By Douglas Bain, vice president, Global Operational Excellence, Medidata Solutions Worldwide
During the last 10 years, a great number of experts across the eClinical industry have invested a considerable amount of work into an extremely complex balancing act: ensuring that clinical study development tools are flexible and simple-to-use, while also delivering on the promise of interoperability and the ability to output submission-ready data sets. Of particular note are the efforts of the Clinical Data Interchange Standard Consortium (CDISC), resulting in the Study Data Tabulation Model (SDTM), Operational Data Model (ODM) and recently published Clinical Data Acquisition Standards Harmonization (CDASH) standards.
Early on, SDTM and ODM were established as standards that could deliver measurable benefits to clinical researchers. The ODM standard gained acceptance primarily as a means of exchanging clinical data between systems, but also—typically with extensions—as a syntax for study definition metadata. Meanwhile, SDTM offered a common basis to create a library of domains that can be used for clinical data submission. For sponsor companies, especially the larger ones comprising numerous mergers and acquisitions and with disparate geographical entities, SDTM was a good independent structure that could be more easily accepted, as compared to proprietary, internal standards.
However, a key obstacle still remained toward standardizing outputs: The uncertainty at the beginning of the clinical trial design process—specifically, the variability of inputs.
CDASH Helps Close the Loop
The publication of CDASH in October helped close the loop in standards for data capture, interchange and submission. CDASH focuses on the development of consensus-based data collection domains and fields in case report forms (CRFs) applicable across therapeutic areas and sponsors. Prior to CDASH, considerable data mapping was required from a CRF in order to achieve SDTM datasets, which significantly increased the cost and complexity of eClinical studies. Now, combining CDASH, ODM and SDTM, sponsor organizations can use a standard set of all CRFs as the basis for integrations between disparate systems.
For example, if both an electronic data capture (EDC) system and an integrated voice response (IVR) system meet CDASH and ODM standards, the two systems can virtually automatically communicate and transfer data leveraging CDASH, as opposed to requiring development of a custom interface. In non-CDASH enabled configurations, each study would require case by case mapping of data between the target and source systems. This typically demands a full specification and validation process, which is expensive and a point of potential failure. With CDASH applied, the two systems can communicate virtually out-of-the-box with only “delta” changes, which reflect specific study requirements.
eClinical systems now also offer Web services interfaces more frequently. These services allow for the secure exchange of data and metadata on Internet-based systems on a real-time, or near real-time, basis. When leveraging CDASH with ODM over Web services, supporting systems are also able to communicate rapidly and reliably. For example, an EDC system might capture key subject demographic and screening information, and then request a randomization code from an IVR system. The IVR system can check the information provided and then issue the appropriate randomization code, after which the EDC system can then present the results—all of this within a few seconds. In this scenario, the communication mechanism is Web services, the syntax is CDISC ODM and the metadata standard is CDASH.
CDASH’s Impact on the Future
CDASH provides an additional level of standard detail beyond what has already existed by providing standardization at the front-end of data collection, allowing the preparation of standard data capture and integrations within and between eClinical systems. The pre-validation of interfaces that meet these standards will open up the potential for more widespread usage of integrated systems where formerly cost, complexity or risk has ruled out their deployment. The timely availability of centralized clinical data will enable earlier decision making across studies and remove barriers for novel study execution methods such as adaptive clinical trials.