1. e-data Implementation Impact
Discussion Panel
Summary of Experiences & Challenges for Pharma, CROs, Software Vendors, Consultants & Regulators
2018 CSS
06 Mar 2018

2. Breakout 2 - Impact Panel Discussion
Method:
Distribute the attendees to 5 tables - each dedicated to their “industry”
Pharma, CROs, Software Vendors, Regulators, Consultants
Each group to answer, from their perspective:
What are the best impacts of e-data submission requirements on your organization?
What challenges have you overcome?
What current challenges are you facing?

3. Pharma 1 of 3
Best Impacts
Data visualization increase data review speed & enhanced interpretation
Use of data visualization able to influence organization (company?)
Forcing standardization
Structured data warehouses
Organized way to retrieve data from CROs (data repatriation)
Surfaced assumptions that were not previously identified

4. Pharma 2 of 3 Challenges Overcome Senior management SEND buy-in
Terminology mapping (somewhat still a challenge)
Learning lingo/terminology
Establishment of a data management group
Merging of datasets (multisite studies) into one submission
Challenge of 2 systems talking to each other (e.g., ECG, TK, worklists)
Pharma/CRO alignment on what should be in a dataset
Use of data visualization able to increase engagement of scientists with SEND
User communities coming together to influence tools

5. Pharma 3 of 3 Current Challenges
Study Data Standardization Plan (SDPS) – getting alignment across clinical, nonclinical & regulatory
Generating SEND datasets for studies with challenging designs (e.g., escalating dose and what is day 1 for x group)
Delays in getting datasets – can’t visualize data internally before submitting/reporting
Multisite studies – coordinating communications, handoffs, timing, SDRG generation, org limitations
Gaps in data collection systems that can’t generate SEND datasets (manual generation required)
Getting quality & conformance when dealing with different CROs (lack of harmonization)
Unscheduled activities can be more difficult to handle in SEND than scheduled
Some domain variable still subject to interpretation – leading to inconsistencies – examples?
Study endpoints that aren’t in SEND yet won’t be included
Concern that nonclinical could become rate limiting at NDA submission
Projecting resources & costs dependent – can be dependent on strategy when generating a dataset
Getting more frequent feedback from agency to understand if we are doing a good job

6. CROs 1 of 2 Best Impacts Challenges Overcome
Consistency in data collection and controlled terminology
All around better communication
Streamlining work processes
Driven upgrades/updates
Challenges Overcome
Tool flexibility/manual manipulation
Adjusting to change
Purchasing & implementing software/commercial products

7. CROs 2 of 2 Current Challenges Lack of feedback
Educating clients (needs & costs)
Verification/QC
Resources (timelines, employee needs, charge costs)
Keeping up with the standards
Organizing data from outside vendors

8. Software Vendors 1 of 2 Best Impacts Challenges Overcome
SEND investment
Data collection – submission warehouse
Upgrades
Data sharing projects
Challenges Overcome
Consumption SEND
Automation checking
Trial summary population
Collection  output

9. Software Vendors 2 of 2 Current Challenges Visualization desired
But don’t have any scope to provide
Data from multiple LIMS including handwritten records
Harmonizing into a single cohesive dataset
Data gaps
Specimen or other data not captured in the LIMS system.
Filling and adding information traditionally not stored in the LIMS
Unstructured data (correlations)
Data correlation via narrative
Data collected – SEND aware CT
Study Summary tables based on same system creating SEND output
Increase harmonization of report and SEND dataset
Moving target of SENDIG 3.0 interpretation - TCG requires changing implementation of SEND variable

10. Consultants Best Impacts Business opportunities Revenue stream
It is the business
Innovation
Renewed interest in the usefulness of legacy data
Moving to future use of data – it’s not just one drug, one study, one submission

11. Consultants 1 of 2 Challenges Overcome Custom SDTM datasets
QC checking – finding inconsistencies
Filling the gap for expertise, education & workforce
Providing a level of reassurance, confident responses
Tools – creation, vetting & using

12. Consultants 2 of 2 Current Challenges
SEND still an infant – change happens, but movement is not synced up
Lack of consistent feedback from regulators
Lack of platform for information sharing
Data collection practices have not caught up – unwillingness to optimize input
Trial design
Access to multiple tools
Divergence between SEND output and the study report

13. Regulators/Standards Organization
Best Impacts
Getting usable data
Using CDISC standards
Challenges Overcome
Issuing guidances
Pilots
Current Challenges
Training
Heterogenicity of implementation
Harmonization of standards