Dan Shiu

May 10, 2012Dan Shiu

July 28, 2012

Agenda

Overview of major CDISC standards

Common implementation strategies and tools

FDA (CDER) guidance and concerns

Lessons learned

A Glance of CDISC Standards CDISC is a global, open, multidisciplinary, non-profit organization that

has established standards to support the acquisition, exchange, submission and archive of clinical research data and metadata.

PRM CDASH

LAB

SDTM

SEND

ADaM

Controlled Terminology

ODM XML

Define.xml

CDASH – Implementation

CDASH (Clinical Data Acquisition Standards Harmonization) defines basic standards for the collection of clinical trial data

Implement in EDC system to directly map CRF forms/folders/fields to SDTM domains/visits/variables

EDC edit checks can be built based on SDTM checks to catch conformance issues early

CDASH – Domains Adverse Events (AE)

Comments (CO)

Prior and Concomitant Medications (CM)

Demographics (DM)

Disposition (DS)

Drug Accountability (DA)

ECG Test Results (EG)

Exposure (EX)

Inclusion and Exclusion Criteria (IE)

Laboratory Test Results (LB)

Medical History (MH)

Physical Examination (PE)

Protocol Deviations (DV)

Subject Characteristics (SC)

Substance Use (SU)

Vital Signs (VS)

CDASH – SDTM Mapping

Scenario 1: Central ProcessingScenario 2: Local Processing Scenario 3: Central Processing with Secondary Site Assessment of Clinical Significance

CDASH – Medidata Rave Example

CDASH – FDA Guidance “… The use of case report forms that incorporate SDTM-

standard data elements (such as with CDASH-style case report forms) allows for a simplified process for creation of SDTM domains. This approach is preferred to the alternative of collecting data in a non-standard format and then converting to SDTM format after the trial (legacy data conversion). Legacy data conversion is often complex and difficult, and CDER has received submissions in which reviewers have occasionally encountered difficulties while reviewing converted data.”

SDTM – Implementation SDTM (Study Data Tabulation Model) benefits regulatory agency:

Reviewers can be trained to work with the standardized data more effectively with less preparation time

Support FDA‘s efforts to develop a repository for all submitted trial data Support standard review tools to access, manipulate, and view the

source data

To create SDTM from raw data: Standard mapping specifications Maximal macros/utilities to execute mapping Minimal custom programming

To validate SDTM conformance: WebSDM/Janus rules OpenCDISC Validator

To validate SDTM mapping: Independent programming Risk-based QC

SDTM – Integrated System Example: Kinship EXACT

SDTM – Custom-built Tool Example: Standard Mapping Template

SDTM – Custom-built Tool Example: Standard Mapping Template

SDTM – Conformance Validation Using OpenCDISC Validator

SDTM – Risk-based Mapping Validation Using SAS Business Intelligence Tools

SDTM – FDA Guidance Custom Domains: “Prior to creating a custom domain,

sponsors should confirm that the data do not fit an existing domain …”

EPOCH: “Please include the variables EPOCH for every clinical subject-level observation …”

SE Domain: “The Subject Elements domain should be included …”

Column Length/Size: “Lengths/Sizes of columns should not arbitrarily be set to 200. For example, if your USUBJID column has a maximum length of 18 being used throughout the dataset, the USUBJID’s column size should be set to 18, not to 200.”

ADaM – Implementation ADaM (Analysis Data Model) datasets are source data plus

derived data , i.e. ADaM = (SDTM-) + derivation: To support statistical analysis by simplifying the programming steps

necessary to perform an analysis (a.k.a. analysis-ready) To enable traceability from analysis results presented in the CSR

back to the original data elements collected in the CRF and represented in SDTM datasets

To create ADaM from SDTM: Standard mapping specifications Maximal macros/utilities to execute mapping Minimal custom programming

To validate ADaM conformance: CDISC ADaM validation rules OpenCDISC Validator

To validate ADaM derivation: Independent programming Risk-based QC

ADaM – Integrated System Example:Kinship EXACT

ADaM – Custom-built Tool Example: Standard Mapping Template

ADaM – Mapping Specification Challenge for Parameter Value-Level Metadata

PARAM AVAL BASE AVALCAT CRIT CRITFL

LAB XYZ (mg/dl)

Calculate as medianvalue of multiple measures if present

Calculate as average of all pre-baseline measures

CT of: HIGH, NORMAL, LOW

Value is > 120 Y or N

LAB ABC (g) Select maximum value if multiplemeasures are present

Select value immediately prior to 1st dose

CT of: PANIC HIGH, TELEPHONE PANIC, PANIC LOW,

RATIO OF XYZ / ABC

AVAL of LAB XYZ divided by AVAL of LAB ABC

BASE of LAB XYZ divided by BASE of LAB ABC

65<=ratio<=80 Y or blank

From CDISC ADaM document:

In reality you want a dataset like this:

ADaM – Conformance Validation Using OpenCDISC Validator

ADaM – Risk-based Mapping Validation Using SAS Business Intelligence Tools

ADaM – FDA Guidance “… the submission of data in SDTM format does not

eliminate the need for submission of analysis datasets, whether in the ADaM format (preferred) or in an alternative format.”

“Analysis datasets should be derivable from the SDTM datasets, in order to enable traceability …”

“In addition to the variables specified for ADSL …, it is expected that the sponsor will include multiple additional variables representing various important baseline patient characteristics.”

“It is often useful for sponsors to provide their programsused to create ADaM datasets to help clarify how exactly they were created.”

Controlled Terminology –Implementation

Data values for CDISC standards-specified variables should use the CDISC Controlled Terminology, which is maintained and distributed as part of NCI Thesaurus

CDISC Controlled Terminology standards have been developed from CDASH, SDTM, SEND, and ADaM

Implementation must be built into EDC, SDTM/ADaM mapping, and validation tool

Controlled Terminology – Load to Custom-built Mapping Template

Controlled Terminology – Load to OpenCDISCValidator Template

Controlled Terminology – Guidance

FDA: “For variables for which no standard terminology exists, or if the available terminology is insufficient and needs to be extended, the sponsor may propose their own terminology.”

CDISC: “All possible values for the variable within the trial should be included in the define.xml, except in the case of external dictionaries.” What would be included in define.xml for a female-only

study? F=Female M=Male UNKNOWN=Unknown UN=Undifferentiated

Define.xml – Implementation Define.xml is a Data Definition Document describing the

content and structure of submission data in a machine-readable format (.xml)

To create define.xml OpenCDISC Validator

Custom-built macros/utilities to collate all submission artifacts (datasets, mapping template, aCRF, Reviewer’s Guide, SAP, TLG, analysis programs)

To validate define.xml WebSDM/Janus rules

OpenCDISC Validator

Define.xml – Example Created by OpenCDISC Validator

Define.xml – Example Created by Custom-built Tool for SDTM

Define.xml – Example Created by Custom-built Tool for ADaM

Define.xml – ADaM Parameter Value-Level Metadata Challenge

From CDISC ADaM document:

Variable Metadata for Dataset ADLB

Variable Name

Variable Label

Variable Type

Display Format

AVAL Analysis Value

float 7.3

Parameter Value Level Metadata for ADLB.AVAL

Variable Value

(Where PARAMCD=)

Label

(Where PARAM=)

Variable Type

Display Format

Source / Derivation

AVAL XYZ LAB XYZ (mg/dl)

float 7.3 Calculate as median value of multiple measures if present

AVAL ABC LAB ABC(g)

float 7.3 Select maximum value if multiple measures are present

AVAL RXYZABC RATIO OF XYZ / ABC

float 7.3 AVAL of LAB XYZ divided by AVAL of LAB ABC

AVAL HYS1FL Elevated Transminase

float 7.3 <blank>

AVAL HYS2FL Elevated

Transminase

and Elevated Bilirubin

float 7.3 <blank>Orange are PARAM represented by *DEFAULT*Yellow are other non-default PARAMS

But the define.xml should display:

Define.xml – FDA Guidance “A critical component of data submission is the define file.

A properly functioning define.xml file is an important part of the submission of standardized electronic datasets and should not be considered optional.”

“As a transition step, CDER prefers that sponsors submit both the define.pdf and define.xml formats. The define.pdf is primarily for printing purposes and need not include hyperlinks.”

“…, sponsors should make certain that every data variable’s codelist, origin, and derivation is clearly and easily accessible from the define file. An insufficiently documented define file is a common deficiency that reviewers have noted.”

“Please provide supporting documentation that describes the non-standard terminology that is used. The define.xml is the preferred mechanism for documenting such terminology issues.”

SDTM Annotated CRF – Example

Reviewer’s Guide – Example

CDISC Submission Package in eCTD

Tabulations SDTM datasets (.xpt) SDTM annotated CRF Define.xml Define.pdf Reviewer’s Guide

Analysis ADaM datasets (.xpt) Define.xml Define.pdf Reviewer’s Guide Analysis programs

(.sas) SAP TLG

Lessons Learned – Maintain a CDISC Study Registry Current CDISC standard versions

CDASH: 1.1 SDTMIG: 3.1.3 (released 7/16/2012, TR and TU under review) ADaMIG: 1.0 Define.xml: 1.0 Controlled Terminology

CDASH: 6/29/2012 SDTM: 6/29/2012 ADaM: 7/22/2011

Software tools/macros/utilities/templates versions Keep track of versions of CDISC standards used for each

study can help: Plan for migration Document the Reviewer’s Guide

Lessons Learned – Maintain a CDISC Standard Hierarchy

Implementing CDISC standards on project-by-project basis is:

Inefficient

Inconsistent

Difficult for ISS and ISE

Confusing to regulatory agency

Standard Hierarchy

CDISC Sponsor TA Product Study Analysis

References CDISC: http://www.cdisc.org

FDA CDER Common Data Standards Issues: http://www.fda.gov/downloads/Drugs/DevelopmentApprovalProcess/FormsSubmissionRequirements/ElectronicSubmissions/UCM254113.pdf

Medidata Rave: http://www.mdsol.com/products/rave_overview.htm

Kinship EXACT: http://kinshiptech.com/products.htm

OpenCDISC Validator: http://www.opencdisc.org

Thank You

Q&A – Dan Shiu, ProSAS Inc.

dan.shiu@prosasinc.com

669-241-4035