several modifiers. Clinical staging depends on what site you are looking at such as breast cancer and lung cancer may vary. The descriptions also vary versus C1 tumor and C2 tumor. All this is built off the actual epic terminology. Specimen- cell, fluid, molecular and tissue. Tumor registry- full, tumor registry-abridged Pathology reports include the following: · Co path: it is sun quest co path. There are lots of issues of co path into epic. It looses information such as header that is lost so don’t know where the information is coming from. So actually went into the co path. · Cytogenetic, ER · Flow data: marker panel, specimen type was run and the various tests. · Image files from the pathology system · Heme data Minimal Data Set: Per RW there was a researcher who had an excel spreadsheet and brought it in red cap and then into i2b2. Per Phillip most other things are standardized such as medications, allergies/ drug class. Medications include current medication and ordered medications. Cancer gene and Risk profile: A risk profile suggests the probability of cancer. It suggests if the patient had a greater than 20% chance of cancer. This is an outside data set brought in and integrated with Epic. Reports: Visit notes a) note concept b) note types 6. I2b2 demo by KUMC (DC and RW): i2b2 access and metrics: Per DC i2b2 at KUMC is integrated with the enterprise directory system. There is some governance stuff rolled into this. There are 3 criteria to access heron, i.e. KUMC faculty or sponsored, complete human subjects training and signed system access agreement. Russ and others are big on metrics. DC demonstrated queries by user that report the usage, volume by month for researchers and other 30-40-50 customers. They have also started to look at real time performance such as there are some queries that run for a long time and some that take 2 seconds. Training videos…there is oversight committee. Usage audit report if anyone is looking at queries that don’t relate to his research. Use red cap for workflow. You see a summary of the data each time you log in to Heron that you need to acknowledge prior to logging in. Per DC KUMC integrates data from red cap projects. Some features don’t work if you are in red cap mode. I2b2 demonstration focused on the following data elements: Allergy Cancer cases: the terminology follows the NAACCR guidelines SEER site summary: Ability to get a patient list and show a statistical integration. Demonstrated 5year cancer survival using the Analysis tools. Breast cancer cases- group 1 and black in group 2- roughly 900 patients. On good days you get a data script that can run into R. Same thing with our data requests as well. Used it for demographics that can be used for recruiting. Zip codes: here are HIPAA issues around reporting zip codes. Vital status: KUMC integrates data from the SS administration. Diagnosis: comes from UMLS since the one from i2b2 is incomplete. Flow sheets: did an AMIA paper on this. There is the model that Epic gives you however the nurses enter information in varied ways. Per RW in KU IT there is a grouper/ template and then the concept of temperature and blood pressure is still from Epic. It would be interesting to see how the model or foundation is similar for other sites. There would be Apgar scores where everyone as done their custom stuff. Labs- no LOINC this is whatever our lab system does. Modifiers: last lab result and median lab results Medications- included dispensed, inpatient, and historical medications, etc. Microbiology Procedure orders: standard CPT stuff. Per RW one thing is remarkable we integrate data from clinics from 1980 and for hospitals form 2007. Visit Note: Note concepts Biospeciman Repository: there is a spreadsheet but not a whole lot of stuff. UHC data: Per RW there is a great amount of data for folks who do inpatient research. Example: UHC visit details. We get the files from UHC and then stage the files and add them to i2b2. 7. Epic User group session: by Jim Campbell and James McClay. 2014 Epic User Group meeting session<http://www.mail-archive.com/[email protected]/msg00076.html>. Per RW It would be good to get feedback from the research user group at Epic. 8. Feedback on i2b2 data model: Jay Fuehrer sent some feedback that did not get forwarded to the group. Please look at the wiki site. 9. Data Standards discussion: JC reported he has done a couple of calls w/Epic and one of the issues he has been pursuing is how we can standardize and collapse the flow sheet data. Epic is actually building a data standards committee for the researchers. JC also discussed the NDF_RT ontology. Marshfield and other sites use First Data Bank (FDB). We need to think about the FDB ontology such as therapeutic class, etc. Per RW FDB is proprietary however we can talk about it more later on. ________________________________ From: Greater Plains Collaborative Software Development [[email protected]] on behalf of Supreet Kathpalia [[email protected]] Sent: Wednesday, February 05, 2014 2:45 PM To: [email protected] Subject: GPC-DEV Meeting minutes 020514 Please see attached meeting minutes. We did cover a lot of information. Please let me know if I missed anything or any edits needed. Thanks, -- Regards, Supreet Kathpalia Clinical Data Analyst Research Development & Support AHC Information Systems, U of MN Ph. 612.624.5848 [email protected]<mailto:[email protected]>
