We Need a Next-Generation Patient Registry: Introducing the Unified Patient Registry
Advances in medical technology and informatics profoundly expanded the ways we can measure and process real-world patient data. These advances create opportunities for accelerated research and development that can improve healthcare outcomes for patients. For example, the pharmaceutical industry faces pressure to release newer, better drugs to market as quickly as possible. Innovations at the intersection of informatics and medical tech can accelerate approval processes, improve clinical trial site choice and reduce site numbers, speed up trial enrollment, and shorten trial lengths, thus reducing the cost of clinical trials. These cutting-edge research methods can be used in most scientific research settings even after regulatory approvals, including phase 4 clinical studies, real-world evidence (RWE), and drug repurposing.
However, barriers to data access impede our ability to harness these advancements. Many hospitals are reluctant to share patient data due to privacy concerns. Acquiring a large amount of data doesn’t necessarily mean it is useful, either — the data collected must be fit for use.
Siloed data is data that is stored in different organizations and different systems, and privacy laws make it harder to compare data across silos, reducing its generalizability and utility. Data bias may be reflected in limited sample sizes or underrepresentation of subpopulations, whether demographics, disease types or disease stages, all of which reduce the statistical power of data and affect outcomes.
Secure AI Labs’ Unified Patient Registry resolves the issues described above. The registry will allow the researcher access to a large volume of quality, current and diverse data. It provides researchers not only quantitative improvement but also qualitative improvement.
Federated learning — also known as collaborative learning — is a machine-learning technique that enables algorithms to access data from multiple databases at the same time without pooling them into a central location. The Unified Patient Registry, which is powered by federated learning and analytics, allows researchers to train models across traditionally siloed data. This means that researchers can perform analytics on data from multiple hospitals to predict outcomes while preserving patient privacy and researcher algorithm rights.
Good governance of the Unified Patient Registry is vital to its success, and advocating for the best interest of the patients is extremely important. Since patient information never leaves the data owner’s control in the Unified Patient Registry, data owners can easily confirm the implementation of good governance standards, thus ensuring patient data remains private and secure.
Privacy is built into the framework, and the registry’s early success with SAIL’s first collaborator, the Kidney Cancer Association, will prove its efficacy and allow expansion to more patient advocacy groups, hospitals, and pharmaceutical companies. This expansion allows access to a patient’s full story — from diagnosis to treatment and beyond. At scale, those stories can ensure that research is more comprehensive, thus allowing patients to receive better care, more quickly. As research groups grow and connect with one another, the resulting network will transcend disease or condition, and provide researchers access to universally useful data.
SAIL’s Unified Patient Registry allows data owners to ensure data privacy by maintaining data ownership through the entire research process and solves the issues of the siloed structure of data today to bring true innovation to all aspects of healthcare including academic and public health researchers, pharmaceutical companies, and patients.