We need health data banks not unlike what we have for organ donors
by Eric Lefkofsky
The practice of medicine often requires making the most of a tragic situation, learning from illness to help people stay healthy. Look no further than the act of organ donation. The donor’s selfless act can transform a loss into hope for the sickest patients—those identified most in need of an organ transplant by the United Network for Organ Sharing (UNOS), the nation’s transplant system. Every organ donor has the potential to save up to eight lives and improve quality of life for 75 others. So many of us check the Organ Donor box on our driver’s license, often without giving it much thought. One of medicine’s great paradoxes is that the end of one life can spawn the chance for others to live. When we think of the benefit of a breakthrough therapy, or the clinical impact of an individual life, it’s always a story of one-to-many.
I have spent the better part of the last decade building a dedicated team focused on advancing precision medicine through data and technology—which is itself a one-to-many model. As the entire industry has made immense progress in leveraging data to inform medical research and decisions at the point of care, it dawned on us: Why should data be treated any differently than organs?
When data from diverse sources—including electronic healthcare records, imaging from pathology slides and radiology scans, and results of genomic sequencing—are combined and de-identified across the healthcare system, we can significantly advance therapeutic discovery and development, and find better treatment options. This longitudinal view of the current landscape of cancer, of what’s working and what isn’t, can help researchers identify gaps where novel treatments are most needed. Think of it as a system like UNOS, but for data. Real-world evidence is applied to guide other patients’ treatment more precisely, and more effectively. Applying the data lens to healthcare, it’s fair to say that the history of medicine is the progressive application of knowledge to biology. In our own time, patient data has become the most sophisticated expression of applicable knowledge. So, what can we do to make it a model of many-to-many?
A decade ago, if a patient responded well to a novel therapy, this was the limit of the learned information. The doctor couldn’t be sure that the next patient to walk through the door would experience the same. But now, with data-driven insights, made possible by the use of artificial intelligence that allows researchers to quickly interpret vast amounts of clinical information to derive unique insights, there is so much more to inform treatment choices and decisions. Connections can be made from this data that might explain certain positive outcomes among patients, whether it be specific tumor biomarkers or hereditary signals. The more de-identified patient data that can be collected, organized, and harmonized, the greater the chance that we’ll find the next breakthrough in care, and, potentially, the difference between life and death for future patients.
Like organs, healthcare data is powerful. When appropriately secured and de-identified in compliance with applicable laws and regulations—as is the case with organ donation—data can enable doctors and researchers to learn from the experience of today’s patients to help tomorrow’s. We are already seeing it successfully applied each and every day, when physicians leverage data to personalize each of their patient’s care by identifying targeted therapies and promising clinical trials.
So, how do we get there? As UNOS reports, organ transplants were managed entirely by individual hospitals only until the 1970s; however, that’s still where we are with data, which is managed at the individual health system level, with little infrastructure to enable de-identified data sharing at a national scale. We can only best use data if it’s harmonized across the industry, and it is inclusive of all patient populations. We need to remove technological barriers (e.g. interoperability, data standardization), share data across hospitals and healthcare institutions, and work together to create insights that will help generations of patients.
Health data can create just as much value as organ donation, especially in oncology. But, while a cancer diagnosis will always be devastating, perhaps we can advance care driven by the insights from collective de-identified patient data to turn those experiences into hope.
Eric Lefkofsky is the founder and CEO of Tempus.
Source: Fast Company
