Transcript Clinical Trial Management & Part 11 Compliance

Emerging Data File
Initiatives & Standards:
Potential Impact on Global
Clinical Studies
Co-Presenters
Ginger Clasby
Don Hurst
Executive VP
Business Development
Director
Clinical Information Technology
[email protected]
[email protected]
Corporate Goal
Optimize revenue stream by getting
products to market as quickly as possible
Accomplishing the Goal
Enroll subjects as quickly as possible
•
Sites in US & OUS
•
Experienced sites
•
Sophisticated px recruitment programs
Accomplishing the Goal
Maximize clinical trial operational
efficiencies to minimize time from last
data in to database lock
•
Standardization
•
Internet applications for study management docs
•
Electronic Data Capture (EDC)
•
Detailed monitoring of subject exam schedules
Accomplishing the Goal
Minimize time for statistical analysis &
CSR development
• Standardization
• eXtensible Markup Language (XML)
• Develop/validate templates prior to database
lock
Accomplishing the Goal
Minimize regulatory agency review
period
•
Standardization
•
Electronic submissions
•
XML
Oops!
Differing data capture & reporting formats
across countries/regions
Differing Formats
• Data conventions: mo/day/yr or day/mo/yr?
• Measurement units: kg or lb?
• Variable names: English, French, German or ?
• Report content/structure varies widely by
country or region
CDISC
Clinical Data Interchange
Standards Consortium
CDISC Background
• Established 1997
• Assumes use of EDC
• Collaboration to produce functional standard
data models facilitating data interchange
between industry stakeholders
• Supports end-to-end data flow within trials,
from source document to regulatory submission
CDISC “Deliverables”
Operational Data Modeling (ODM)
Data acquisition
Data archiving
Submission Data Modeling (SDM)
Operating database
Reg submission
CDISC “Reach”
• Biopharmaceutical study focus
• Active collaboration with FDA
• Active collaboration with analogous
regulatory organizations in Europe &
Japan
CDISC vs. ICH
• ICH – working toward global submission
standards
• CDISC – working on standardization of
submissions at the data level
CDISC vs. HL7
• HL7 – data standards for all health care
operations, including reimbursement &
order processing – no clinical trials focus
• CDISC – oriented to biopharmaceutical
product development only
Three Technological Trends Impacting Clinical Study
and Data Management Design for Global Trials
1.
EDC: environment manipulating the files
A. Hardware characteristics
B. Software & database/file characteristics
2.
XML: a vendor neutral file standard
A. Brief history
B. SGML
C. HTML
D. XML
3.
CDISC
A. ODM – Operational Data Model
B. SDM - Submissions Data Model
Electronic Data Capture
Hardware Characteristics:
1.
Web enabled, wired or wireless
2.
Transfer devices
A. PDA (Palm Pilot) for patient diaries, etc.
B. Blackberry devices (Raspberry devices??)
C. IVRS (Interactive Voice Response Systems)
D. Scanning (CRFs)
E. Keyboard
F. Smart phones
Electronic Data Capture
Hardware Characteristics:
3.
Servers
A. Many different types of servers (authentication,
mail, database, backup, Citrix, etc.)
B. Server to server communication
C. Device to server communication
4.
Communication protocols
A. TCP/IP
B. FTP
C. Bisynchronous
D. Asynchronous
Electronic Data Capture
Software & Database/File Characteristics:
1.
Operating systems
A. Linux
B. Windows 2000
C. Solaris
2.
Front end (Graphical User Interface)
A. HTML
B. JAVA
C. C+ and other languages
Electronic Data Capture
Software & Database/File Characteristics:
3.
Backend database
A. Flat
B. SQL/Server (Relational)
C. UDB/DB2 (Relational)
D. XML (hierarchical)
4.
Client/server type
A. Thin (Explorer based only)
B. Hybrid (Explorer + some client download)
C. Thick (Major client download…usually for
offline followed by scheduled uploads)
Electronic Data Capture
Software & Database/File Characteristics:
5.
Types of files used between devices
A. Only one type of file: flat and sequential
B. Characteristics of flat file are dependent on what
is encoded from the source and how file is
interpreted, or reconstructed, by the target. Or
another way of saying it, is what association is
with the file (icon).
C. A flat file can be many different types:
1) SAS transport, Excel, Word, unloaded
relational, zipped, Adobe PDF, etc.
2) XML file
1.
What is it?
A. eXtensible Markup Language
B. Non-proprietary, platform-independent meta-language
for hierarchically structuring information
2.
What is XML’s history?
A. Subset of the International Standards Organization (ISO)
Standard Generalized Markup Language (SGML), ISO
8879:1986
B. SGML is an ISO Standard - ISO 8879:1986
C. SGML Established Standard for 12 years.
D. SGML released as ISO8879 in 1986
E. At SGML 96 Conference in Boston, XML released by a
working group associated with the W3C.
F. SGML ## was previous name for XML ####
G. XML 1.0 is W3C recommendation (32 pages)
H. XML became recommendation in February, 1998
XML
3.
Where is it used?
A. Almost all industry groups
B. Heavily used in the banking sector
4.
How is XML related to SGML and HTML?
A. XML + HTML = SGML
B. SGML – Standard Generalized Markup Language
1) Context (HTML)
2) Content (XML)
C. HTML - HyperText Markup Language
1) Used in Web pages
2) Can emphasize text, graphics, links, etc.
D. XML
1) Hierarchical in organization
2) Uses data tags to classify data
XML
XML
5. What does XML look like (normal view not indented)?
6. Things to remember about XML
XML
A. Already an international standard
B. An ISO standard
C. Standard for the electronic exchange of
information in most industry groups
D. Although a flat and sequential file, it has
structure
E. “Content” based, while HTML is “context”
based
CDISC
1. Clinical Data Interchange Standards Consortium
2. Developing a common interchange standard for
clinical data
3. Accomplish through the development of metadata models
CDISC
4. Benefits of standard meta-data models:
A. Clin. labs do not have to support different formats
B. CROs no longer have to develop completely new
databases to meet unique specifications of each
trial
C. EDC providers do not need to customize exchange
format for each new client
D. Sponsors can receive data in standard format in less
time
E. Regulatory reviewers can more quickly and easily
review submission data
CDISC
CDISC
5. Current Meta-Data models
A. ODM – Operational Data Model
B. SDM - Submissions Data Model
C. RIM - Reference Information Model
D. ADaM – Analysis Dataset Model
E. LAB – Laboratory Standards (LOINS)
CDISC
6. ODM – Operational Data Model Goals
A. Develop vendor independent models for
interchange and archive of clinical data using
metadata
B. Base models on XML technology
CDISC
7. ODM – Overview Structure
CDISC
8. SDM – Submissions Data Model (Domains)
• Demographics
• Concomitant Medications
• Disposition
• Adverse Events
• Exposure
• Vitals (Horizontal)
• Labs (Chemistry, Hematology,
Urinalysis)
• Vitals (Vertical)
• Physical Exam
• Medical History
• ECG (Horizontal)
• EKG (Vertical)
CDISC
9. Other data models receiving top priority
A. Pharmacogenomics
B. Microbiology
Summary (and peek into the future)
1. Three trends are merging to bring a drug/device
to regulatory review in a faster, more standardized
package.
2. The merging of the trends will help lower one of
the major costs of trials and remove one of the
traditional bottlenecks prior to submission.
3. The merging trends have gained such momentum
that the SAS Institute has invested and created a
new procedure called “PROC CDISC”.