I thought this comment on modifiers from Shawn worth sharing Jim
________________________________ From: Murphy, Shawn N. [[email protected]] Sent: Monday, April 28, 2014 5:15 PM To: Campbell, James R Cc: Mandl, Kenneth; Russ Waitman; Connolly Dan Subject: RE: i2b2 ontology work based on ONC standards...... Hi Jim, There are no explicit constraints, but there is this guidance to using modifiers: Guiding principles to using modifiers In order to represent data and respond to queries in ways that researchers expect, we advocate the following use of modifiers: 1. The modifier does not fundamentally alter the concept. For example, it would be inappropriate for the modifier to be a negation (“does not have”) or a designation that it applies to the family history (“has family history of”). These should be placed in their own top level folders. This way the creator of a query can drag the term without modifiers and get an overall representation of the concept without regard to modifiers. It is also necessary for proper backward compatibility. 2. Modifiers cannot be used for entries in the dimension tables. For a universal applicability of modifiers, all patient and encounter variables should be places in the fact table. 3. Modified facts have values that are pertinent to the modifier. For example, the “dose” modifier on a medication will have the dose amount as the value. 4. Although not technically incorrect, if a column already exists as a modifier (like “provider”) it should not be made a modifier as well. Thanks, Shawn. From: Campbell, James R [mailto:[email protected]] Sent: Monday, April 28, 2014 9:16 AM To: Murphy, Shawn N. Cc: Mandl, Kenneth; Russ Waitman; Connolly Dan Subject: RE: i2b2 ontology work based on ONC standards...... Thanks for responding Shawn. Note that the analysis I sent out yesterday maintains attribution of EHR source material for data domains. This seems to be implicit in the PCORI principles and will be required for CDM V1 compliance. Hence, for diagnoses we are extracting from the encounter, the problem list, PMH and possibly orders. For medications we have administration events, inpatient orders, prescriptions and pharmacy dispense events. We are proposing to maintain distinction between those subclasses using modifiers. Do you have explicit constraints in the elements of the domain you extract for SHRINE or has this been a matter of interpretation across each site that has implemented? Kenneth Any comments about the data domains and ontologies to be supported for SCILHS? Jim ________________________________ From: Murphy, Shawn N. [[email protected]] Sent: Monday, April 28, 2014 7:54 AM To: Campbell, James R Subject: RE: i2b2 ontology work based on ONC standards...... Yes for the Harvard SHRINE, not sure yet for SCILHS. Thanks, Shawn. From: Campbell, James R [mailto:[email protected]] Sent: Monday, April 28, 2014 7:48 AM To: Murphy, Shawn N. Subject: RE: i2b2 ontology work based on ONC standards...... Thanks Shawn. Is your network still running with with the four main data domains? Jim ________________________________ From: Murphy, Shawn N. [[email protected]] Sent: Sunday, April 27, 2014 8:34 PM To: Campbell, James R; Kahn, Michael; Russ Waitman; Phillips, Lori C. Subject: RE: i2b2 ontology work based on ONC standards...... Hi Jim, It’s in this paper: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0055811 Thanks, Shawn. From: Campbell, James R [mailto:[email protected]] Sent: Sunday, April 27, 2014 5:05 AM To: Murphy, Shawn N.; Kahn, Michael; Russ Waitman; Phillips, Lori C. Subject: RE: i2b2 ontology work based on ONC standards...... In preparing an analysis of data harmonization across i2b2 platforms, SHRINE came up as a reference. Shawn, is there a paper which documents the information model and domain ontologies currently employed at SHRINE sites? Jim ________________________________ From: Murphy, Shawn N. [[email protected]] Sent: Wednesday, April 23, 2014 9:25 PM To: Kahn, Michael; Campbell, James R; Russ Waitman; George Hripcsak ([email protected]<mailto:[email protected]>); Bonnie Westra; Thomas Campion ?[[email protected]]?; [email protected]<mailto:[email protected]>; Keith Marsolo; Charles Bailey; Phillips, Lori C.; Wehbe, Firas; Matthew Hoag; [email protected]<mailto:[email protected]>; [email protected]<mailto:[email protected]>; Connolly Dan Cc: Flores, Dianna L; [email protected]<mailto:[email protected]>; '[email protected]'; [email protected]<mailto:[email protected]>; Klann, Jeffrey G. Subject: RE: i2b2 ontology work based on ONC standards...... Hi All, It occurs to me that I’m not sure specifically what are we trying to achieve. I think the discussion is drifting around two very important concepts. One concept is that the i2b2 ontology can be used to represent different information models. In this sense, the information model behind meaningful use can be expressed, and this is what Mike Buck has achieved. A second concept is that we need an i2b2 ontology to help us to transform i2b2 data into the PCORNet mini-sentinel data model. What it means for us to be able to transform i2b2 data into the PCORNet mini-sentinel “data model” is that we need to be able to represent the PCORNet mini-sentinel Information Model in i2b2. What many people have pointed out by comparing i2b2 to OMOP, and i2b2 to mini-sentinel, is that the tables that correspond to a data model in i2b2 do not define an Information Model. In i2b2 you also need the addition of the ontology to define the Information Model, and depending on the ontology that is constructed one can define different Information Models in i2b2 (often quite ad-hockly as others have pointed out). In principle, if the i2b2 ontology adequately describes the PCORNet mini-sentinel Information model, the ontology should then be able to define a mathematical transformation from one to the other, because they now both represent the same information Model. So we could make an ontology for the Meaningful Use Information Model, the PCORNet mini-sentinel Information Model, the OMOP Information Model, or any of these, but they are not all the same thing, and each one will not be a trivial task or even have just one solution (thus the benefit of standardization for any of these). So do we want to divide up on this task? From Jeff Klann’s and my perspective, right now, we absolutely have to make an i2b2 ontology that represents the PCORNet mini-sentinel Information Model to fulfill our obligations to PCORI, so is there a specific group forming around that task? Thanks, Shawn. From: Kahn, Michael [mailto:[email protected]] Sent: Wednesday, April 23, 2014 10:10 AM To: Campbell, James R; Russ Waitman; George Hripcsak ([email protected]<mailto:[email protected]>); Murphy, Shawn N.; Bonnie Westra; Thomas Campion [[email protected]]; [email protected]<mailto:[email protected]>; Keith Marsolo; Charles Bailey; Phillips, Lori C.; Wehbe, Firas; Matthew Hoag; [email protected]<mailto:[email protected]>; [email protected]<mailto:[email protected]>; Connolly Dan Cc: Flores, Dianna L; [email protected]<mailto:[email protected]>; '[email protected]'; [email protected]<mailto:[email protected]>; Klann, Jeffrey G. Subject: i2b2 ontology work based on ONC standards...... Added Lisa Dahm, Daniella Meeker, Doug Bell, and Jeffrey Klann to this email thread. Please add them to the core distribution list. Daniella pointed me to work done as part of the ONC S&I framework to create an i2b2 ontology based on CCDA. The link to that ontology is at https://community.i2b2.org/wiki/display/queryhealth/Home, about half way down the page past the HQMF stuff. Looks like it was done by Jeff Klann, who I added to this distribution list. The ontology was posted 9/2013 so it is reasonably recent. Could this ontology be the starting point for a MU-standards based ontology??????? Michael Kahn From: Campbell, James R [mailto:[email protected]] Sent: Wednesday, April 23, 2014 8:00 AM To: Kahn, Michael; Russ Waitman; George Hripcsak ([email protected]<mailto:[email protected]>); Murphy, Shawn N.; Bonnie Westra; Thomas Campion [[email protected]]; [email protected]<mailto:[email protected]>; Keith Marsolo; Charles Bailey; Phillips, Lori C.; Wehbe, Firas; Matthew Hoag; [email protected]<mailto:[email protected]>; [email protected]<mailto:[email protected]>; Campbell, James R; Connolly Dan Cc: Flores, Dianna L Subject: Thank you everyone who could participate in yesterday's call on i2b2 standardization. I apologize to those who were not familiar with UTC for meeting planning, since everyone in US I suppose it is frivolous. I have updated the working document with attendees and deliverables from yesterdays meeting. My secretary Dianna will be sending out a doodle poll for best available time for our next call in two weeks, Tuesday May 6 Jim 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. The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
_______________________________________________ Gpc-dev mailing list [email protected] http://listserv.kumc.edu/mailman/listinfo/gpc-dev
