JanusCRIX Clinical Data Repository CDR

1
Janus/CRIXClinical Data Repository (CDR)

• Project Overview for CRIX Executive Group
• June 30, 2005

2
Agenda

• Background (Review)
• Goals and Objectives
• The Vision
• What is the Janus/CRIX CDR?
• Pilot Project Overview
• Project Approach
• Project Activities and Deliverables
  (In-Scope/Out-of-Scope)
• High Level Timeline and Deliverables
• Project Organization, Roles and Responsibilities
• Working Groups Today

3
Purpose and Expected Outcomes

• The purpose of this Janus Project Overview is to
• Provide project introduction and outline overall
  goals of this project
• Discuss project team member roles and
  responsibilities
• Review high-level project approach key tasks,
  deliverables and timing
• Start defining requirements
• The expected outcomes are
• Facilitate communication
• Common understanding of processes that will be
  followed to achieve project success
• Shared expectations regarding projects goals and
  and commitment to achieving the overall goals of
  this project
• Initial list of use cases and requirements

4
CRIX Objectives (Reminder)

• More efficient submission based on standards
• Reproducible, custom datasets for analysis
• Reusable tools for analysis and review
• Less data redundancy
• Less time on orientation to data by reviewers
• Less ambiguity in communications of information
• More auditable data
• Less manual, paper processing
• Use of common standards across the entire
  community (government, industry, academia)
• Interoperability with other caBIG data sets,
  tools, capabilities

NO MORE PAPER!
5
CRIX Vision (Reminder)

• An infrastructure that will
• allow secure transmission of clinical research
  information between sponsors, researchers, and
  regulatory authorities
• facilitate the adoption of electronic data
  standards, standardized terminologies,
  e-transactions, and e-submissions
• reduce the overall cost of existing information
  gathering and submissions development processes
  as well as review and analysis of information
  and
• be accessible to all.

6
CRIX Development Projects

• Firebird
• Global investigator registry for commonly
  used/referenced data
• SAFE
• Legally enforceable digital signatures compliant
  with Title 21 Regulations
• CRIX CDR (Clinical Data Repository)
• Sustainable, secure and standards-based
  repository for electronic IND submissions of data
  and documents
• Tools for reviews, reporting and analysis
• Electronic IND Submissions
• Sustainable, secure and standards-based
  infrastructure for electronic submissions of data
  and documents

7
CRIX Development Plan
Electronic Submissions
CRIX CDR

Firebird/SAFE
ProductionInfrastructure
Firebird/SAFE
Firebird/SAFE
operationalfeasibility
technicalfeasibility

CRIX CDR
CRIX CDR
Infrastructure Development
Electronic Submissions
8
CRIX 2.0 Clinical Data Repository (CRIX CDR)

• Partnering with industry to
• Implement a prototype a Clinical Data Repository
  that is standards-based and caBIGTM compliant
• Develop a data import facility that accepts data
  in standard formats
• Provide a data services layer that supports
  collaborative research for multi-site studies on
  a global basis
• Provide tools for review, analysis and reporting

9
Janus/CRIX Scope

• Whats in
• All clinical trial data
• Protocol
• Analysis plan
• Animal tox data

• Whats out
• Spontaneous reports
• Documents
• Reviews

Pilot includes implementation of existing Janus
data model defined under IBM CRADA with FDA only
10
Janus/CRIX Functional Building Blocks

• Data what happened
• Findings
• Events
• Domains, attributes
• Interventions
• Protocol what was supposed to happen
• Trial structure (arms, visits)
• Planned assessments
• Like actual findings, but no result
• Planned interventions
• Like actual interventions
• Analysis plans and results
• Analysis datasets (query rule)
• Analytic plans
• Analytic results

11
Out of Scope for Janus/CRIX PilotPotential
Extensions

• Implement the data services layer for the entire
  Janus/CRIX model
• Implement a fully validated, complete Janus/CRIX
  Model that is compliant with CFR 21 Part 11
• Conduct performance analysis of Janus/CRIX
  repository to project expected production scale
  system requirements and identify any performance
  constraints
• Extend the Janus/CRIX data model so as to support
  data now gathered in the CDUS and CTMS
  applications managed by NCIs Cancer Therapy
  Evaluation Program (CTEP) as well as for SAEs,
  AdAM, and 1571/1572 form information

12
CRIX CDR - Data Access Vision
Tools for Review, Analysis, Reporting

Ad Hoc Query Tool
FDA Review
SAS
I-Review
Spread- sheet Tool
WebSDM
ToxVision
Study Specific Data Marts
Data Access Layer (CDISC/HL7)
CRIX CDR
Master Index
Janus Extensions
Base Janus Model
13
CRIX CDR Collaborative Research Environment
Vision
Local DBs
NCI
Clinical Center
Biopharma
Virtual Database
Local DBs
Local DBs
CRO
Clinical Center
Local DBs
Local DBs

• Grid/Web Services
• Security Services (CSM)
• Semantic Services (caDSR)
• Vocabulary Services (EVS)
• Data Services (caBIO)
• etc.

14
CRIX 3.0 Electronic Submissions

• Develop and deploy a secure, electronic
  infrastructure for facilitating submissions to
  the FDA including
• Documents unstructured text or non-standard,
  special data sets submitted via
• ICH Electronic Common Technical Document (eCTD)
• HL7 Regulated Product Submission (RPL) message -
  under development
• Data Standard, structured data submitted via
• ODM and SDTM Define.XML
• HL7 v3 Electronic Data Collection Instrument
  (eDCI) message under development
• Proposal for submission of standardized
  electronic study data from clinical studies
  evaluating human drugs and biologics would revise
  FDA regulations to require clinical study data to
  be provided in electronic format and require the
  use of standard data structure, terminology and
  code sets Federal Register 69 FR 73119

15
Putting it All Together
Documents (Unstructured Text Non-Standard Data
Sets)
eCTD or HL7 RPL XML
Annual Reports Investigator Brochures etc.
Secure Transport
Secure Transport
Message Exchange Services (ebXML)
Message Exchange Services (ebXML)
Structured Data (Standard Data Sets)
Secure Transport
Structured Protocols Findings Events Interventions
etc.
ODM and SDTM Define.XML or HL7 eDCI XML
CRIX Clinical Data Repository
Secure Transport
Data Load API
Government, Academic or Industry Site
16
Janus/CRIX Project Organization
Executive Steering Group (IOTF Informatics Subcomm
ittee)
Relevant caBIG Workspaces and SIGs (CTMS, DSIC,
V/CDE, Arch)
FDA Lead (R. Levin)
Program Lead (S. Dubman)
CRIX Exec Steering Group
BRIDG/CTOM (C. Andonyadis, Scenpro)
ebXML, HL7 SDK (S. Liu, SAIC)
CDUS/CTMS Extensions Analysis (CTIS)
Technical Support (Terpsys)
Analytical Tools (SAS, Lincoln, Pharmquest, I-Revi
ew, etc.)
Janus Pilot (IBM)
17
Janus/CRIX CDRPilot Project Overview

• Bill Tulskie, IBM

18
Development Team Assignments (Leads and
Staff/SMEs)

• Project Manager Ginny Wagner
• Architect Bill Tulskie
• Requirements Team
• Lead Kathy Martin
• Staff/SME Elise Blaese
• Data Load Data Warehouse Team
• Lead Doug Del Prete
• Staff/SMEs Joyce Hernandez
• Data Access Team
• Lead Joel Dubbels
• Staff/SMEs Bridget Meyer
• CSM Team
• Lead Gary Sullivan
• Staff/SMEs Jeff Uehling

19
Janus/CRIX CDR Vision
Analysis and Reporting Applications
Analytical and Reporting Applications (SAS,
WebSDM, ToxViewer, I-Review, etc.)
Query Application
BIOGopher (Excel with caBIO Driver)
caBIG
FDA Submissions Source
caGRID
Study Specific Data Marts
C S M

JDBC 2
caBIO/BRIDG/HL7 RIM
Potential Extensions 1571/1572 form data,
CDUS/CTMS, AdAM, SAEs, etc.
ODM and SDTM Define.XML (HL7 eDCI)
Data Load API
Janus data extensions
Janus/CRIX Clinical Data Repository
caDSR
EVS
20
Janus/CRIX CDR Vision
Analysis and Reporting Applications
Analytical and Reporting Applications (SAS,
WebSDM, ToxViewer, I-Review, etc.)
Query Application
BIOGopher (Excel with caBIO Driver)
caBIG
FDA Submissions Source
caGRID
Team 1
Study Specific Data Marts
Team 3
Team 4

C S M
JDBC 2
caBIO/BRIDG/HL7 RIM
Team 2
Potential Extensions 1571/1572 form data,
CDUS/CTMS, AdAM, SAEs, etc.
ODM and SDTM Define.XML (HL7 eDCI)
Data Load API
Janus data extensions
Janus/CRIX Clinical Data Repository
Team 3
caDSR
EVS
21
Project Plan Overview
22
Janus/CRIX User Requirements

• Kathleen Martin/Elise Blaese

23
JANUS/CRIX User Requirements

• Scope of Responsibilities Phase Ia
• Understand how Janus data warehouse users will
  access NDA submissions data, based on roles. IBM
  will compose 10 use cases.
• Develop user requirements workshop/interview
  guides
• Conduct workshops/interviews with FDA and
  NCI/caBIG staff to confirm our current
  understanding of the intended usage.
• Develop use cases for accessing data in the Janus
  data warehouse.

24
JANUS/CRIX User Requirements

• Project Plan Phase 1a
• Develop Interview Guides (input from technical
  team members and NCI)
• Conduct requirements interviews/workshops and
  develop usage test cases. Plans to date include
• CRIX Exec Committee - conducting 3 parallel
  workshops Design, Business, and Functional
  Requirements (June 30, DC)
• FDA audience (July 14, Rockville)
• Internal NIH/NCI audience (July 6 and July 8)
• Cancer Centers (July 29, Chicago)
• Conduct post-interview debriefs and follow up
• Develop Requirements Analysis

25
JANUS/CRIX User Requirements

• Deliverables
• Workshop/interview guides
• Workshop output/interview summaries
• User requirements analysis
• Use test cases

26
JANUS/CRIX User Requirements

• Identify User Roles
• FDA Roles
• Medical Officers
• Pharmacologist
• Biostatistician
• Data Administrator
• etc.
• NCI/Cancer Center Roles
• Principle Investigator/Investigator
• Study Coordinator
• Data Manager
• Biostatistician
• etc.

27
JANUS/CRIX User Requirements

• User Interviews / Workshops Overview
• Project background
• Interview / Workshop subject areas
• Roles within categories (researchers, clinicians,
  reviewers, and data management)
• Types of data accessed
• Raw and summary data
• Within studies/across studies,
• Queries, reports, etc.
• Volume of users
• Current tools used and features/functionality/purp
  ose
• Data metrics for large, medium, and small studies
  (role-specific)

28
JANUS/CRIX User Requirements

• Dependencies
• Confirmation of roles and finalize interview
  guide
• Availability of interviewees and workshop
  attendees
• Identification of additional roles
• Identifying/providing reference material or
  background information

29
Data Access

• Joel Dubbels
• Bridget Meyer

30
Data Access Overview (development primarily
Phase 1B timeframe)

• Scope definition
• Develop Java program APIs enabling application
  access to Janus database (a subset of the overall
  model)
• APIs will be developed using NCIs caCORE Toolkit
  V3
• Development to caBIG silver level compatibility
  guidelines

31
Janus/CRIX CDR Vision
Analysis and Reporting Applications
Analytical and Reporting Applications (SAS,
WebSDM, ToxViewer, I-Review, etc.)
Query Application
BIOGopher (Excel with caBIO Driver)
caBIG
FDA Submissions Source
caGRID
Study Specific Data Marts
Validate caCORE toolkit functionality /
compatibility
C S M

JDBC 2
caBIO/BRIDG/HL7 RIM
Potential Extensions 1571/1572 form data,
CDUS/CTMS, AdAM, SAEs, etc.
ODM and SDTM Define.XML (HL7 eDCI)
Understand/map to existing NCI EVS and CDE
services as appropriate
Data Load API
Janus data extensions
Janus/CRIX Clinical Data Repository
caDSR
EVS
32
Data Access Overview (Phase 1B timeframe)

• Dependencies
• Development of use case for each user role
• Definition of Janus/CRIX model subsets
• Definition of data access requirements over the
  data subsets
• Technical assistance from NCI for EVS and CDE
  work
• Need NCI contact person to help resolve
  functional issues encountered
• Need NCI assistance with running semantic
  connector to establish appropriate correlation to
  existing caCORE EVS and CDE

33
Data Access Overview (Phase 1B timeframe)

• Key milestones
• Successful db access using one object built via
  IBM UML modeling tool and caCORE Toolkit
• Fully defined logical model (UML) over the
  agreed-to subset of the Janus/CRIX model
• Successful processing of UML model through caCORE
  semantic connector
• Successful code generation from UML model via
  caCORE toolkit build process
• Successful data access via caCORE interface

34
Janus/CRIXData Load and Warehouse

• Doug Del Prete

35
JANUS/CRIX Data Load and Warehouse

• Scope
• Given specific study data and set of domains
• Load the data from the identified study/domains
  into the JANUS/CRIX data warehouse,
• Encrypt and support the identified data access
  under the defined CSM user access controls

36
Janus/CRIX CDR Vision
Analysis and Reporting Applications
Analytical and Reporting Applications (SAS,
WebSDM, ToxViewer, I-Review, etc.)
Query Application
BIOGopher (Excel with caBIO Driver)
caBIG
FDA Submissions Source
caGRID

• Source data
• Study/Domains
• Data transport format
• (XML preferred)
• define.xml to populate Codelists

Study Specific Data Marts

C S M
JDBC 2
caBIO/BRIDG/HL7 RIM
Potential Extensions 1571/1572 form data,
CDUS/CTMS, AdAM, SAEs, etc.
ODM and SDTM Define.XML (HL7 eDCI)
caDSR
EVS
37
JANUS/CRIX Data Load and Warehouse

• Plan
• Phase 1A
• Build a prototype system including JANUS data
  model and sample data (with encryption)
• Phase 1A/B Deliver System to NCI
• Get identified study/domain data from NCI
• Develop and deliver Janus/SDTM V1.0 or V1.1
  education
• Assist NCI to map their clinical trial data to
  SDTM (based on the final outcome from Phase 1A)

38
JANUS/CRIX Data Load and Warehouse

• Plan (continued)
• Phase 1B
• Develop final Janus-SDTM data element mapping
• Load initial data set
• Verify all data loaded correctly (joint with NCI)
• Encrypt selected portions of the data warehouse
• NCI to test with JANUS data warehouse and related
  access tools
• Implement user access control

39
JANUS/CRIX Data Load and Warehouse

• Plan (continued)
• Phase 1C
• Adjust the Janus Data Warehouse based on NCI
  feedback
• Load final data set and encrypt
• Exercise test plans to verify correct data load
• Conduct basic load testing (as possible)
• Develop recommendations for production system
• Other data formats and sources
• System configuration and sizing
• Final Report

40
JANUS/CRIX Data Load and Warehouse

• Dependencies
• Study/domain data
• Consultation from NCI on the data mappings and
  data encryption
• NCI to evaluate pilot implementation and provide
  feedback for Phase 1C

41
CSM Overview

• Garry Sullivan

42
Janus/CRIX CDR Vision
Analysis and Reporting Applications
Analytical and Reporting Applications (SAS,
WebSDM, ToxViewer, I-Review, etc.)
Query Application
BIOGopher (Excel with caBIO Driver)
caBIG
FDA Submissions Source
caGRID
Study Specific Data Marts

C S M
JDBC 2
caBIO/BRIDG/HL7 RIM
Potential Extensions 1571/1572 form data,
CDUS/CTMS, AdAM, SAEs, etc.
ODM and SDTM Define.XML (HL7 eDCI)
caDSR
EVS
43
Common Security Module Overview (Phase 1B
timeframe)

• Scope
• Application Authentication will be done through
  AuthenticationManager. (Choices RDBMS or LDAP)
• Authorization checks will be provided through the
  AuthorizationManager run-time methods.

44
Common Security Module Overview

• Customer dependencies
• Definition of roles and data access requirements
  over the subset Janus/CRIX model for each user
  type
• Subset Janus/CRIX model populated with test data
• Development of a logical use case scenario for
  each user type
• Implementation of use case scenarios

45
Common Security Module Overview

• Key milestones
• Install and configure CSM
• Application sign-in and authentication
• Successful data access for each user role per
  assigned authority
• Reject unauthorized data access

46
For use by NCI and IBM team, FDA, and other
interested parties . . .

• Being used for initial project tracking, exchange
  of reference materials and discussion, meeting
  info, project calendar, etc.
• All artifacts will migrate to caBIG CVS
  repository as they are completed

47
Secure login based on IBM registered ID
All artifacts will migrate to the caBIG CVS
repository as they are completed
48
Recap

• Discuss issues/concerns
• Document meeting action items
• Answer questions

Questions ??