To some at a recent Capitol Hill panel discussion, the idea of turning patient and population data into on-the-ground treatments may sound like science fiction, but Dr. Summerpal Kahlon, director of care innovation at Oracle Health Sciences and Physician and Infectious Diseases Specialist at Veterans Administration Medical Center, says, “I think we are already there.”
Kahlon and other panelists at the event, entitled “Industry Intersect: Advancing Precision Medicine,” spoke about the many forms of data that can be used to unlock precision medicine—treatments tailored to specific groups and even individuals. The challenge is bringing all the necessary data together in the right ways to understand diseases and create personalized treatments that improve patient outcomes.
Innovations such as genomics, EHRs, population health analytics, and patient wearable devices, provide data that are leading to new ways of understanding disease and how the individual patient’s biology and lifestyle affect treatments.
However, amid all the new science and innovation around precision medicine, “we’ve lost some of the communication and coordination in the process,” Kahlon said. “We’ve moved into hyper-specialized, highly fragmented care environments,” he said. Different healthcare providers don’t communicate well, “so now it’s on me as an individual to figure out how to coordinate my care and move data around.”
In this environment, how can the healthcare industry bring isolated data together to create better personalized therapies and deliver better outcomes in more patient-empowering and innovative ways? Those questions took center stage at the panel event, which was presented by Oracle, Intel and Pfizer. Here are some of the insights and answers that medical experts, industry leaders, government officials and policymakers shared:
Data Tells a Story
With the advent of EHRs, genomics patient-monitoring devices, etc., more healthcare data is collected today than ever before. “The pieces now coming into play are really putting together a story around an individual patient, linking data together in a person-centric way,” Kahlon said. “The data starts to give not just a snapshot but a full story of what’s happening to a person over time.” But only if this data available and in a usable form to the clinician.
Genomics is a key driver for personalized medicine since it provides information about the individual’s DNA. A patient’s genomics can identify risk factors for disease, insight into which treatment is most likely to be beneficial, and more. Genomics is also, by far, the largest—and most unwieldy—piece of the big-data puzzle. “The big data of genomics adds a whole new complexity,” said Alice Borelli, director of global health and workforce policy at Intel. “Just to give you some perspective of how big this is--there are 14 million patients diagnosed with cancer worldwide every year--if we were to sequence the genes of these patients it would be 5.6 exabytes." That’s more data than would be generated if you put every word ever spoken by human beings into text form.
As approaches to care delivery evolve, healthcare systems must develop better tools to identify and leverage genomic signatures that align with disease and can guide treatment planning. It will be important to find ways to integrate this massive amount of genomic data into the clinical care systems in ways that enable personalized medicine to be practiced by a wider range of clinicians and with different training. Such a collection of patient information will enable precise, targeted care for an individual or specific group regardless of their care setting.
“For precision medicine to really develop, we need a personalized approach to patient care,” Borelli said.
Sharing the Wealth (of Data)
To enable precision medicine, data from many systems and providers need to be available—for clinicians treating patients and for analysts pursuing population health insights. Several obstacles stand in the way of sharing healthcare data on a large scale, including fragmentation or data siloes and the challenges of transporting, storing and aggregating large quantities of data. But the biggest concern in sharing healthcare data is trust, according to John Sotos, chief medical officer at Intel. The data system must be trustworthy: Patients have to trust physicians with what happens to their personal data, and physicians have to trust that when they do share data, they get credit for work they’ve done.
Many patients are also eager to share their own data—from wearables and other monitoring devices, and that data holds great promise, according to Craig Lipset, Head of Clinical Innovation at Pfizer. “When patients have access to their health data and the trusted ability to share, that will be one of the most transformational tools available for medicine development,” he said.
To create and preserve an atmosphere of trust, Sotos envisions a federated data sharing model rather than a centralized model—taking analytics to the network edge, at the collection point, rather than sending it to a central location. “If I send data to another location, I’ve lost control of it,” said Sotos. “If I keep my data right here, I retain full control and there’s only one copy.”
A federated analytics model ensures swift access to the data, says Yentram Huyen, managing director of Americas Health and Life Sciences Solutions at Intel. “Since you can keep the data local, you’re not moving it around,” Huyen said. “You’re actually sending the query to that data, so it’s being computed in place—that takes care of your data movement problem, which is huge, along with the trust issue.”
Huyen believes a federated data analysis model will accelerate the rate at which medical discoveries are made. “By enabling that level of federation, you can get innovations to the pharmaceutical companies, through FDA approvals, and then into the hands of physicians more quickly.”
Telling the Whole Story
Oracle and Intel have collaborated on healthcare solutions for two decades to enable high-performance analysis of large, complex volumes of healthcare data. Oracle Healthcare Foundation, a data solution for healthcare organizations, reflects these efforts to optimize system synergy in support of complex data analysis. “We’re working to bring data together and enable an ecosystem that can facilitate integration of all types of healthcare data—whether it’s collected from wearables, clinical encounters, or genomic testing,” says Sean Sigmon, Director of Business Development and Strategy, Oracle Health Sciences. “By working together, we can optimize the way our systems are engineered so the data collection and analysis process can be accelerated.” The result? Clinicians and caregivers are able to leverage more data than ever before.
“It starts with people,” Kahlon said. “This is about data reflecting people and their stories, with the goals of providingbetter access and better outcomes for all.”
To view video highlights and a full presentation of the panel event, “Industry Intersect: Advancing Precision Medicine,” click here.