Informatics Core

Gustavo Jimenez-Maggiora, MBA

Washington, DC

April 19th, 2015

Updates

New Data Portals

Amazon Web Services (AWS)

• We have migrated the majority of our Data Portals to AWS
• Remaining Data Portals:
  • A4
  • TCAD

Application Improvements

• Based on user feedback, we have made improvements to the following apps:
  • Images
  • IP Tracking
  • MedDRA
  • Queries

Images

Improvements

• Multiple modalities
• Study-specific, multi-user collaborative workflows
• Traceability and transparency
• Tagging capabilities allow users to flexibly organize and search for images
• Manage processed or derived images
  • Link derived images to source images to establish provenance across modalities
  • Visualize and navigate using inheritance trees
  • Multi-file upload, linking and tagging
• Improved reporting and tracking of workflow inventory and performance metrics

Release Schedule

• Launched on FYN
• Plan to upgrade in the next 6 months:
  • A4 / LEARN
  • INI
  • TCAD

IP Tracking

Improvements

• Cleaner, more intuitive interface
  • Configurable data per menu
  • Smart action links
  • Additional details…
• Advanced filters
• Color statuses
• New “Cart” interface
• Automated drug ordering/resupply
• Additional new features based on user feedback
  • Download CSV
  • Status report (interactive chart report)

Release Schedule

• Launched on FYN
• TCAD - May, 2015
• INI - ?

MedDRA

Improvements

• Support for MedDRA updates
• Text-based search and filtering
• Improved batch actions (coding and confirmation)

Release Schedule

• Launched on all active ADCS data portals!!!

Queries

Improvements

• Usability and Layout
  • All query related-information is visible
  • Modal vs drawer
• More filters
  • Created-by
  • Touched-by
  • Query Age
  • Tags
• Saved searches
• Tagging
• Starring - follow a query

Release Schedule

• Launched on FYN and CCF-RAS
• Plan to upgrade in the next 6 months:
  • A4 / LEARN
  • INI
  • TCAD

How do we ensure that these changes are validated?

Validation 1.0

When any change (even a small change) is made to the software, the validation status of the software needs to be re-established.

Initial Validation

• Completed in December 2012
• Focused on Core Modules and Functionality:
  • Electronic Data Capture (EDC)
  • Queries
  • Audit Trail
  • Data Export
  • Document Repository
  • Authentication and Authorization
  • Meta-data Management

Initial Validation

• Initial timeline: 3 months
• Actual time to completion: 9 months
• Traditional approach
  • Manual test execution
  • Paper-based
• Required 50k+ person hours

Initial Validation

• Observations:
  • Traditional approach is resource intensive and difficult to replicate
  • Our user community demands constant innovation
  • Our business requirements are evolving and study-specific
  • By design, we have a key dependency on web browsers, which are rapidly changing

"Timeline of web browsers" by I, ADeveria. Licensed under CC BY-SA 3.0 via Wikimedia Commons

Rapid Change in Browsers

• Firefox and Chrome release a new version every 6 weeks
• Microsoft recently deprecated support for Internet Explorer
  • Phaseout of Internet Explorer Stirs No Grief
• Microsoft is launching Spartan

Browser Share on Data Portals (Q1 2015)

Browser Usage on Data Portals (6 months to 3/31/2015)

Initial Validation

• Conclusions:
  • Develop a new approach that effectively combines processes, standards and technology to maintain a continuous, documented, reproducible, and structured quality review of our clinical trials data management system
  • Furthermore, this new approach should allow us to rapidly validate new functionality and browser versions

Automated Validation

Software Development Life Cycle

• Request
• Approval
• Design
• Specification
• Requirements
• Development
• Testing
• UAT
• Release
• Configuration
• Deployment

Software Development Life Cycle

• Request
• Approval
• Design
• Specification
• Requirements
• Development
• Testing
• UAT
• Release
• Configuration
• Deployment

Tools

• Git (Version Control)
• Github (Cloud-based Version Control Service)
• Jenkins (Task Queue)
• Amazon Web Services (Platform as a Service)
• Robot (Automated Functional Testing)
• Foreman + Puppet (Provisioning and Configuration Management)
• Spacewalk (Package Management)

Benefits

• Manage and test all software and system deliverables on a ongoing basis
  • Reproducibility
  • Flexibility
  • Scalability
  • Visibility
  • Metrics
• Testing becomes central to the overall development process

Status

• We completed development of the automated validation framework in July 2014
• We completed the first round of automated validation activities in December 2014
• We achieved full coverage of all Data Portal apps in March 2015

Status

• The following tests are running nightly:
  • 1k+ Functional Tests
  • 8k+ Unit Tests
• The functional tests are running in Chrome and Firefox
• We plan to continue to expand our test coverage

Robot in Action

test-video

Roadmap

• Improvements to the Help app
• Working on with LONI to add ADCS data to GAAIN
• Preparing for A5 and future studies
• Developing new approaches to integration of external data and data sharing
• Working with Safety Stats on RBM
• Working with Stats on Reporting
• Working with Biomarker on LIMS

Thank You