PhUSE CSS Working Groups An Experiment in

Report
PhUSE Computational
Science Symposium
Working Groups
An Experiment in Collaboration
What is PhUSE?
Pharmaceutical Users Software Exchange
• Global group of programmers, statisticians, and data
scientists
• Vendor neutral, Inclusive, and Open
• www.phuse.eu
Mission:
Provide platform for creating & sharing ideas,
implementing tools & standards around data, and
exploring innovative methods and technologies
Old Paradigm
FDA
Please tell us
what to do
Paradigm Shift
CSS Timeline
2008: PhUSE and FDA collaborate on Wiki
2009: FDA forms CSC
2010/11: FDA hosts first two CSC meetings
2012: First Annual CSS Meeting
2013/14: Annual CSS Meetings
What CSS WGs Are?
Implementation
• Work on best practices for standards and technology
implementation
• Discuss and summarize challenges with implementation
Incubator
• Try ideas using new technology e.g., semantic modeling, cloudbased implementation
• Test and provide feedback on new standards and processes
Collaboration
• Open forum for sharing ideas across FDA, Industry, and other
organizations
• Provide platforms (wikis, code libraries) for sharing information
What CSS WGs are Not
Policy
• Projects do not create, define and/or
recommend policy or guideline changes
• FDA individuals provide input only
Standards
• CDISC, HL7 and other SDOs create standards;
• CSS Working Groups tackle implementation and
best practices around standards
Commercial
• Projects are not a place to ‘market’ technology
• Open collaboration in a non-competitive
environment
CSS Working Groups
Optimizing Use of
Data Standards
Emerging
Technologies
Steering
Committee
Standard Scripts for
Reporting and
Analysis
Non-Clinical
Roadmap and
Implementation
Steering Committee
Industry
• Chris Decker – Industry PM (d-Wise)
• Scott Bahlavooni (Biogen Idec)
• Anne Russotto – (Celgene)
• Michael Brennan (J&J)
• Jim Johnson – Optimizing Standards
Liaison (Summit Analytical)
• Susan DeHaven – Non-Clinical Liaison
(Sanofi)
• Mary Nilsson – Standard Scripts Liaison
(Eli Lilly)
• Frederik Malfait – Emerging
Technologies Liaison (IMOS Consulting)
FDA
• Crystal Allard – FDA PM
• Lilliam Rosario
• Steve Wilson
• Suzie McCune
Working Group
Leadership Teams
• Emerging Technologies
–
–
–
–
Frederik Malfait (Roche)
Isabelle deZegher (Paraxel)
Matteo Ditommaso (Pfizer)
Crystal Allard, FDA Liaison
• Optimizing Data Standards
– Jim Johnson (Summit
Analytical
– Jingyee Kou, FDA Liaison
– Steve Wilson, FDA Liaison
• Standard Scripts
– Mary Nilsson (Eli Lilly)
– Hanming Tu (Accenture)
– Steve Wilson, FDA Liaison
• Non-Clinical
– Susan DeHaven (Sanofi)
– Bob Dorsam, FDA Liaison
Lot of other volunteers leading projects!
Optimizing the Use of
Data Standards
Overview:
The development and implementation of data
standards has shown significant promise in
improving efficiencies in the product submission
and review process. However, there are challenges
in the interpretation and use of the standards. This
working group identifies specific gaps and best
practices to enable FDA and industry to maximize
the benefits of standards implementation.
Optimizing Data Standards
Active Projects
Traceability and
Data Flow
ADaM Data
Reviewer’s
Guide
Best Practices
Study Data
Standardization
Plan
ODS Projects
Traceability and Data Flow
• Leaders: Paul Bukowiec, Sandra Minjoe, Tanja
Petrowitsch, Natalie Reynolds
• This project will discuss and define traceability
considerations and best practices for study level
dataset and integrated datasets conversion for a
variety of different data flow scenarios.
ODS Projects
Traceability and Data Flow
• Meeting Accomplishments:
– Developed first draft of Study Level Traceability
white paper
– Noted specific traceability details to be added to
current Basic Linear Data Flow white paper
– Determined the need to develop a template for the
Legacy Data Conversion Plan (mentioned in the new
draft FDA Study Data Technical Conformance Guide)
ODS Projects
Analysis Data Reviewer’s Guide (ADRG)
• Leaders: Susan Kenny, Gail Stoner
• Development of an ADRG template and instructions for
industry to consider that will enable consistent and
usable ADaM documentation in submissions.
ODS Projects
Analysis Data Reviewer’s Guide (ADRG)
• Accomplished at Meeting
– Completed final review of Template
– Completed final review of Completion Guideline
– Presented poster and expect to receive additional
comments. Comments due by MARCH 28.
– Began review of one ADRG example
• By April 30:
– Complete review and updates to 2 examples
– Finalize ALL documents (template, guidelines, samples)
– Post all materials to PhUSE Wiki
ODS Projects
Best Practices for Standards Implementation (NEW)
• Leaders: Mike Molter, Lisa Lyons
• Goal: develop a set of recommendations for best
practices in optimizing the data standards.
• Each topic will be ~3-5 months duration for delivery of
the recommendation
• Each best practice will have a sub-team
• First topics include:
– Lab Unit Standardization
– EPOCH/VISITNUM assignments
– USUBJID assignments
Data Optimization: Best
Practices
• Future Best Practices Topics Identified
– Programming Documentation
– Reference Range Standardization
– Unit Standardization Beyond Labs
– Trial Design Setup
– Treatment Emergent Flags
– ARM/ARMCD assignment and standard implementation
– Others welcome – Will be a place to identify new topics
on Best Practices Wiki
ODS Project
• Study Data Standardization Plan (SDSP)
– New Project Starting in 2014
– Goal of the Project is to develop a Template to be
made available for Industry
– Template Model will follow a similar development as
the SDTM and ADaM Reviewer’s Guide
Standard Scripts for
Analysis and Programming
Overview:
The development and implementation of data
standards provides a great opportunity to develop
standard analyses and displays to support the
needs of FDA medical and statistical reviewers. This
working group will identify potential standard
scripts for data transformation, analyses, and
displays.
Refined Projects
• P01: Look for existing scripts and store them in the repository Work on FDA scripts first
• P02: Define qualification steps for scripts in the repository
• P03: Maintain and enhance platform (repository) for sharing
scripts
• P04: Create templates and metadata for documenting scripts
and coding practices
• P05: Implement and further develop communication plan for
standard scripts
• P06: Create white papers providing recommended display and
analysis including Table, List and Figure shells
Scriptathon Summary
• Scriptathon – Success!!!!
–
–
–
–
26 attendees/15 coders
13 Scripts developed
Volunteers for the remaining scripts
Organize local and annual events
Whitepaper Summary
• White Papers
– Great Discussion!
– Six whitepapers
•
•
•
•
•
•
ECG, Vitals, Labs – Central Tendency: finalized in 2013
ECG, Vital, Labs – Outlier/Shifts
Adverse Events
Demographics, Disposition, Medications
Hepatotoxicity
PK
– One additional white paper planned to be added (QT
Studies)
– Look for writing support
Plans for Upcoming Year
• Continue to build Script Repository
– Continue creating code for tables and figures as outlined in
white papers
– Look for other existing code sharing libraries
• Continue white paper development
–
–
–
–
Consider updating the first white paper
Finalize PK white paper this month
Finalize remaining 4 white papers in 2014
Consider additional white papers to develop
• Communications
– Presentations at various conferences
– Implement other ideas to increase awareness
Non-Clinical Roadmap and
Impacts on Implementation
Overview:
A need exists to improve nonclinical assessments
and regulatory science by identifying key needs and
challenges then establishing an innovative
framework for addressing them in a collaborative
manner. This working group establishes the
collaborative framework and identifies and
executes projects to support nonclinical informatics
and specific implementation solutions for SEND.
Nonclinical Working Group
6 projects with significant accomplishments over the past year
- White paper published (3/14):
“Interconnectivity of Disparate Nonclinical Data Silos for Drug Discovery and
Development”
- Manuscript developed on industry use of Nonclinical Historical Control Data
- Expansion of SEND Wiki, with QA perspectives on SEND – poster
- Compared Clinical vs Nonclinical Study Data Reviewer’s Guide – poster
- Development and testing of interorganizational flow of e-data – 2 posters
- “How to Design a Custom SDTM Domain for Nonclinical Data”
CSS PhUSE 2014 Theme
Developing Collaborations
Our Sessions
1. Panel discussion – “How are clinical and nonclinical data coming together in
your environment?”
• 3 presenters shared experiences (vendors, sponsors, consortia)
2. Develop collaborations across PhUSE and others –
• Interactive discussion on possible projects with Emerging Technologies,
IMIeTox, and Transcelerate
• Goal to increase potential for collaborative new projects
3. Team time
• Opportunity for NC projects to get feedback from full WG on
accomplishments and future directions
• New project idea triage and action plans
4. Leveraging Deliverables
• “Table Teams” contributed their experience Socializing the Deliverables
Actions for 2014
• AND……..
Turbocharge the SEND Implementation User
Group!
– New members, WIKI tools, communication
subteams
Emerging Technologies
Overview:
Regulatory science, drug, biologic, and new
device development challenged present a
unique opportunity to apply underutilized
existing technologies and/or new and emerging
technologies. This working group provides an
open, transparent, pre-competitive forum for
the exploration and application of said
technologies.
Metadata Definition
Leader: Isabelle deZegher
•
•
30
No formal meeting
Finalize document through TC – Available Apr2014
ST: Representing
PRM/SDM.xml in RDF
• Review the current draft model
• Discuss modeling of study parameters/parameter
groups, study activities and planned activities
• Discuss next steps for the RDF model
– Do not want it to become another “under-used”
representation of structured protocol content
ST: Analysis Results
Metadata
• Introduce team and concept to the broader
community
• Developed draft metadata for count and
summary statistic analysis procedures
• Identified central tendency, t-test, ANOVA as next
analyses to develop
• Identified strategy for researching publically
available existing models
• Identified alignment needs with the define.xml
team and standard scripts repository white paper
team
Summary
• Great opportunity for FDA and Industry to
work together
• Evolving and changing
• Typical Challenges of volunteers and time
• Exciting interactions and discussion
Get Involved
• See the PhUSE CSS Dashboard for an overview of active
projects
– http://www.phuse.eu/CSS-dashboard.aspx
• More detailed project and working group information is
available on the PhUSE Wiki
– http://www.phusewiki.org/wiki/index.php?title=PhUSE_Wiki
• Email and working group lead or sign up for a mailing list
– http://www.phusewiki.org/wiki/index.php?title=Listbox_Instr
uctions

similar documents