Life sciences companies recognize the benefits of implementing CDISC standards. Why have they been slow to adopt them?
With the US Food and Drug Administration (FDA)’s December 17, 2016 deadline approaching rapidly—a deadline that requires all clinical and non-clinical studies to use standards in the CDISC Data Catalog to support new drug submission and biologics license applications—Accenture recently talked with 77 global life sciences executives about how they are implementing clinical data standards and whether or not they see value in the standards.
Although most companies see the advantages of complying with the CDISC standards and recognize that the standards offer a number of potential commercial and operational benefits, including reducing costs and speeding time to market of new drugs, they have been slow to adopt the standards.
Beyond migrating or mapping proprietary data to CDISC standards, respondents said that a number of challenges pose significant demands on their already stretched resources, including:
Training people to implement the standards.
Putting management processes in place to verify compliance with the new standards.
Developing governance structures to verify value.
Other significant findings include:
Agreement on business importance and drivers of standards
The vast majority of respondents cited regulatory compliance as the most important reason for adopting data standards. A slightly smaller percentage believed CDISC standards will actually improve regulatory compliance over the long run. Of the respondents, 94 percent currently submit to the FDA, 76 percent to the European Medicines Agency (EMA) and 53 percent comply with the Prescription Drug Marketing Act (PDMA).
While regulatory compliance is the primary driver of data standardization, nearly 80 percent of respondents also see the need for consistent data across the clinical data life cycle, and 66 percent believe standards will enhance operational efficiency.
Different data standards are used
When it came to selecting CDISC standards, the highest number of respondents chose Study Data Tabulation Model (SDTM), followed by Analysis Data Model (ADaM) and Define-xml. About half of respondents said they integrated SDTM, ADaM ad CDASH standards cross-functionally, while 36 percent have integrated standards separately in functional areas.
Organizational barriers to implementing standards are significant
Companies are clearly at different starting points and maturity in their journey toward data standardization. Consequently, different implementation issues take precedence depending on where companies are on the adoption curve. Over half of respondents viewed the difficulty in building governance processes as the primary organizational challenge. Other challenges include contending with:
Insufficient knowledge of, or experience with CDISC standards, and lack of support of new standards, processes and infrastructure.
Multiple versions of CDISC standards.
Difficulty in hiring professionals with CDISC knowledge.
Resistance from study teams in moving away from the legacy approach.
Governance and integration pose current and future challenges
Approximately half of the respondents noted that implementation of their selected data standard was cross functional, but more than a third revealed that standards were implemented separately within functional areas.
Nearly half of the respondents have governance challenges, with specific challenges ranging from limited resources within functions to a dearth of cross-functional resources. Many respondents stated that only part-time resources were deployed to their standards organization and that only modest technology support was provided to facilitate governance workflows.
Nonetheless, there is wide recognition that more structured governance will yield potential benefits mostly to improve standards awareness, verify standards compliance, and centralize accountability and authority.
Metadata management is valuable capability, but gaps persist
A clear majority of respondents expected metadata management to both improve efficiency and allow faster maintenance of standards. Yet, organization of metadata was fragmented, with some companies using spreadsheet-based standards metadata, and less than a third using metadata repositories (MDRs). A significant number of respondents indicated that they were not using standards metadata. Despite the challenge of metadata management, more than two-thirds see the value in consistency of data across studies.
Use of outsourcing varied
Nearly two-thirds of those using STDM outsourced data creation and exchange, while approximately a third of companies chose that route for ADaM or Define-xml.
What’s clear from the responses of these life sciences executives is that companies are at different levels of maturity in their journey toward meeting the deadlines and that many are struggling with issues related to data standardization capability, metadata management and governance.
In my next post, I’ll tackle the question: Could a standards maturity model help companies working toward data standardization?
In the meantime, for a summary of the research findings, see the CDISC Standards: A Catalyst for Industry Transformation infographic.