NINDS Common Data Element (CDE) Project –The Use of Common Data Standards (CDS)

September 8, 2015

Joanne Odenkirchen, MPHNINDS, Office of Clinical Research

3

Motivation & Overall impact of the NINDS CDE Project

MotivationTrials were costing too much: no one believed in re-use of CRFs Trials were taking too long and costing to much to get up and going Data quality varied, no standardsData collection was not consistentComparisons of data between studies was not possible

ImpactReduce time/cost to develop data

collection toolsReduce study start-up time and cost of

overall trial Improve data qualityFacilitate collection of dataFacilitate data sharing/comparisons

between studies and meta-analyses

4

Objectives of the NINDS CDE Project

• Identify CDEs used in clinical research (age, gender, race, etc.)

• Provide information to clearly describe a question Identify common definitions, permissible values, range

checks, etc.• Present data elements in a standard format available to all • Standardize case report forms and instruments

5

Collaborative Effort: CDE Development & Implementation• Expertise from close to 1000 specialists worldwide

Experts from every continent except Antarctica

• NIH institutes (17+), federal agencies (including FDA, CDC, VA, DOD, ACL, USUHS, AHQR, PCORI), and Nebraska State Health Department representative

• Collaborate with non-profits/foundations: American Heart Association, American Academy of Neurology, Muscular Dystrophy Association,

Prize4Life, ALS Association, Fredric's Ataxia Research Alliance, CHDI Foundation, MA, United Mitochondrial Disease Foundation, Neurocritical Care Society, MRG, National Multiple Sclerosis Society, American Epilepsy Society, American Spinal Injury Association, ISCoS, Myasthenia Gravis Foundation of America, Sarah Jane Brain Foundation, Craig H Nelson Foundation, Parkinson’s Disease Foundation, MJF, AACPDM, Chiari & Syringomyelia Foundation, etc…

• Pharmaceutical/Laboratories/ Companies Allergan, Isis Pharmaceuticals, Cytokinetics, Glaxo Smith, Medtronic, BioMarin, Apotex Research,

Merck, Stealth Peptides, Acorda Therapeutics, Edison, Novartis, Teva, EMD Serono, Biogen Idec, BrainScope, Bayer HealthCare, Sarepta Therapeutics, Banyan Group, MNG Labs, Coriell, etc…

6

Collaborative Effort: CDE Development & Implementation• National Library of Medicine - NLM Repository Pilot Project and Assigning

SNOMED, LOINC, RxNORM Values to NINDS CDEs

• NIH CIT BRICS – FITBIR and PDBP using the NINDS CDEs

• NIH-wide: PhenX, BRAIN, NIA’s Alzheimer’s Coord . Center, CaDSR, PROMIS, BMIC, etc…

• Harmonization with international data standards

• Clinical Data Interchange Standards Consortium (CDISC) & C-PATH

• NIH funded studies –part of our funding announcements

• Collaborate with non-profits/foundations to develop new CDEs and share CDEs for use in registries/studies

• Public (in the public comment period during development and feedback on our website)

NINDS CDE Disease Areas – over 11,000 CDEs & 575 Instruments

Epilepsy*HeadacheMitochondrial disorders*Movement disorders• Parkinson’s disease• Huntington’s disease

Multiple sclerosisSpinal cord injury (SCI)*Stroke*Traumatic brain injury*

* Pediatric Specific Recommendations

Neuromuscular disorders*• Amyotrophic lateral sclerosis • Friedreich’s ataxia• Muscular dystrophies

–Congenital, Duchenne/Becker, Facioscapulohumeral, Myotonic

• Myasthenia gravis• Spinal muscular atrophyCerebral palsy (in development)Subarachnoid hemorrhage (in development)Chiari & Syringomyelia (in development)

General CDEs

88

9

CDE Development Process

Development Step Typical Timeframe

Introductory Working Group (WG) Web conference meeting(Includes information on deliverables, COI, transparency, re-use of current CDEs & instruments, expectations, support, & end of project publication)

1-2 hours

Subgroups* meet every 3-5 weeks via conference call to develop CDEs for assigned areas

6-9 months(has been done in 6 mo’s)

Internal WG Review of all subgroups’ CDEs 1 -2 months

Subgroups revise CDEs based on feedback from Internal WG Review 1 month

Public Review of WG’s CDEs 6 weeksSubgroups revise CDEs based on feedback from Public Review 1-2 months

Post Version 1.0 of CDEs on Website 1 monthTOTAL 12-16 months

* Not all disease/disorders have subgroups

10

Example of Disease-Specific Working Groups (WG)

• Biomarkers• Demographics• Cognitive/Behavioral/Psychological Outcomes• Endocrinology/Diabetes/GI/Nutrition• Exercise Physiology• Genetics• Imaging• Neurological Assessments• Patient Reported Outcome/QOL• Vision• Hospital/Care Management• Pulmonary• Cardiac• Outcomes and Endpoints• TherapiesNote: not all Disease- specific working groups have broken into sub-groups

NINDS Common Data Elements: TBI http://www.commondataelements.ninds.nih.gov/tbi.aspx#tab=Data_Standards

12

Specific Recommendations: TBI Guidelines, CRF, and CDE Development

TBI Global Outcomes:http://www.commondataelements.ninds.nih.gov/tbi.aspx#tab=Data_Standards

TBI Outcome Domains and Descriptions Recommendations:http://www.commondataelements.ninds.nih.gov/tbi.aspx#tab=Data_Standards

14

TBI Outcomes Guidelines CRF, and CDE Development – CRF:

TBI Global Outcomes: Disability Rating Scale (DRS), Notice of Copyright: http://www.commondataelements.ninds.nih.gov/Doc/NOC/Disability_Rating_Scale_NOC_Public_Domain.pdf

15

Disability Rating Scale - CDEs

TBI Disability Rating Scale (DRS) CDEs:https://commondataelements.ninds.nih.gov/ReportViewer.aspx?/nindscdereports/rptAllCDE&rs:Command=Render&rc:Parameters=false&CDEName=&DiseaseName=Traumatic Brain Injury&SubDiseaseName=Comprehensive&DomainName=+&SubDomainName=Global Outcome&Classification=+&IsCopyright=&Population=+&CdeId=+&keyword=&CrfId=F0031&CrfName=Disability Rating Scale (DRS)

Examples of Working Group Instrument Considerations:

Example of Mitochondrial Disease Working Group:

• Description of scale• Population• Where applied before?• Was it validated?

– If validated, give validation criteria• Scoring• Notable studies used scale• Limitations in use for in mitochondrial disease• Advantages in use for the mitochondrial disease• Time required to complete• Should scale be evaluated further for use in

disease?

Example of Congenital Dystrophy Working Group:

• What are the specific population where instrument is used (include age and criteria if applicable)?

• What are the specific diagnoses/diseases where instrument has been applied?

• What are the specific diagnoses/diseases where instrument has been validated?

• How is the instrument scored?• Are there notable studies which have used the

instrument? (List references if notable)• Limitations for use in CMD patients• Advantages for use in CMD patients• Time required to complete instrument• Fee associated or copyright limitations• Should this instrument be recommended for use in

CMD studies?

17

NINDS CDEs Lessons Learned Need for continuous QA/QC of CDEs Promote re-use of CDEs Investigate other CDE sources 1st Scientific findings or new valid instruments change the CDEs Involve various/different groups Listen to input and different perspectives Need for better documentation CRFs help Version 1.0 is NOT the end Education of the clinical researcher:

Once the discussion of the CDEs moves towards finding better diagnosis, treatment, prevention, and other research study goals and objectives, the promotion of the CDEs by researchers is easily accomplished.

18

NINDS Vision for CDEs• Future NINDS-funded trials use CDEs or be CDE-

compatible – it is part of FOA and Terms of Award• All types of clinical research can use part of the CDEs

Observational clinical studies can be linked to trial datasets All human subject grantees are asked to consider using CDEs

• Clinical research progress will be accelerated New investigators can build on consensus data elements Start-up of multi-center and international clinical research

efforts will be facilitated

19

NINDS Funding Announcements/Terms of AwardNINDS Phase III Investigator-Initiated Efficacy Clinical Trials (U01) PAR-11-173

Section I. Funding Opportunity Description: Objectives of the ProgramThe DCC application should present and discuss the statistical and data management plans for the study, including plans for using NINDS Common Data Elements (http://www.commondataelements.ninds.nih.gov/#page=Default) and sharing of data following the conclusion of the trial. 

 

Section IV. Application and Submission Information:  SF424(R&R) Other Project InformationStatistical methods…  Data management plans should incorporate Common Data Elements (CDE; see http://www.commondataelements.ninds.nih.gov).  

 

Section V. Application Review Information:  Scored Review Criteria Criteria:  Approach

Are the overall strategy, methodology, and analyses well-reasoned and appropriate to accomplish the specific aims of the project? …  Are NINDS Common Data Elements considered? Are the milestones appropriate? 

__________________________________________________________________________Use of NINDS Common Data Elements (CDEs) for Phase III Trial – Terms of Award

“The NINDS strongly encourages researchers who receive funding from the Institute to use the NINDS Common Data Elements (CDEs) available at www.commondataelements.ninds.nih.gov or document how they will ensure their data collection is compatible with the CDEs. Investigators should use the common definitions and the standardized case report forms and other instruments identified by the CDE Project. The CDE Project has developed uniform formats by which clinical data can be systematically collected, analyzed and shared across the research community.

The NINDS Program Official must approve awardee's plan for using CDEs before enrolling the first study participant. Justification must be clearly provided in the plan if any general CDE or disease-specific (as available) will not be used.”

20

NINDS CDEs Project Future Plans in the next 5 years• Provide easier access to all CDEs for investigators• Continuing promotion of the use of NINDS CDEs • Promoting CDEs along with their standard associated coding

(e.g., LOINC, SNOMED, ICD9/10, HL7) for data exchange• Researcher-centered promotion of CDEs• Collaboration and Harmonization• Collaborating with NLM to develop a CDE Development

Model thru the NLM CDE Repository • Envision more organizations/associations will develop CDEs

using NINDS’ paradigm with only assistance from NINDS

21

Accessing the NINDS CDEs

NINDS Common Data Elements Website

www.commondataelements.ninds.nih.gov

Submitting Feedback on CDEs Feedback form on NINDS CDE website

http://www.commondataelements.ninds.nih.gov/ProjReview.aspx#tab=Feedback_and_Suggestions

For more information on the NINDS CDEs, please contact: Joanne Odenkirchen jo21x@nih.gov