Transcript Analysis Databases: A Regulatory Perspective

Analysis Databases:
A Regulatory Perspective
Steve Wilson, DrPH
Deputy Director
Division of Biometrics II/FDA/CDER
2003 FDA/Industry Statistics Workshop
"Statistics: From Theory to Regulatory Acceptance“
Parallel Session II: Analysis Databases
Hyatt Regency, Bethesda, Maryland
September 18-19, 2003
Analysis Databases:
Let’s Get a Guidance Pronto
Steve Wilson, DrPH
Deputy Director
Division of Biometrics II/FDA/CDER
2003 FDA/Industry Statistics Workshop
"Statistics: From Theory to Regulatory Acceptance“
Parallel Session II: Analysis Databases
Hyatt Regency, Bethesda, Maryland
September 18-19, 2003
Disclaimer
Views expressed in this
presentation are those of
the speaker and not,
necessarily, of the Food and
Drug Administration
Outline
• The way we were
• The way we are
• Standards development
• The urgent need for a guidance on the
submission of analysis files
• New eCTD guidance
• Developing a guidance
• CDISC and ADaM
The Way We Were
• Statistical review focused on efficacy
• Data submitted to reviewers (“desk copies”)
– No EDR
– No archive standards (e.g., Transport 5, PDF, …)
•
•
•
•
•
•
•
The luxury of time (pre-PDUFA)
Less workload
Delayed decisions
Encourage, but not insist
Every study a new challenge
Routine analyses
Pharmaceutical development separate from
healthcare.
The Way We Are
•
•
•
•
•
•
•
•
Risk-Benefit / Safety - Efficacy
Medical reviewers “want” data for “analysis”
PDUFA limits time and options
Tough realities related to Science (no more kneejerk LOCF)
Approvable cycles
Submission standards (Transport 5, PDF)
Reviewer templates
Data “warehouse” – more standards (HL7,
Version 3, SDS, CDISC, HIPAA, etc.)
Standards Development :
A Sea of Terms & Acronyms
SDM
CDISC
Version 3
RCRIM
ADaM
ODM
JANUS
WEBSDM
eCTD
MeDRA
PPV
ICH
HIPaa
XML
HL7
AERS
SNOMED
911
21 CFR 11
Standards Development
• ICH E3 – study description
• eCTD – electronic Common Technical
Document (submission guidance, Booze-Allen
viewer tools)
• Submission Data Standards (CDISC Version 3,
PPV Pilot, Lincoln Tech CRADAs –PPV and
WEBSDM HL7/RCRIM, JANUS)
• Protocols
– ICH E6, E9 and E3
– Statistical analysis plan
• Consistency – impossible task?
I Contend That …
• It is important that we (statisticians, the
industry, clinical trialists, regulators,
researchers) need to agree on and use
standards for describing an analysis
database and analyses -- for a study and
for combined studies.
• This is true for Science and regulation
• We need a guidance.
• We need to use the guidance to document
and communicate the scientific decisions
we make in analyzing data
eCTD Draft Guidance: August 2003
Guidance for Industry
Providing Regulatory Submissions in
Electronic Format — Human
Pharmaceutical Product Applications and
Related Submissions
DRAFT GUIDANCE
This guidance document is being distributed for comment purposes only.
Comments and suggestions regarding this draft document should be submitted within
60 days of publication in the Federal Register of the notice announcing the availability
of the draft guidance. Submit comments to Dockets Management Branch (HFA-305),
Food and Drug Administration, 5630 Fishers Lane, Rm. 1061, Rockville, MD 20852.
All comments should be identified with the docket number listed in the notice of
availability that publishes in the Federal Register.
For questions regarding this draft document contact (CDER) Randy Levin 301-594-5411,
or (CBER) Robert Yetter at 301-827-0373.
eCTD Draft Guidance:
Finding a “Home” for Analysis Files
1. Study reports
Typically, a single document should be provided for each study report
included in this module. However, if you provide the study reports
as multiple documents, you should confine the subject matter of
each document to a single item in the following list…
Individual subject data listings …
-- Analysis datasets
- Analysis datasets
- Analysis programs
- Data definitions
[my emphasis]
Developing Guidance for Analysis
Databases: Some Challenges
• Keeping Science alive – complexity (The
Battle of the Comments)
• Improving the Science (data quality; missing
data – “gets hard real fast” – O’Neill, 2003
FDA/ASA Biopharm Workshop; safety and
efficacy
• Efficient, unambiguous communication
(recent experience)
• Everybody already knows what to do
• Include health research enterprise and new
technologies
Regulatory Environment
Regulation
Guidance
S. Woollen
Guidance Begets Guidance:
Be Careful of the Consequences
• Providing Regulatory Submissions in
Electronic Format — NDAs
• Provided advice on submitting “CRT” (case
report tabulation) data
• Details on safety data and demographics
• “lists contain suggested data elements for the
individual datasets. This listing is meant to serve
as a starting point for discussion between you
and the review division on the content and
organization of the datasets”
• Our advice led to confusion
Developing Guidance
• Best advice – reflect best practice
• Pay attention to current practice &
Science (meetings with sponsors)
• Talk to “customers” – series of meetings
• Ask experts – Dave Christiansen, SGE
• Coordinate with other guidances – HL7
• We need to do it – eStat Group
• Provide draft for comments
• Not regulation – comment: very fuzzy
• New urgency – guidance on submission
of analysis files
Did Get People Together for the First
Time to Talk Standard: CDISC
• Workgroups
–
–
–
–
SDM (submission data)
ODM (operational data – XML
ADaM (analysis data)
Lab (laboratory data)
• Challenge – coordinate efforts
• ADaM – doing the hard work
• We (FDA) are “observers”
ADaM focus – FDA Statisticians
• “provide guidelines for the creation of
files and associated documentation that
are submitted to the FDA statistical
reviewer in support of the primary and
important secondary study objectives”
• “statistical analysis data models represent
consensus across a large number of
reviewers experienced in regulatory
review”
• A big jump start for us – re. guidance
ADaM Models: One PROC Away
• “Statistical Analysis Dataset Models”
–
–
–
–
Change-from-baseline
Time-to-event
Categorical
Safety (under construction)
• Detailed examples
– Detailed description of data files and
corresponding metadata (SDM)
– Example based on real endpoint
– Display tables
– Test data
ADaM Models:
The Power of Standards
Are We There Yet?
More Guidance
• Statistical Analysis Plans
• E3 – Appendix 16.1.9 Documentation of
statistical methods
• Protocols
THANK YOU
See You Next Year!
[email protected]
301-827-5583