All, I'll try to respond to some of the questions from Phillip and Lindsay based on the approach we at KU have taken so far. Of course, this approach may not be optimal (and, I'm not done yet) so I'm eager for input from the group.
Phillip: "What is the current proposal for how the counts should be generated?" - I don't know that we've decided as a network - please correct me if I'm wrong. However, as suggested by Dan<https://informatics.gpcnetwork.org/trac/Project/ticket/145#comment:3>, I've tried to first map the i2b2-style paths for the PCORI CDM to the i2b2 paths we have in our i2b2 instance at KU (HERON). Then from there, I tried to build views that look like the Common Data Model Specification from PCORI<http://listserv.kumc.edu/pipermail/gpc-dev/attachments/20140319/da2b9f05/attachment-0001.docx>. I then generate the counts requested in PCORnet ETL Annotated Data Dictionary V1.0.xlsx<https://informatics.gpcnetwork.org/trac/Project/raw-attachment/ticket/144/PCORnet%20ETL%20Annotated%20Data%20Dictionary%20V1.0.xlsx> by querying against the CDM views. The idea was that we could share much of the code across the network changing the mapping between the i2b2 path for PCORI CDM and local terminology. I've made some progress with this approach (see pcori-annotated-data-dictionary project on BitBucket<https://bitbucket.org/njgraham/pcori-annotated-data-dictionary>) but I'm not done yet. Phillip: "Nathan just sent an email about materializing the PCORI CDM view of the data warehouse. Should we all plan to do this so that we will have a consistent process for generating the counts?" - I ran in to performance issues using the views and decided it would be better to just create CDM tables. So far this is working better. I don't know if we should all take the materialized approach or not - I did it out of necessity. I suspect others may experience similar performance issues using the view approach but I don't know. I'm not sure it matters if we all do exactly the same thing as long as we can support queries against CDM-shaped data. Phillip: "Will we be populating the CDM with all patients, regardless of the two visits requirement? Historic data? Etc?" - So far, I've taken the approach of counting all patients we have in i2b2. As I recall, the "two visit in the past 30 day" constraint was for answering the "enrollment" questions on earlier PCORI characterization questions (since I don't think we didn't have firm guidance from PCORI on what enrollment means for a CDRN). I'm not sure if we should add these constraints or not - the ETL Annotated Data Dictionary spreadsheet says: "The purpose of this form is to provide a mechanism for PCORnet data partners to communicate information about how they transformed their local data into the PCORnet Common Data Model (CDM) Version 1.0." so I was just taking the approach of counting everything we ETL for now. Lindsay: "I think that the Enrollment Table is one of the biggest question marks in my mind. What do we plan on using for this table?" - I haven't done anything yet. In the notes from the gpc-dev call on July 29<http://listserv.kumc.edu/pipermail/gpc-dev/attachments/20140730/98039483/attachment-0001.pdf>, Jim Campbell noted that PCORI hasn't clearly defined it yet and we can bypass until we get better guidance. I don't remember seeing better guidance yet - please let me know if I missed it. From: [email protected] [mailto:[email protected]] On Behalf Of Lindsay Cowell Sent: Tuesday, August 19, 2014 9:21 AM To: Phillip Reeder Cc: Dipti Ranganathan; [email protected] Subject: Re: [gpc-informatics] #144: ETL Annotated Data Dictionary form required by PCORI On Aug 18, 2014, at 10:44 AM, Phillip Reeder <[email protected]<mailto:[email protected]>> wrote: I think that the Enrollment Table is one of the biggest question marks in my mind. What do we plan on using for this table? PCORI is expecting a start date/end date for the enrollment of a patient in the dataset, so the two visits in 3 years definition doesn't seem to applicable. I'm assuming our basis would be algorithmic, since we don't have regional isolation and don't have insurance enrollment data. What algorithm should we apply? How will we store this in i2b2? I think for us at UTSW we would use encounter based basis. (defined in the ETL spreadsheet) Thanks What is the current proposal for how the counts should be generated? Nathan just sent an email about materializing the PCORI CDM view of the data warehouse. Should we all plan to do this so that we will have a consistent process for generating the counts? Will we be populating the CDM with all patients, regardless of the two visits requirement? Historic data? Etc? Phillip From: Dan Connolly <[email protected]<mailto:[email protected]>> Date: Sunday, August 17, 2014 at 1:15 AM To: Lindsay Cowell <[email protected]<mailto:[email protected]>> Cc: Phillip Reeder <[email protected]<mailto:[email protected]>>, "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>>, Dipti Ranganathan <[email protected]<mailto:[email protected]>>, Susan Morrison <[email protected]<mailto:[email protected]>>, "Campbell, James R" <[email protected]<mailto:[email protected]>> Subject: Re: [gpc-informatics] #144: ETL Annotated Data Dictionary form required by PCORI Would you please point us to your earlier questions? Or give a couple examples? Let's try to get as much done between calls as we can. On Aug 16, 2014 3:53 PM, Lindsay Cowell <[email protected]<mailto:[email protected]>> wrote: Hi Jim, The GPC team at UTSW met yesterday to work on completing the ETL Annotated Data Dictionary. We have many of the same questions now as we had when generating counts for the original count table. We would like to request a call during which we attempt to agree across the 10 sites on a standardized way to generate the counts (or as close as possible). Perhaps this can happen during the next two gpc-dev calls? Or perhaps we need a separate time. Thank you, Lindsay On 8/14/14, 10:17 PM, "Campbell, James R" <[email protected]<mailto:[email protected]>> wrote: >Nathan >I think that I have responsibility for this deliverable and so you can >forward Wisconsin response as attachment to the ticket and I will monitor >and collate for the network response to PCORI. >Jim > >James R. Campbell MD >[email protected]<mailto:[email protected]> >Office: 402-559-7505 >Secretary: 402-559-7299 >Pager: 402-888-1230 > >> On Aug 14, 2014, at 2:56 PM, "Wilson Nathan" >> <[email protected]<mailto:[email protected]>> >>wrote: >> >> Where or to whom should I send the Wisconsin data dictionary? >> >> Nathan Wilson >> >> -----Original Message----- >> From: >> [email protected]<mailto:[email protected]> >>[mailto:[email protected]] On Behalf Of GPC Informatics >> Sent: Thursday, August 07, 2014 1:17 PM >> To: [email protected]<mailto:[email protected]>; >> [email protected]<mailto:[email protected]>; >> [email protected]<mailto:[email protected]> >> Cc: [email protected]<mailto:[email protected]> >> Subject: Re: [gpc-informatics] #144: ETL Annotated Data Dictionary form >>required by PCORI >> >> #144: ETL Annotated Data Dictionary form required by PCORI >> ---------------------------------+------------------------ >> Reporter: campbell | Owner: ngraham >> Type: task | Status: assigned >> Priority: major | Milestone: popmednet >> Component: data-stds | Resolution: >> Keywords: ETL Data Dictionary | Blocked By: 145, 146 >> Blocking: | >> ---------------------------------+------------------------ >> >> Comment (by ngraham): >> >> I worked on adding the [https://bitbucket.org/njgraham/pcori-annotated- >> data-dictionary/commits/eb6e04281dfd9fd92758da1d51c7baf1d38e8e2e >>diagnosis CDM view] but ran in to performance issues - see >>ticket:146#comment:5. I plan to spend a little time trying to tune the >>query. >> >> -- >> Ticket URL: >><http://informatics.gpcnetwork.org/trac/Project/ticket/144#comment:9> >> gpc-informatics <http://informatics.gpcnetwork.org/> >> Greater Plains Network - Informatics >> _______________________________________________ >> Gpc-dev mailing list >> [email protected]<mailto:[email protected]> >> http://listserv.kumc.edu/mailman/listinfo/gpc-dev > >The information in this e-mail may be privileged and confidential, >intended only for the use of the addressee(s) above. Any unauthorized use >or disclosure of this information is prohibited. If you have received >this e-mail by mistake, please delete it and immediately contact the >sender. > >_______________________________________________ >Gpc-dev mailing list >[email protected]<mailto:[email protected]> >http://listserv.kumc.edu/mailman/listinfo/gpc-dev ________________________________ UT Southwestern Medical Center The future of medicine, today. _______________________________________________ Gpc-dev mailing list [email protected]<mailto:[email protected]> http://listserv.kumc.edu/mailman/listinfo/gpc-dev
_______________________________________________ Gpc-dev mailing list [email protected] http://listserv.kumc.edu/mailman/listinfo/gpc-dev
