Best practices for creating and maintaining a library of data collection standards

This video provides an in-depth exploration of best practices for creating, implementing, and maintaining a library of data collection standards within clinical research, with a particular focus on leveraging Veeva Vault CDMS/EDC. Richard Young, VP Strategy at Veeva, sets the stage by outlining the critical benefits of data standards, such as reducing cycle times, improving data quality, enhancing compliance, and ensuring consistency across studies. He highlights common challenges in standardization, including difficulty getting started, lack of ownership, proliferation of standards, overstuffed standards, and outdated practices. The presentation emphasizes aligning with industry standards like CDISC (C-Dash) and SDTM, keeping the end in mind for downstream processes, and adopting a site and patient-centric approach.

Jen Showalter, an industry expert in data standards, then presents a case study from Eli Lilly, detailing their journey in building and maintaining a library-driven approach for Veeva Vault CDMS. She describes Lilly's phased approach, starting with identifying frequently used forms, collaborating with end-to-end stakeholders (from sites to statisticians and medical teams), and establishing robust governance. A key recommendation is to build new forms from scratch within the library rather than copying old practices, leveraging new tool capabilities like progressive displays and form linking, and ensuring a unique "library key" is assigned to each object for effective tracking. The discussion delves into the specific components managed within their Veeva Vault CDMS library, including events, forms, item groups, items, code lists, and various user-defined rules (dynamic, data verification, set value, subject status, and email rules). Beyond the system, Lilly also manages external documentation like allowable changes, data check specifications, and completion guidelines, stressing the importance of aligning these with the library's version.

The presentation further elaborates on the maintenance and governance of the library. Lilly employs a master-level management approach to minimize variations, using guidance to inform study designers on allowable changes. They set clear expectations for library usage, aiming for high adherence (e.g., 80% usage) while having a tiered governance process to review and approve necessary deviations quickly. A crucial tool for monitoring compliance and informing maintenance is the system-generated difference report from Veeva EDC, which tracks divergences between study objects and library objects. This report also enables a risk-based approach to testing, reducing rework. Changes to the library can originate from external sources (regulatory, industry standards like CDISC/SDTM, system enhancements from Veeva releases) or internal feedback from studies. All change requests are tracked in a centralized system, informing quarterly maintenance cycles and allowing for impact assessments. The presentation concludes by highlighting the realized benefits, including allowing study teams to focus on unique scientific aspects, improved site experience, downstream efficiencies for data management, medical review, and biostatistics, enhanced automation, reduced study build cycle times, minimized rework, and improved data quality.

Key Takeaways:

• Strategic Importance of Data Standards: Data collection standards are crucial for reducing clinical trial cycle times, improving data quality and consistency, enhancing regulatory compliance, and driving efficiencies across the entire data lifecycle, from design to reporting.
• End-to-End Stakeholder Collaboration: Successful library implementation requires collaboration with all stakeholders, including sites (for user experience), medical affairs, statisticians, and data management, to ensure content meets diverse needs and facilitates adoption.
• Veeva Vault CDMS/EDC as a Foundation: The Veeva Vault CDMS/EDC platform provides robust capabilities for building and managing a data standards library, including features for forms, items, code lists, user-defined rules, and form linking.
• Library-Driven Approach for Efficiency: Starting with a well-defined library of standard templates (forms, events, rules) significantly reduces study build times and rework, allowing teams to focus on scientifically unique aspects of a study.
• "Library Key" for Tracking and Maintenance: Each object in the library should have a unique identifier (library key) to track its lineage and usage across studies, enabling effective monitoring of adherence and informing future maintenance decisions.
• Build New, Don't Copy Old: When establishing a library, it's more beneficial to build new forms and objects from scratch, leveraging modern tool capabilities, rather than replicating potentially inefficient legacy practices from existing studies.
• Comprehensive Governance Model: A robust, tiered governance process is essential to manage change requests, ensure adherence to standards, and facilitate timely review of necessary deviations without slowing down study development.
• Minimize Variation, Maximize Guidance: Instead of creating numerous variations of a standard within the library, manage core "master" standards and provide clear guidance (e.g., "allowable changes") to study designers on how to adapt them for specific study needs.
• Leverage System-Generated Reports: Tools like Veeva's "difference report" and "library report" are invaluable for monitoring library usage, identifying divergences from standards, and informing a risk-based approach to testing, thereby reducing redundant testing efforts.
• Regular Maintenance Cycles: Libraries are living entities and require regular maintenance cycles (e.g., quarterly) to incorporate new regulatory requirements, industry standards (CDISC, SDTM), system enhancements, and feedback from real-world study usage.
• Centralized Change Request System: Implement a system to track all change requests, whether from internal feedback or external sources, to prioritize updates, document decisions (including reasons for not implementing a change), and inform maintenance cycles.
• Documentation Alignment and Accessibility: All supporting documentation (completion guidelines, data check specifications, allowable changes) must be version-aligned with the library and easily accessible to all stakeholders to ensure consistent understanding and usage.
• Training and Communication are Key: Continuous training and communication are vital to ensure all stakeholders, from new hires to experienced teams, understand the standards, their importance, and how to effectively use the library.
• Downstream Benefits Beyond Data Collection: Standardization extends benefits to downstream processes, including improved data extraction for SDTM, streamlined statistical programming, enhanced medical review, and better integration with other clinical systems (e.g., CTMS).
• Resource Allocation for Library Support: Dedicate a subset of study designers, including a "lead librarian," to support library maintenance, stakeholder discussions, change request management, and training, even if not fully dedicated to the library.

Tools/Resources Mentioned:

• Veeva Vault CDMS (Clinical Data Management Suite)
• Veeva Vault EDC (Electronic Data Capture)
• CDISC (Clinical Data Interchange Standards Consortium) / C-Dash (CDISC standards)
• SDTM (Study Data Tabulation Model)
• Veeva Studio (platform for library report)
• System-generated difference report (from Vault EDC)
• Library report (from Vault EDC Studio)
• Lilly's internal request system/change log

Examples/Case Studies:

• Eli Lilly's Implementation: The entire presentation serves as a case study from Eli Lilly, detailing their multi-year effort to establish and mature their data collection standards library within Veeva Vault CDMS.
• Common Standard Forms: Examples of forms prioritized for standardization include Demography, Inclusion/Exclusion, Informed Consent, Adverse Events, Medical History, Prior Therapies, and Exposure.
• Code List Variations: Using distinct code lists for different therapeutic areas (TA1 vs. TA2) for a single disposition form allows for flexibility without creating multiple form variations in the library.
• Metadata for Downstream Transformations: Using metadata like form name, release date, and version within the standard helps drive downstream transformations and identify differences across studies.
• FAQ Guidelines: An example of an FAQ for units of measurement demonstrates how to provide instructions for geographically varied standards (e.g., US vs. global sites using Celsius vs. alternative units).