Tech Snapshot® captures today’s cutting-edge tools and technologies that will help drive drug discovery tomorrow. This installment was written by xCures, whose mission is to improve cancer outcomes using A.I. and predictive modeling.
With the explosion of AI, the emergence of next generation sequencing (NGS) and the seeming ubiquity of electronic health records, the promise of clinico-genomic real-world data (RWD) has never been more intriguing in the world of oncology.
And yet, research projects capitalizing on RWD from claims data and/or EMR data seem to be getting more and more frustrating. Blame it on the curse of dimensionality, (or as Jerry Jones might say, “circumcising the mosquito.”) As products get more targeted for hyper-specific patient populations, the data needed to answer questions about those populations becomes more difficult to actually find amidst the ocean of data (both structured and unstructured) available in retrospective real-world datasets. It is the abundance of data that can make it quite difficult, costly and time consuming to mine datasets to answer research questions about hyper-tailored subpopulations.
Determining if a dataset exists that has the right patient volume to answer your question is the first hurdle that burns time and money, and even more time and money is spent conducting a feasibility assessment of that dataset to determine the potential sample size that meets the inclusion/exclusion criteria for that research question. This exercise routinely manifests in spending precious resources only to have researchers come back and tell the sponsor that there is not enough data to answer the research question with any statistical significance. It becomes a needle-in-the-haystack issue.
Generating insights from RWD is even more challenging when trying to answer questions about recently launched products. When a product launches, companies want to know how that product is performing in a real-world setting. Outside of setting up a registry (which can be extremely costly), companies will rely on retrospective real-world data to gather insights on how these products perform in the real world. But waiting for data on recent products to appear in retrospective RWD sets can be an exercise in patience. Companies generally wait at least a year – and often much longer – for data on new products to become robust enough in these datasets for their commercial and medical teams to use.
Wouldn’t it be great if, instead of spending resources searching datasets for the right patients, we could prospectively target patients that we know would fit the inclusion/exclusion criteria?
The first challenge facing prospective real-world data capture is patient participation. Will patients be willing to participate in observational research? Consider these statistics from a recent survey by COTA, inc. of more than 1,100 Americans who have either had cancer or had a family member with cancer:
- Approximately 85% of respondents would agree to share their anonymous data if asked by their doctor
- 86% of respondents believe oncologists should be actively discussing the value of sharing data as part of patient interactions
- 89% support all cancer patients sharing their health data anonymously for advancing treatment research and discovery
- 87% indicated they wouldn’t care if their data had already been anonymously shared.
Of course, patients also own their data, and through HIPAA third party right of access, they can consent their data to third parties, which would enable them to share their data for research purposes. If presented with the proper information, and the proper mechanisms through which they can share their data, it certainly appears that patients would be willing to participate in prospective observational research.
If we can identify early on the patients taking new, exciting therapies and empower them to share their data, we can begin tracking them in a real-world setting and begin gathering real-world outcomes data from them much earlier than we’re typically used to. This is a win-win for everyone. Manufacturers can get a much clearer picture on the value of the therapy in specific populations, adjusting their medical and payer strategies earlier. Shared knowledge of real-world performance will only improve how these therapies are used in real-world populations by physicians, which ultimately benefits patients.
But how can this be executed? xCures has solved this by creating a value exchange with patients. Patients consent into an observational registry (XCELSIOR, NCT03793088). In exchange for access to their medical records, patients receive a structured Cancer Journey, receive personalized treatment option recommendations to consider with their physicians and access expert insights from KOLs at leading institutions through xCures’ Virtual Tumor Boards and shared learning networks. Through this registry and consenting of records, patients are followed prospectively, generating valuable longitudinal real-world data.
It is possible to capture real-world data from patients prospectively. All we need to do is identify them and empower them to participate in observational research. The motivation, and the mechanism, exist.
Here at xCures, we offer a free service for patients in exchange for access to their full medical records, which we are able to use for observational research. If there is a patient population of interest that you need longitudinal real-world data from, please reach out to xCures to discuss how to begin a prospective real-world data capture engagement.
For more information, please contact Max Goldstein, Vice President, Research Partnerships. email@example.com