For years, data scientists have conducted research using clinical trial data the old-fashioned way by:
- Contacting a wide variety of companies, each with their own data request process—some with a secure access platform and some without
- Conducting parallel research projects using various data sets
- Manually consolidating the results to uncover clinically relevant insights
This time-consuming, complex process creates significant barriers to healthcare advancements. For example, researchers cannot always combine data sets from different providers. Keeping data sets separate is a huge hindrance for those performing large-scale meta-analyses on everything from drug interactions and drug safety to optimal drug dosing.
It was this problem that led to the formation of the Clinical Study Data Request consortium (CSDR), a clinical trial data request and sharing solution that vastly streamlines and accelerates how researchers request, access, and analyze clinical data from multiple organizations. Launched by ideaPoint and SAS in January 2014, and supported by the first data sponsors, GSK, Roche, Sanofi, ViiV Healthcare, and Boehringer Ingelheim, CSDR quickly became the leader in multi-sponsor clinical trial data sharing. CSDR is a “one-stop shop” for data scientists and researchers seeking to access and analyze third-party clinical patient data from multiple studies to advance human health. Through ClinicalStudyDataRequest.com, CDSR’s publicly available website, at no cost researchers can:
- Request access to global clinical data from 17 sponsors and study funders
- Centrally manage and track proposals, questions, and inquiries
- Combine multi-sponsor data sets in a secure workspace
- Analyze these data sets1 on a single analytics platform with the most complete set of analytics tools available today, including open source, proprietary, and SAS analytics software
- Access all shared data in the secure workspace for as long as necessary to complete the research
The power of global collaboration
In July 2013, Pharmaceutical Research and Manufacturers of America (PhRMA), along with the European Federation of Pharmaceutical Industries and Associations (EFPIA), called forgreater data transparency and collaboration within the biopharmaceutical industry to help researchers innovate, discover, develop, and deliver new therapies and vaccines. EFPIA and PhRMA also published the Principles for Responsible Clinical Trial Data Sharing as a set of guidelines. The CSDR platform was designed around these principles to ensure clinical study data could be shared with researchers, but with assurances that:
- The data would be secure
- Patient anonymity would be protected
- Any organization conducting and funding clinical trials—from big pharma to university researchers—could share their data with confidence
Today, CSDR is comprised of 13 pharmaceutical companies and four funders—The Wellcome Trust, The Bill & Melinda Gates Foundation, Cancer Research UK, and the Medical Research Council. “We recognized there were major barriers to sharing data and making it accessible and findable by researchers,” explains Georgina Humphreys, clinical data sharing manager at Wellcome Trust. “By signing up, we removed one barrier for researchers we fund to list available data.”
Similarly, one of the main reasons pharma giant Novartis joined CSDR was to tangibly embrace the principles of data sharing defined by EFPIA and PhRMA. Kevin Poirier, data sharing portfolio manager at Novartis, explains that, “We’re happy to make our clinical trial data available to researchers for free, and at the same time, CSDR gives researchers the option to access other member companies’ clinical data. CSDR is a particularly powerful tool because users can pool our data with data from other member organizations—all in one place.”
For Eisai, a Japanese pharmaceutical company headquartered in Tokyo with 10,000 employees—including 1,500 in research—joining CSDR was the obvious choice. “All of the processes were already in place and implemented in a way that was so well thought out and easy to use,” notes Ray Harris, senior director of biostatistics at Eisai. “Researchers benefit from a seamless process, start to finish, and can combine data from many studies and companies while ensuring patient confidentiality.”
Enabling new research that advances human health
CSDR has broken down the barriers to clinical trial data access, aggregation, and analysis, which has significantly improved the end-to-end research experience—especially for individual patient data (IPD) metadata analysis involving the pooling of large volumes of patient data from multiple clinical studies. Long maligned by pharmaceutical researchers generating actual patient data, IPD metadata research is increasingly recognized as playing a valuable, complementary role. Examples of how researchers are using CSDR data include:
- Running novel analyses
- Performing systemic reviews
- Checking for publication and reporting bias
For instance, IPD metadata studies can be used to compare drug efficacy for patients newly diagnosed with a condition so healthcare providers can make more informed drug and dosing recommendations that will achieve better patient outcomes. With IPD metadata analyses, this kind of knowledge can be gained without the time and cost associated with running an actual drug trial, which can take years to complete and cost a great deal of money.
Consider, for example, research shared at the Clinical Trial Data Transparency Forum 2019 in Heidelberg, Germany, where industry leaders, researchers, and peers discussed how the life sciences industry is building on the success of the data sharing movement. Dr. Sarah Nevitt from the University of Liverpool, U.K., explained how meta-analyses of granular patient data, including unstructured text data, can capture real patient experiences such as side effects experienced or reasons for stopping a treatment. Such details can reveal how different drugs affect patient-centered variables like quality of life—not just medical efficacy, which original clinical studies typically focus on. These findings help healthcare providers consider the full burden of treatment for patients when making drug decisions.
At the Forum, Nevitt showcased her own metadata study analyzing patient data from CSDR and other sources. The analysis involved 45 pairwise comparisons of antiepileptic drugs, some of which had never been compared head-to-head in drug trials. The study, titled “Cochrane Individual Participant Data Network Meta-Analysis (IPD-NMA) of Antiepileptic Drugs (AEDs),” is now influencing guidelines on the choice of drugs—guidelines that are used for clinical decision-making and balancing the benefits and harms of commonly-used drugs for patients with different types of epilepsy. This research could not have been done without clinical trial data sharing enabled through CSDR and similar platforms.
Accelerating research through quicker data access
The growing number of research proposals submitted for review by the CSDR team testifies to its perceived value by researchers and statisticians. Since 2014, over 800 researchers from 500 institutions in 42 different countries have submitted proposals, and recently, the number of requests have jumped from 85 in 2017, to 105 in 2019. This is a good indicator that CSDR is making an impact. “The more studies that are listed, the more subtle are the things that people can analyze,” comments EISAI’s Harris.
CSDR’s member companies also measure the impact of CSDR by assessing time to clinical data access, as faster data access accelerates the delivery of published clinical recommendations. “We’ve been able to reduce the average time from submission of the request form to accessing data from 200 days in 2018 to 139 days in 2019, although we accept there are still some that take much longer,” explains The Wellcome Trust’s Humphreys. “Already, there are 70 published studies that used CSDR data, and another 30 that are awaiting publication.”
Six years ago, CSDR was the first organization to offer multi-sponsor clinical study data access. Other sharing platforms have followed, such as Vivli in 2018, which, like CSDR, covers multiple disease areas and multiple sponsors; and Project Data Sphere, which is focused on cancer research. Since CSDR’s inception, more than 560 research requests have been received, with nearly 60 studies published using data from the platform, and nearly 30 more anticipated. As CSDR has continued to seek to enhance its data sharing platform, a key success and differentiator has been the streamlining of data requests and access in order to enable researchers to merge data from different pharmaceutical companies, and analyze it directly within a secure platform. In addition, with the availability of SAS, open source, and commercial capabilities, CSDR is able to provide a complete solution with a truly comprehensive analytical platform.
However, given that multiple clinical data platforms are now available, the long-term industry goal must be platform interoperability; this will unlock all clinical data offered in various platforms and facilitate even richer data sharing, analysis, and insights that will improve patient health. CSDR is actively pursuing platform interoperability as an important next step for advancing human healthcare.