PDDI Info Model Task Force - CONTENT Sub-team minutes for 6/30

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

PDDI Info Model Task Force - CONTENT Sub-team minutes for 6/30

Richard Boyce-2
Hi Team,

The minutes from the All Team Meeting (7/28), and the standard and content sub-team meetings (7/21 and 7/19 respectively) are pasted below (online at https://goo.gl/ghMvr3). Session recording links are provided for the sub-team meetings. Sorry about accidentally not recording the all team meeting. Thanks to Elizabeth Garcia for writing the minutes!

Kind regards, -R


Minutes for 7/28/2016 (All-Team Meeting)

In Attendance:  Mathias Brochhausen, Ammar Halawa, Brian Hocum, Brian LeBaron, Dan Malone, Evan Draper, John Horn, Maria Herrero, Matthew Breitenstein, Michael Miller, Øystein Nytrø, Serkan Ayvaz, Xia Jing, Harry Hochheiser, Katrina Romagnoli, Daniel Rubin, Guillermo Vega Gorgojo, Elizabeth Garcia, Louisa Zhang

Meeting recording:  None (accidental)

Meeting:

  • Task Force Progress to Date

    • Content sub-group

      • Exemplar potential drug-drug interactions

      • Creation of decision trees for specific interactions; some of which adopted from AHRQ Task Force

        • Team developing new decision trees; there will be 15-20 decision trees total

        • Not yet public (can request via email), but will eventually be published

    • Standard sub-group

      • Scope of knowledge representation for the project

      • Setting down guidelines for how ontology will be developed

  • User-Centered Definitions:  https://goo.gl/xGeUr2

    • Completed

      • Clinical Consequences

      • Evidence

      • Mechanism of Action

    • In Progress

      • Recommended Action:  will be going out as a Qualtrics survey

      • Frequency of Harm within the context of Frequency of Exposure

        • Pending; combination into one user-centered definition

    • Remaining:

      • Seriousness Rating; Contextual Information/Modifying Factors

      • By third week of August

  • W3C Interest Group Note:  http://dbmi-icode-01.dbmi.pitt.edu/dikb-evidence/w3c-ddi/index.html

    • Open source, broadly distributed

    • Lays down a nice set of artifacts that can be further enhanced

    • Everything in draft note right now

    • Google Docs under each section

      • Comments to refer to literature

      • Use “track changes” (Suggestions) within Google Doc to make edits

    • This first note will be the introduction

      • Laying down the purpose and the guidelines

    • o   There will be a second note that focuses on the actual model

      • Ontologies and terminologies; genus differentia

      • Going deep into representation

      • Artifacts:  potential drug-drug interactions implemented within the model

      • Can discuss why certain decisions were made

  • W3C Interest Group Note Sections:

    • Introduction:  https://goo.gl/eDX72y

      • What the motivation is

      • What methods were used to arrive at:

        • User-Centered Definitions

        • User Stories

        • Med Rec Use Cases

        • Knowledge Representation

    • How drug interactions were selected

      • Not the most important, or contraindicated, but rather situations that highlight boundary cases in the minimum information model

        • E.g., applies at the class level, does not apply at the class level, etc.

      • Write-up will discuss how pairs were selected

      • Exemplar potential drug-drug interactions:  https://goo.gl/P1In54

    • User Stories:  https://goo.gl/Pb21wt

      • Giving a face to the user who would need to have some contact with DDI information

        • Drug-drug interacting pairs selected from exemplars

      • Brainstorming stakeholders; led to comprehensive list that was narrowed down to tighten scope

      • Came out of Interviews, reviewing internal interview transcripts, literature

      • Color-coding scheme applying information model definitions

      • Information needs related to different users also listed

    • Knowledge Representation Core Considerations:  https://goo.gl/jhwj9e

      • Mathias Brochhausen and a team within Standard sub-group

      • Goal is to set the scope for how to build the information model as an ontology

      • Level of representation required

      • Reusing terms, building ontologies

      • OWL File:  Mathias (DIDEO), Maria (DINTO)

      • Invitation for comments

    • Questions/Comments

      • Problem within other DDI task groups:  tendency for suggestions to not be tracked

        • Keeping track of all comments in appendices; what was commented, what was suggested, etc.

        • Why they didn’t make it into final versions

      • Everyone is welcome to comment on these Google Docs

  • Medication Reconciliation Use Cases (Elizabeth):  https://goo.gl/x86CLm

    • Øystein Nytrø discussed medication reconciliation use cases as a rallying point for the task force

    • Workshopped through the three use cases

      • Goal is to show a detailed user workflow that incorporates elements of the task force information model

      • Modifying factors:  being able to tie to controlled vocabularies

      • Evidence:  pull out from other cases

      • Edits will be made to tie the use cases to the model, highlight definitions

      • Qualtrics will be sent out to elicit feedback

        • Enhancements, improvements

        • Other elements the information model could or should support

    • Please feel free to comment on the Google Doc as well

  • Next Steps

    • Sub-teams will continue to meet

      • Content team will wrap up decision trees

      • Standard team will wrap up definitions

      • Qualtrics will be sent out re: user-centered definitions, medication reconciliation use cases

      • Doodle Polls will be sent out July 29 to plan remaining sub-team meetings

    • After October:  more All-Team meetings

      • Progress in terms of writing

      • Goal is to have this all written up by the end of the year



Minutes for 7/21/2016 (Standard subgroup)

In Attendance:  Mathias Brochhausen, Kim Nolen, Chris Vitale, Michel Dumontier, Maria Herrero, Kim Nolen, Xia Jing, Oliver He, Brian Hocum, Brian LeBaron, Richard Boyce, Louisa Zhang, Elizabeth Garcia

Meeting recording:  http://goo.gl/Fb498l

Meeting:

  • Update from Content group:  Progress on decision trees

    • Minutes from Decision Tree group meeting lists all decision trees in progress:  https://goo.gl/Mqi8YC

    • Want to include DDIs that reflect information model elements

      • Suggested PDDI spreadsheet:  https://goo.gl/P1In54

      • Selected to depict boundary cases; minimum information model would benefit from this pair because it highlights a PDDI that should be contextualized

      • Clear definitions for clinicians; want to represent this information as a goal of the task force

      • Seriousness is really dependent on the patient and the situation; not just “this is serious,” but “this is serious when…”

    • Overall changes to clinical scope

      • Recommended actions limited to 3:  no special precautions, assess risk and take precautions as necessary, use only if benefit outweighs risk

        • We inherited the idea of creating decision trees from the AHRQ task force (alerting); footnotes will say what actions to take

      • Limiting work to prescription drug-drug interactions

        • For now, not including over-the-counters or herbals to keep the focus simpler

      • Enzyme inhibitors

        • What’s the list of enzymes?  List of evidence to support recommendations

  • Progress on mechanism of interaction user-centered definition

    • User-Centered Definitions:  https://goo.gl/xGeUr2 (Can comment)

    • Discussion re: scope of what is included under “process”

      • For the purposes of the information model, representing mechanisms in great detail will not be possible

      • At a higher level, there are information sources that are not clinically helpful; this is trying to address that problem

    • Edits discussed and approved

  • Update from Mathias:  PDDI Task Force – KR subgroup

    • Knowledge Representation Core Considerations  will be split into two parts:

      • Basic representation of scope

      • Content document

    • Meeting next week to finalize; close to getting a complete document

    • Next task will be to start building a small web ontology language file (OWL file)

      • If anyone would like to help Mathias with the OWL development process (more than just commenting), please let him know

    • Definitions and Drug Interactions now available to start making parallel progress

  • Update on User Stories

    • PDDI User Stories and Information Needs:  https://goo.gl/Pb21wt  (please comment)

    • Color-coded to show information model user-centered definitions

    • Presented various user stories and made edits based on recommendations

  • Update on Detailed Medication Reconciliation Use Cases

    • In-depth cases that show what needs to be queried, what is needed in a clinical workflow, etc.

    • Please comment on the following Google Doc:  https://goo.gl/x86CLm  

    • Will be discussed in-depth at the upcoming All-Team Meeting

  • Next Steps

    • Please fill out All-Team Meeting Doodle Poll

      • Meeting will be finalized ASAP

    • Doodle Poll will be sent July 25 to schedule August, September, and October meetings

    • User-Centered Definitions

      • Qualtrics re: “Recommended Action” will be sent out July 27/28

      • “Frequency of Harm” definition also being developed; will be developed within the context of “Frequency of Exposure”

    • Detailed Medication Reconciliation Use Cases

      • Please review and add comments prior to All-Team meeting


Minutes for 7/19/2016 (Content subgroup)

In Attendance:  Brian Le Baron, John Horn, Serkan Ayvaz, Evan Draper, Richard Boyce, Dan Malone, Michel Dumontier, Oktie Hassanzadeh, Elizabeth Garcia, Louisa Zhang

Meeting recording: http://goo.gl/vWnKz0

Meeting:

  • Upcoming

    • All-team Doodle poll:  opportunity for all of us to meet

    • Final progress on user needs (user stories, medication reconciliation use cases)

    • Mathias Brochhausen’s sub-team meeting on developing the actual ontology, including definitions and how things will be represented

  • Progress on Decision Trees

    • Good progress on 6 interactions (in addition to 3 already developed for interactions picked)

    • Meeting minutes have been posted in Task Force Meeting Minutes:  https://goo.gl/suIMnS

    • Next steps will be to finalize these 6 decision trees at an upcoming meeting

      • Doodle Poll will be sent out

  • Progress on mechanism of interaction user-centered definition

    • User-Centered Definitions:  https://goo.gl/xGeUr2  (Can comment)

    • “Information Content Entity” discussion will be used to build the ontology

    • Importance of attaching “clinical consequences” to the definition for the purposes of the minimum information model

    • Edits discussed and approved

  • User Needs

    • Goal is to finalize these

      • Putting a fictional face to someone who would potentially be using the minimum information model

    • Write-up is going to describe the minimum information model

    • Will break into two interest group notes

      • One introduces the topic and sets the scope and approach for knowledge representations

      • Another will discuss the model itself and representation

  • Update on User Stories

    • PDDI User Stories and Information Needs:  https://goo.gl/Pb21wt

    • Workshopped through various user stories

      • “Screening” category changed to “Evaluation of Management Options” for several stories

      • “Nurse” user story will focus more on “Screening”

      • “Clinical Decision Support” user story will be split into content user and technical user

        • Turned into a team; need both a content expert and a technical expert in order to design rules and implement them

      • Information Model elements will be highlighted and color-coded

      • Please review and add comments with questions or suggestions

  • Update on Detailed Medication Reconciliation Use Cases

    • In-depth cases that show what needs to be queried, what is needed in a clinical workflow, etc.

    • Please comment on the following Google Doc:  https://goo.gl/x86CLm

    • Will be discussed in-depth at the upcoming All-Team Meeting

  • Next Steps

    • Please fill out All-Team Meeting Doodle Poll

      • Meeting will be finalized ASAP

    • Doodle Poll will be sent ~July 25 to schedule September and October meetings

    • User-Centered Definitions

      • Qualtrics re: “Recommended Action” will be sent out July 27/28

      • “Frequency of Harm” definition also being developed; will be developed within the context of “Frequency of Exposure”

    • Decision Tree Sub-Group

      • Look out for Doodle Poll within the next week (August 25 – September 2 meeting)

    • Detailed Medication Reconciliation Use Cases

      • Please review and add comments prior to All-Team meeting


Reply | Threaded
Open this post in threaded view
|

Re: PDDI Info Model Task Force - CONTENT Sub-team minutes for 6/30

Renato Iannella-3

On 29 Jul 2016, at 20:00, Richard Boyce <[hidden email]> wrote:


This is good to see the use of other ontologies as the foundation for this work (eg BFO etc).

My question is, what is the intended relationship (if any) to the FHIR Ontology?


Cheers...
Renato Iannella
Founder & Principal Architect, Semantic Identity
+61 4 1313 2206  @riannella  http://semanticidentity.com


Reply | Threaded
Open this post in threaded view
|

Re: PDDI Info Model Task Force - CONTENT Sub-team minutes for 6/30

Richard Boyce-2
Hi Renato, we haven't discussed this yet on any calls but I suspect that showing how the standard could be used within FHIR would be very useful and we could add this as a goal  Are you interested in helping with that? If so, I would encourage you to join our calls because we will be starting to build artifacts soon and would benefit from the insight of folks who know FHIR well.

best,
-R

On 07/31/2016 09:40 PM, Renato Iannella wrote:

On 29 Jul 2016, at 20:00, Richard Boyce <[hidden email]> wrote:


This is good to see the use of other ontologies as the foundation for this work (eg BFO etc).

My question is, what is the intended relationship (if any) to the FHIR Ontology?


Cheers...
Renato Iannella
Founder & Principal Architect, Semantic Identity
+61 4 1313 2206  @riannella  http://semanticidentity.com




-- 
Richard D Boyce, PhD
Assistant Professor of Biomedical Informatics
Faculty, Center for Pharmaceutical Policy and Prescribing 
Faculty, Geriatric Pharmaceutical Outcomes and Gero-Informatics Research and Training Program
University of Pittsburgh
[hidden email]
Office: 412-648-9219
Twitter: @bhaapgh