An introduction to CDISC and CDASH Bring on the Standards!

An introduction to CDISC and CDASH Bring on the Standards! Emmanuelle Denis M.Sc., MICR Global Health Clinical Trials Research Programme

What is CDISC?

Clinical Data Interchange Standards Consortium Good data management practices are essential to the success of a trial because they help to ensure that the data collected is complete and accurate.

The objectives of good clinical data management are to ensure: That the trial database is complete, accurate and a true representation of what took place in the trial That the trial database is sufficiently clean to support the statistical analysis and its interpretation

Where do I start?

Start with the protocol Data to be collected are defined by the protocol’s: Primary Objectives Secondary Objectives Other major pre-planned analysis

Designing the data collection instrument A well designed Case Report Form (CRF) is key to obtaining accurate and complete trial data. CRF design should begin in parallel with protocol development as the CRF is essentially the data capture system for the protocol.

Designing the data collection instrument Necessary data only: collect only data that will be used for analysis and avoid collecting redundant data Statistical Analysis Plan to define essential data that needs to be collected and ensure that there are no redundancies Ensure that all members of the study team have adequately reviewed the CRFs before they are finalised Keep the end-user in mind so that CRF is quick and easy for site personnel to complete. Also, consider the source data. Keep the CRF questions clear and unambiguous to ensure that they do not ‘lead’ Avoid collecting ‘free text’ as it will require coding before it can be analysed Use ‘yes/no’ checkboxes whenever possible or to provide ‘picklists’ Ensure that translated CRFs are reviewed to ensure that the questions have a consistent meaning in all languages Prepare CRF completion guidelines to assist site personnel in completing the forms Adapted from CDISC Clinical Data Acquisition Standards Harmonization (CDASH), 2008

How should I process my data?

Clinical Data Management Systems A clinical data management system or CDMS is software used in clinical research to manage clinical trial data CDMS used to process data from source through to validation checks, analysis, reporting and storage CDMS also used for coding (particularly for adverse events and medications) – MEDRA (Medical Dictionary for Regulatory Activities) – WHOART (WHO Adverse Reactions Terminology)

FDA 21 CFR Part 11 Many trial sponsors and funding agencies specify that a CDMS must be 'FDA' compliant. US FDA 21 CFR Part 11 requirements: System validation Robust audit trail Security access controls Specification for system design and edit checks Archiving procedures Electronic signatures

Free, open source, web-based software for EDC developed by Akaza Research Features: FDA 21 CFR Part 11 compliant 7000 users in 76 countries Java-based server system Management of diverse clinical studies Clinical data entry and validation Data extraction Study oversight, auditing, and reporting www.openclinica.org

Paper or electronic CRFs? The choice to use paper or electronic influenced by the CDMS in use and the local infrastructure e-CRFs are less expensive to process but need to factor in the extra cost of computers or mobile devices and internet or GPRS access If paper CRFs must think ahead about the flow of data and account for the extra time that will be required for data entry Consider shipping costs for multi-site trials Once trial enrolment begins participant information must be entered onto the trial database as quickly as possible to enable accurate tracking of trial progress and monitoring of safety data Also importance to track the flow of paper CRFs from the study sites to the coordinating centre.

How do I ensure quality data?

ALCOA Principle “Data quality” refers to the essential characteristics of each piece of data; in particular, quality data should be: Accurate Legible Complete and Contemporaneous Original Attributable to the person who generated the data WHO Handbook for Good Clinical Research Practice (GCP) : Guidance for Implementation, 2005

ALCOA Principle Quality data should be: Accurate Trial database accurately captures the source data Any corrections or changes are documented Audit trail! WHO Handbook for Good Clinical Research Practice (GCP) : Guidance for Implementation, 2005

ALCOA Principle Quality data should be: Legible Clear handwriting on CRFs Do not obliterate information when making changes/corrections All data (including meta-data and audit trails) must be in human-readable form WHO Handbook for Good Clinical Research Practice (GCP) : Guidance for Implementation, 2005

ALCOA Principle Quality data should be: Complete and Contemporaneous Avoid blank data fields or provide explanation (e.g. unknown, unobtainable, not applicable) Data must recorded at the time the activity occurs Audit trails to provide evidence of timing WHO Handbook for Good Clinical Research Practice (GCP) : Guidance for Implementation, 2005

ALCOA Principle Quality data should be: Original Original data (e.g. lab results, study questionnaires Accurate transcriptions of source data WHO Handbook for Good Clinical Research Practice (GCP) : Guidance for Implementation, 2005

ALCOA Principle Quality data should be: Attributable Who recorded the information Only designated study staff should have access to data Audit trail! WHO Handbook for Good Clinical Research Practice (GCP) : Guidance for Implementation, 2005

References Food and Drug Administration (FDA). Guidance for Industry Part 11, Electronic Records; Electronic Signatures - Scope and Application. 2003; Available from: http://www.fda.gov/CDER/GUIDANCE/5667fnl.htm CDISC. Clinical Data Interchange Standards Consortium. Available from: http://www.cdisc.org/about/index.html CDISC. Clinical Data Acquisition Standards Harmonization (CDASH). 2008; Available from: http://www.cdisc.org/standards/cdash/index.html Fegan, G.W. and T.A. Lang, Could an open-source clinical trial data-management system be what we have all been looking for? PLoS Med, 2008. 5(3): p. e6 Mats Lörstad, Data quality of the clinical trial process - costly regulatory compliance at the expense of scientific proficiency. The Quality Assurance Journal, 2004. 8(3): p. 177-182 Knatterud, G.L., Guidelines for Quality Assurance in Multicenter Trials A Position Paper. Controlled clinical trials, 1998. 19(5): p. 477 Baigent, C., Ensuring trial validity by data quality assurance and diversification of monitoring methods. Clinical Trials, 2008. 5(1): p. 49