PhUSE

10-12 September 2005

Heidelberg, Germany

An Introduction to CDISC: Part One

“What is CDISC and Why Standards?”

Rebecca Kush, PhD, CDISC

David Handelsman, SAS
 Lack of standards make our
lives more complicated
• Electrical power

Standards make our lives easier

• ATMs
• Bancomat
• Geldautomat
Clinical Information Dispersion Today
• Healthcare information is found in:
– paper medical records
– disparate databases
– hospital-based information systems

• Clinical research data exists in:

– additional databases
– research notebooks

• Clinical trial data collection:

– 3-part NCR forms in ~85% of trials
– multitude of electronic data capture applications
• eCRF
• IVRS
• ePRO
 The Move to Standards
“Faced with rapid changes, the nation’s healthcare system has fallen short
of its ability to translate information into knowledge that can be used
in practice, and to apply new technology safely and appropriately.
The results are exactly what you would expect. Everyone who uses the
current system constantly confronts large information gaps, whether
it’s at the doctor’s office, on the hospital ward or at government agencies
charged with protecting the public health. That goes for the FDA -- we’re
no exception.”
Dr. Mark McClellan, former FDA Commissioner
CDISC Interchange, October 2003

“Innovation depends upon standardization.”

Dr. Bob O’Neill, Director, Office of Biostatistics, CDER, FDA
Benefits of Standards: Operational
• Increase efficiencies of performing clinical trials
• Simplify processes among investigators, pharma
companies, CROs, vendors, laboratories
• Streamline data collection at investigator sites
• Improve sponsor’s access to data and ability to
understand it
• Provide long-term means for electronic data
archive
• Facilitate FDA review of submissions
• Improve links between healthcare delivery and
clinical research
Benefits of Standards: Scientific
• Facilitate integration of data
– Across trials, compounds, companies
• Enable assessment of trends and patterns in data
– Public health benefit
– Drug development benefit
– Better guidance for studying disease states
• Allow companies to focus on their potential
therapeutic products vs. data issues
• Improve safety monitoring and pharmacovigilance
• Support ad hoc analyses; enhance decision-
making capabilities during data review
Standards: Yours, Mine, Ours
• Internal standards are a good first step,
but are insufficient
– Streamline internal processes
– Little effect of data transfer and interoperability with external
organizations
– CROs are a classic example of the limits of internal
standards

• Development and adoption of public

standards are the next logical steps
– Increase data efficiency of all participating organizations
– Brings a renewed focus to the science of clinical research
 Why Are Standards Important?

Regulatory
IND Annual ISS, ISE, etc. NDA Reporting

Regulatory
Pre-Clinical Clinical Development Post Marketing
Submission
Phase I Phase II Phase III

Early
Discovery
Phase IV

Animal 30-100 200-400 3000+

Studies Volunteers Patients Patients

Up to 10,000 10 - 15 4-8 2-3 1

Approval 10-15 Years
# of Compounds elapsed time
 Why Are Standards Important?
Before Standards
Internal Internal Regulatory
EDC CDM Data
EDC Internal Internal Agencies
EDC System Warehouse
EDC CDM Data
EDC EDC System
Internal
Warehouse
Internal Data
(SDS)
CDM Internal Warehouse
Internal Corporate
EDC
System CDM
System` Data Partners
EDC System Warehouse
EDC Internal Statistical
EDC (ODM) Internal
CDMInternal Programs
DataInternal
Paper
EDC EDC EDC System CDM Warehouse Data
System
Warehouse
EDC EDC
EDC
 After CDISC Standards
EDC Regulatory
(ODM) Agencies

Internal
CDM Internal Corporate
System Data Partners
Warehouse
(SDS)
Statistical
Programs
 Looking into the future
Regulatory
eSource Agencies
Data

Corporate
Internal Partners
Data
Warehouse
Statistical
Programs

…what other options exist?

 Benefits of Standards –
Regulatory View
Before SDTM After SDTM
• Domains = Yes • Domains = Yes
• Standard Domain Names = No • Standard Domain Names = Yes
• Standard Structure = No • Standard Structure = Yes
• Standard Variables = No • Standard Variables = Yes
• Standard Variable Names = No • Standard Variable Names = Yes
• Standard Terms = No • Standard Terms = not yet, but
coming
 Benefits of Standards –
Regulatory View
“The importance of a standard for the
exchange of clinical trial data cannot be
overstated. FDA reviewers spend far too
much valuable time simply reorganizing
large amounts of data submitted in varying
formats. Having the data presented in a
standard structure will improve FDA’s ability
to evaluate the data and help speed new
discoveries to the public.” -Lester Crawford,
Former Commissioner, FDA
 Clinical Data Interchange Standards
Consortium: Original Mission Statement
CDISC is an open, multidisciplinary, non-profit
organization committed to the development of
worldwide industry standards to support the
electronic acquisition, exchange, submission and
archiving of clinical trials data and metadata for
medical and biopharmaceutical product
development.
The mission of CDISC is to develop and support global,
platform-independent data standards that enable
information system interoperability to improve
medical research and related areas of healthcare.
 Current State of Clinical Data
Transfer
Medical
Records

CRF
C Operational
Data

Operational
Data Submission

Submission
Data

Lab
Clinical Development Cycle ‘Tools’ Drug Approval
Data Archiving
Trial Sponsor; Sites
Regulatory Submission
Research Hypothesis Trial Sponsor to Agency
PI or Trial Sponsor
SAS

Protocol Development
PI or Trial Sponsor
Protocol Document
Authoring Simulation
Management
Document
Document Data Analysis Management
Tracking Tool Reporting of Results
Site/Trial Preparation Trial Sponsor
Sites;Trial Sponsor; IRB SAS
Internet Proj.Mgmt.
CTMS Visualization
Subject/Site
Tools
Tracking/Mgmnt
Call Centers Trial Management
Trial Sponsor; Sites
Subject Enrollment
Sites SAE
Internet Clinical Data Tracking
Management
System
Internet Data Collection,
Monitoring, Processing Query
Trial Sponsor; Sites Resolution

EDC
Clinical Back Office

1 6

5
2
3 4

EDC Without Standards, courtesy

Charles Jaffe, MD, PhD
The Future: Standards to Facilitate Data Flow
.from Source to Reviewers
Sources of electronic data

ECG
eSubmission for
LAB Regulatory Review

ODM Operational
SDS
LAB Database ADaM

EDC
CRO
 CDISC
• Formed in 1997 as a volunteer group
• As of 2000, incorporated and funded as non-
profit organization
• Now supported by >170 member companies:
– pharmaceutical companies; biotech companies;
CROs/service providers; technology providers
• CDISC Groups now growing in Japan, Europe,
India
• Standards developed through consensus-based
approach with public reviews; open standards
 CDISC Principles
• Lead the development of standard data models
that improve process efficiency while supporting the
scientific nature of clinical research.
• Recognize the ultimate goal of creating regulatory
submissions that allow for flexibility in scientific
content and are easily interpreted, understood, and
navigated by regulatory reviewers.
• Acknowledge that the data content, structure and
quality of the standard data models are of
paramount importance, independent of
implementation strategy and platform.
 CDISC Principles

• Maintain a global, multidisciplinary, cross-

functional composition for CDISC and its working
groups.
• Work with other professional groups to encourage
that there is maximum sharing of information and
minimum duplication of efforts.
• Provide educational programs on CDISC
standards, models, values and benefits.
• Accomplish the CDISC goals and mission without
promoting any individual vendor or organization.
CDISC Structure CDISC
Board Committees
Board of Directors •Financial Oversight
•Governance
•Nominating
Industry Advisory Board •Scientific
•Technical Support & Guidance

CDISC Working and Alliances

Support Teams •HL7
•Others
•Operational Data Modeling (ODM)
•Submission Data Standards (SDS)
•Analysis Data Set Modeling (ADaM) CDISC Regional Groups
•Laboratory Data (LAB)
•Protocol Representation •Regional CDISC Coordinating Committees (r3C)
•SEND •Regional CDISC Groups
•Terminology r = regions, including Europe, Japan, India

Operations and Infrastructure (OIS)

•Education •Technical Direction; TCC

•Membership Services •Implementation Group Coordination
•PR/Communications •Standards Maintenance
•Operations and Financial Management •Glossary Group Leadership
 Standards
• Definition of Standard: An object
considered by an authority or by general
consent as a basis of comparison; an
approved model*
• Standard data models, metadata
models, codes, terminology, content,
technology .
• Definition of Model: A standard or
example for imitation or comparison*
* The Random House Dictionary
of the English Language, unabridged version
 Data and Metadata: Definitions
• Data
– representations of facts, concepts, or instructions in a
manner suitable for communication, interpretation, or
processing by humans or by automated means. [FDA]
(Synonym: Information)
• Metadata
– Data about data
• Metadata Elements
– The metadata of a study describes the types of study
events, forms, item groups, and items that are allowed
in the study.
 How Important is Metadata?

Event Time D S F
Begin 9/23/99 02:01:00 121,900,000 12,300 143.878
End 9/23/99 02:17:23 9,840

Event Time D S F
Start 19990923 05:01:00 196,200,000 5.5 640
Finish 19990923 05:17:23 4.4

Slide developed by David Christiansen, DrPH

 In this case
$125,000,000:
Mars Climate Orbiter

Mars Orbit M/D/Y HH:MM:SS PDT Distance (miles) Speed Force

Insertion Burn (Earth Receive Time, 10 (miles/hr) (Pounds)
min. 49 sec. Delay)
Begin 9/23/99 02:01:00 121,900,000 12,300 143.878
End 9/23/99 02:17:23 9,840

Mars Orbit YYYYMMDD EDT Distance (km) Speed Force

Insertion Burn (Earth Receive Time, 10 (km/sec) (Newtons)
min. 49 sec. Delay)
Start 19990923 05:01:00 196,200,000 5.5 640
Finish 19990923 05:17:23 4.4

Slide developed by David Christiansen, DrPH

 Data Interchange: Definition
• Data Interchange
– the transfer of information between two or
more parties. Data interchange is
distinguished from data transfer in that the
integrity of the contents of the data is
maintained. [CDISC]
• CDISC Glossary is available on the
website (www.cdisc.org) and in December
resource issues of Applied Clinical Trials.
 CDISC Approach
• The CDISC models are the products of
contributions from numerous
organizations, functional groups, and
individuals; they do not have a sole
source.
• Consensus building
– Involves different disciplines within the industry
– Involves ‘consolidating’ existing models, review comments and
testing
– Takes time, but results in widely accepted models
 CDISC Standards Development Process
(COP-001)
Stage I: Standard Definition/Team Initiation
Need for
Proposal Team Leader ID Working Plan
Specific Review per
to Board Approved And Team (timelines, deliverables
Standard(s) strategy, budget
of Directors Formation communication mech.,
Identified priorities
(via OIS) (multidisciplinary) resources req’d)
(any stakeholder) (OIS) (Team )
Not Approved

Stage II: Standards Development/Review/V 1.0 Release

Testing
Consensus Harmon- Comments addressed Released
Review (Production)
(Initial) ized External Version
TCC Version Public
Version Version Focused Review 1.0
Review
Review
Comments to address by team

Stage III: Education & Support

Respond Educational
To Comments Programs
And Questions (EDU, OIS)

Stage IV: Standards Update & Maintenance

TCC
Annual Review of Working Plan Review New
Released Version (timelines, deliverables, Consensus Harmon- Public Released
(comments, chg communication mech., (Revised) Optional ized Review (Production)
reqsts, tests, plans) resources req’d) Version Version
as
needed
Version
(Team) (Team) Ex
Focused
Review Note: Occasional bug fix releases may be
issued as needed with team review only.
 Scope of CDISC Models

Data Sources
Submission
Operational Data
• Site CRFs Operational Submission
•Laboratories Data Database Data
•Contract Interchange Interchange •CRT/Domain
Research & Archive: •Study Data & Archive: Datasets
Organizations ODM, LAB •Audit Trail SDM, ADaM •Analysis
•Development •Metadata Datasets
Partners •Metadata

ODM = Operational Data Model SDM = Submission Data Model

LAB = Laboratory Data Model ADaM = Analysis Data Models
 Evolution of CDISC Standards:
Started with the end in mind

Data Sources
Submission
Operational Submission Data
• Site CRFs Operational
Protocol

Database Data
•Laboratories Data
Interchange •CRT/Domain
•Contract Interchange
& Archive: Datasets
Research & Archive: •Study Data
SDS, ADaM •Analysis
Organizations ODM, LAB •Audit Trail
SEND Datasets
•Development •Metadata
Partners •Metadata

ODM = Operational Data Model/Std SDS = Submission Domain Standards

LAB = Laboratory Data Model/Std ADaM = Analysis Data Models
SEND = Standards for the Exchange of Non-Clinical Data
 CDISC Standards: Implementation
(Production) Versions
Standard Description Implementation
Release Version
Study Data Tabulation Content Model - regulatory submission July 2004
Model (SDTM), including of case report tabulations (CRT) and
non-clinical (animal) data
SEND for non-clinical
(animal) data
Operational Data Model XML-based Transport Standard – Mid-2001
(ODM) acquisition, exchange and archive

Clinical Laboratory Model Content Model -exchange of laboratory Mid-2002

(LAB) data between labs and sponsors/CROs

Analysis Dataset Model Content Model - General Considerations Late 2004

Document and Examples of Datasets

Case Report Tabulation Transport Standard -Describes how to Late 2004

(CRT) Data Definition create data description doc (metadata)
for submission in ODM XML standard.
Specification (Define.xml)
Structured Clinical Trial Machine- and Human-readable model In progress
Protocol (SCTP) that enables exchange of protocol
information amongst stakeholders.
FDA has endorsed the standards by including them
as specifications in FDA Guidance.

SDTM and define.xml: Specifications for FDA implementation

of the ICH eCommon Technical Document.
 Speaking of health benefits/opportunities to be
realized from making more effective use of IT .
• “I think that CDISC will be a big part of moving
FDA onto an electronic information architecture
where we can realize all of these opportunities. I
think this will have a profound and positive impact
on our drug review process, allowing us to design
trials that can be less expensive and still tell us
more about the risks and benefits of a new
medical product. And I think that the most
significant and perhaps enduring legacy to your
efforts could be the very immediate and significant
impact it has on improving the lives of patients.”
-Mark McClellan, MD, PhD, FDA Commissioner, September 2003
 CDISC Teams and Projects - 2005
Single
Source

HL7-CDISC Harmonization; BRIDG Model

eSource
Data
Interchange
Metadata – end-to-end consistency
LAB and AE scenarios

eDCI
Define.xml
HL7 V3

Terminology (Codelists)
NIH Roadmap
CV, TB Stds

PRG ODM SEND ADaM

LAB SDS

Maintenance, Member Relations, Education and Implementation Groups, Glossary

 The mission of CDISC is to develop
and support global,
platform-independent data standards
that enable information system
interoperability
to improve medical research and
related areas of healthcare.
Knowing is not enough;
we must apply.
Willing is not enough;
we must do.
- Goethe-

To the gracious supporters who

‘apply’ and ‘do’ .
THANK YOU!
Rebecca Kush
[email protected]
 Information and Contacts
• For standards and information, see www.cdisc.org
• eNewsletters available via e-mail; contact Shirley
Williams [email protected] or sign up on the
CDISC website.
• Technical questions: Julie Evans [email protected]
or Public Discussion Forum
• Education and Membership: Frank Newby
[email protected]
• Rebecca Kush: [email protected]