The NINDS mission is to reduce the burden of neurological disorders through research. Patient studies, including clinical trials that test or compare treatments, can provide important insights for understanding disease, lead to better treatments, and are essential to fulfilling our mission. Many clinical research data sets, however, are only used to address the researchers’ original study question, in part because NINDS has left it to researchers to determine what data they would collect and how they would collect them. As a result, many different formats have been used for data collection which has made it very difficult to compare and/or combine the data sets. Creating the data sets takes significant time and resources as well as considerable effort on the part of the patients who participated in the studies. NINDS is therefore working to make sure that we learn as much as possible from them.
To ensure the best possible use of clinical data sets, NINDS has launched the “Common Data Elements” or CDE project. Central to the project is the identification of common definitions and the standardization of case report forms and other instruments. The goal of the project is to help investigators develop and use uniform data formats so that they systematically collect and analyze data that can be shared across the research community. The CDEs are organized hierarchically and in several categories. General CDEs (for example gender and age) should be collected in all clinical research. The “core” CDEs are recommended for clinical research in a given disease. In addition, there are optional “supplemental” elements, or innovative “exploratory” elements. For the last two categories, the CDEs provide a suggested format, but do not require the use of these elements.
To create CDEs for a particular disease, the NINDS convenes experts to make recommendations on what should be collected as part of clinical research. We provide support for the process, but don’t dictate the elements. The initial version of the CDE is posted online for public comments, and then provided as Version 1.0 on an open website. The website shows the expert recommendations, the data elements with their definitions, and printable case report forms (CRFs). Version 1.0 is not the end; the CDEs are dynamic and will evolve over time through an iterative process of periodic review and updates. To date, we have focused on neurological diseases with relatively high prevalence or high clinical research activity, including stroke, Parkinson’s disease, the epilepsies, traumatic brain injury, spinal cord injury and headache (see www.commondataelements.ninds.nih.gov).
What does the NINDS do to disseminate the CDEs?
To increase awareness and use of CDEs, NINDS is partnering with stakeholders in neurological clinical research including patient organizations. For example, the Muscular Dystrophy Association is working with the NINDS to include CDEs in their new clinic registry for neuromuscular disorders. We also work with professional organizations, for example, the American Spinal Injury Association, which can foster the use of CDEs for research. We collaborate with other federal agencies, such as the DoD and VA in coordinating data collection from research studies of traumatic brain injury. The biopharmaceutical industry is a key stakeholder in bringing new treatments to patients. We have therefore involved industry investigators in our CDE working groups and requested industry case report forms when applicable. Because the NINDS is responsible for research on rare neurological diseases, we are reaching out to international investigators and organizations. One example of this is our coordination with the Canadian Institute for Health Information related to their Multiple Sclerosis Registry.
Building partnerships to make data accessible is another long-term goal of the CDE project. In some cases, we have been able to partner with other stakeholders to support a database that uses the CDEs (e.g. TBI and PD biomarkers). We have also conducted a pilot project to map our CDEs for Parkinson’s disease to a widely used data standard, CDISC. We have received assistance from the National Library of Medicine in resolving inconsistencies and in improving metadata of CDEs.
NINDS strongly encourages the use of CDEs in our funding opportunity announcements. We provide hands-on training and support to investigators, and are developing online training materials. We work to make the CDEs user-friendly so that they can accelerate the start-up time for trials. Instead of investing time and resources to create data forms for each new trial, researchers can now build on existing standards.
Creating the CDEs requires a joint effort of clinical investigators, patient groups, industry, NIH staff and contractors, and data professionals. We are grateful to all who have volunteered their time and expertise to the CDE initiative. We are also grateful to those researchers who are using the CDEs. We realize that change requires time and resources, but we think that it will be a good investment because standardized data will facilitate sharing, help ensure that valuable clinical research datasets realize their full value, and accelerate progress towards better treatments for neurological diseases.
Last updated December 26, 2012