Rapid Standardization in Clinical Trials with Metadata Repository (MDR)

The norms and mandates governing clinical trials are highly convoluted. After passing through multiple radars, a successful trial vouches the efficacy of medical devices or drugs that detect, diagnose, nourish, and treat. Strict mandates, geopolitical regulations etc. are therefore, obligatory across the trial process. To complicate it further, each organization follows its own standards while conducting the study. So, to simplify the diversity, stakeholders involved in the trial process formed industry consortiums that establish common standards for clinical trials. Standards alone, however, do not yield effort optimization. It requires a fully functional metadata repository (MDR) to connect all sources of truth across systems and enable reusability to make the standardization efforts a success. In this blog article, we will learn about the two broad challenges faced by the industry, and how a well-constructed MDR can tackle those.

Standards and version updates

Multiple initiatives by enterprises have resulted in the creation of various standards which are siloed and incoherent. They are maintained and published separately, for instance, Clinical Data Acquisition Standards Harmonization (CDASH) for source collection, Study Data Tabulation Model (SDTM) for submission data, Analysis Data Model (ADaM) for data analysis and so on. The versioning of each of these standards is unsynchronized which is worrisome since they are interlinked. For example, a new version of CDASH potentially may have a significant impact on SDTM. Analyzing updates at study level is tedious and time-consuming, therefore detrimental when the objective is timely delivery. With every new release, variations in standards complicate matters where decisions about inclusion or exclusion of items in the new release are to be taken. Version upgrading is a time-consuming and labor-intensive process which involves a lot of cross-functional stakeholders. In such cases, technology can help reduce the effort by providing user-friendly comparison reports and quickly resolve the versioning issues.

Reusability and Governance

The second key challenge faced in standardization is its reusability and governance aspect. Two drivers that would sustain and prolong the usage of standards are the scope of automation in reusing the standards and setting up studies faster and deploying a robust governance framework.

Standards implementation requires equal support from all users. Users typically exhibit reluctance or are complacent in using standards due to not paying attention to the frequent updates. Consequently, if the workflow and governance processes are not streamlined, maintaining standards would be cumbersome with every user deviating and carrying out their own activities.

So how does MDR help?

Let us look at a few ways how MDR helps in overcoming the standardization challenges:

• Robust repository - By design, MDR acts as a one-stop repository expanded both horizontally (include new standard models such as Protocol Representation Model (PRM), CDASH, SDTM, ADaM, or sponsor specific) and vertically (introducing levels like Therapeutic Area (TA), Project, Phase, and Study) to cover any additional functionality. For instance, attributes as Source distinguishes between company specific standard and industry specific standard (SDTM, ADaM). Such categorization helps in maintaining complete metadata at a single location. Additionally, data model can accommodate linking of elements allowing for end-to-end lineage and scope to optimize automation in the study setup. This provides flexibility to assess the impact of probable change on other metadata across all levels.
• Version updates - Periodic updates to standards [e.g. the new National Cancer Institute, Controlled Terminology (NCI CT) version] are assessed before incorporating them within the MDR. By doing this, the existing library remains intact and ongoing studies remain unhindered. The new versions must be implemented in such a way that any company-specific metadata created in the previous version can be easily implemented in the new version, if required. This reduces efforts by curbing redundancies in identifying and recreating such elements in the new version.
• Governance framework - The metadata repository (MDR) cannot function optimally if it is not supported by a strong governance framework. Each stakeholder in MDR, whether it is the governance team or the study team, is required to have well-defined roles and permissions that can be configured within the MDR. A strong and well-defined user management is the foundation to have the governance workflow and establish processes.

A robust governance mechanism makes available all the necessary data points to the applicable owners for them to take informed decisions on whether to accept or reject a change.

Conclusion

Metadata standardization is one of the most required, in-demand, and popular initiatives to be undertaken, but it would lose its transformational impact if we are unable to realize its optimal value, something that the MDR can help us with. A well-designed functional MDR can overcome the existing challenges by not only centralizing metadata management but also facilitating automation and impact analysis. This foundation will enable the implementation of advanced techniques, like artificial intelligence, to identify related processes and enable rapid study setup.