Update on SDTMIG v. 3.1.2 and SDTM v. 1.2

Update on SDTMIG v. 3.1.2and SDTM v. 1.2

Bay Area User Group MeetingMay 2008

Fred WoodOctagon Research Solutions

SDTMIG Comment Statistics

• Total - >950• Fix - 319• Check - 165• Assign - 64• Defer - 22• Discuss - 161• No action required - 101• No action suggested - 13• ADaM - 22 (many duplicates)• Duplicates - 57

Outline• Comment Statistics• Findings Issues/Actions• “Findings About”• SUPPQUAL Comments• Other SDS Team Work• SEND• FDA Issues

– Resolved– Proposals Ready– Not in Scope– Needing Clarification

• Other SDTM-Based Standards Work• Harmonization Work• The Future

SDTM Comment Statistics

• Total - 88• Fix 44• Check - 5• NA - 9• Discuss - 23• Duplicate - 7

Findings Team Actions (1)

1. Standardization of character values in --ORRES to numeric in --STRESC. QS examples retained. New LB examples inconsistent with data-flow diagram removed.

2. Population of --ORRES for derived records, wording to state that it generally should not be.

3. --BLFL for more than one record per test, dependent upon other keys (e.g., position), clarified.

4. Other sets of results and units for lab tests (standard or conventional), deferred

5. Compound test codes (test details). Decision to remove concatenation example to discourage “lumping.” Address via value-level metadata with CRT-DDS and Metadata Team.

6. Numeric variable for original results such as lab data, not implemented.

Findings Team Actions (2)

7. Length of time in a position for a Finding, under development.

8. Decreasing the emphasis on the use of RELREC for PC and PP, adding that relationship data, which is often part of the analysis metadata, does not need to be brought into the SDTM.

9. Algorithm for scoring questionnaire data

Clinical Findings (“Findings About”) Team

1. Discussion of Findings about and Event (FE) and Findings about an Intervention (FI)

2. More Examples of when to use FI and FE versus a custom Findings domain

SUPPQUAL-Related Comments

1. Value-level metadata (e.g., Origin) in SUPP--; add QTYPE to SUPPQUAL since other value-level metadata are there; numeric QVAL variable (value-level metadata), deferred.

2. Requirements for use of SUPPQUAL variables in Appendix C5, note to be added.

3. Use of domain prefixes for QNAM values, not required.

4. Examples for using SUPPQUAL variables in IDVAR for RELREC, deferred.

5. Proposal to allow relationships to be identified with multiple keys (multiple IDVAR and/or IDVARVAL) values, under discussion.

6. Deprecation of QORIG, deferred.

Other SDS Team Work

1. Revision of Appendix C2 (Domain Codes) and guidance of assigning codes for custom (sponsor-created) domains.

2. Specifying 20-character limit for ARMCD (no limit cited previously), and 200 for ARM.

3. Removal of SDTM general-observation-class tables from SDTMIG.

4. Dataset-naming and domain-code proposals to be reviewed within SDS Team

5. Change Log creation

SEND

1. “Rebaselining” - Alignment with SDTM and other CDISC standards

2. Adding new variables to SDTM for SEND (Exclusion Flag, Reason for Exclusion, Species, Strain, Substrain, AGETXT, POOLID.

3. Addition of new SEND Domains

4. Pilot Project

5. Reproductive Toxicology

Issues for Discussion with FDA: Resolved

4. Dictionary Version

6. Promotion of SUPPQUAL Variables

7. Test Details

8. Variable Order

12. Permissible vs. Required

14. No Imputed Data

Issues for Discussion with FDA: Proposals Ready (1)

1. Day in Period/Phase/Cycle

2. Baseline Flags• Algorithm for selecting baseline observation• Baselines for multiple treatment periods• Baselines for multiple tests with the same

TESTCD value

3. Population Flags

5. MedDRA Secondary Organ Class

9. Submission of Null Values for Y/N Questions

10. Exposure Records

a. Placebo Dose

b. Doses Not Given

c. Creation and requiredness of EX

11. Multiple Qualifier Values

16. Discourage Custom Domains

18. No duplication of data (need clarification)

20. Difficulty associating findings with the most recent experimental intervention or a concomitant intervention (related to 24; TD requirements)

Issues for Discussion with FDA: Proposals Ready (2)

24. ARM and ARMCD Not helpful in Crossover Studies (related to 20; TD requirements)

28. CT for AEREL (Y, N vs. Possible, Probable, etc.)

29. Expectations for use of CT in Submissions (lists of CT are not tied to SDTM versions). Versions of CT are different from versions of SDTMIG. Subset discussion included? Does FDA access published list?

30. Questionnaire Scores

32. CRF Annotation and Bookmarking Preferences

Issues for Discussion with FDA: Proposals Ready (3)

33. Referencing ECG Waveforms

34. Size of Datasets

35. MedDRA and WHO Drug numeric codes in addition to text values -add defined variables?

Issues for Discussion with FDA: Proposals Ready (4)

13. Relationships between Source Data

17. All verbatim terms need a corresponding coded value

19. Lack of Adherence

21. Lack of a clear and consistent method to capture concomitant interventions,

22. Lack of a clear and consistent method to capture device characteristics

25. Missing or Incomplete Values for Date and Time Records (similar precision for dates that might be subtracted) (from Intrachange) (Education issue)

26. Death Information (from Intrachange)

27. Not Measurable Versus Not Done (from Intrachange)

Issues for Discussion with FDA:Not in Scope for 3.1.2

15. Include Derived Data

17. “Regional Requirements” for Variables

Issues for Discussion with FDA: Needing Clarification

Other SDTM-Based Standards Work

• Oncology Subteam of the SDS Team– Oncology domains (Tumor Identification and

Tumor Measurements)• Devices Team

– Device Properties domain• Other Teams

– Cardiology and TB standards

Harmonization Efforts

• BRIDG (Biomedical Research Integrated Domain Group)

• CDASH (Clinical Data and Acquisition Standards Harmonization)

• Controlled Terminology• ADaM (Analysis Data Model)• SEND (Standard for the Exchange of Nonclinical

Data)

Beyond 3.1.2

• Change data type for STNRHI and STNRLO to character to allow for “<“ and “>”, since --STRNRC might not be used for a range, but for only for one of the limits

• Multiple Date for a single Event (e.g., start date, diagnosis date, most recent episode)

• More direction and harmonization on anatomical locations (right, left; dorsal, ventral; anterior, posterior; lateral, medial; proximal, distal; liver lobe, kidney region).

• Publishing in a better electronic format• Addition of new Intervention variables for Concentration,

Concentration Units, and Dose Form Amount.• Addition of --SSCAT• Deprecation of TI