Whereas access to representative data has long hamstrung research and development of treatments and algorithms, healthcare and life sciences companies can now access the structured and unstructured notes for nearly a third of the U.S. population.
OMNY Health has now onboarded the de-identified data of 100 million patients onto its data platform and achieved one of the company’s key goals in its seven years of operation. The breadth of the data could power rare disease research, bring new drugs to market and power artificial intelligence models with a nationally representative data set, according to the company.
OMNY’s living data layer reflects the diversity of the nation, including demographic characteristics and types of care settings.
“Whether you go to a big academic medical center in a major city or you go to the community ophthalmology practice, it's different care, it's different resources, it's different access points, different types of providers,” Mitesh Rao, M.D., CEO and co-founder of OMNY Health, said in an interview. “When we think about data that really represents the country, you want to be able to capture that. The second dimension is one we always think about, which is populations. How do you make sure that every geography, ethnicity, race, every economic situation is represented? And so that's been our real big effort [to] get to this 100 million mark.”
By the end of the year, OMNY wants to get more than 125 million patient lives on the platform.
OMNY has achieved its scale by making its living data layer free for healthcare organizations to join and slowly gathering the data one health system at a time. While early on, investors warned that the approach would fail, emergency medicine physician Rao said he wanted to stick to the democratic approach, knowing that access to clean data would be transformative for healthcare.
It took two years for the company to get its first provider systems engaged. Five years later, the OMNY data engine powers 46 health systems, hundreds of hospitals and 650,000 providers. Systems like St. Luke’s University Health Network, Bon Secours Mercy Health and Baptist Health System KY & IN have transformed their data with OMNY, executives said.
The white-glove service can now onboard a health system to the platform in four weeks, compared to six months at the beginning of the company’s journey.
“We could have been bigger, faster by buying access to data,” Rao said. “We could have raised crazy amounts of money. We could have partnered with a lot of people to do things that I think would have cut corners. We never cut corners. We always stay focused. We only do partnerships around supporting development and research that we know will actively benefit patients.”
OMNY transforms the data into a common format, making them into regulatory- and AI-grade data. Health systems, researchers and developers that want to leverage the data to create AI systems or develop treatments pay a subscription model access fee for the data. The healthcare delivery organization that owns the data then gets paid a royalty for the use of their data.
“Everyone's data is completely different within a health system, between the different hospitals, you'll see data being incorporated in the different floors, like the cardiovascular unit, the general ICU and the cardiovascular ICU, may code things differently, like you just will see so much difference that you look at that and then you just have this messy pile of data that's not really usable,” Rao said.
The company started with a focus on pharma and specific therapeutic interventions. Now, the secure data repository has grown so large that it has helped bring rare disease products and biologics to market. Hospitals can also use their own data for quality and safety reporting or training their own large language models in-house.
“A lot of our major health systems now are partnering with biopharma through our platform to do things like quality improvement initiatives, build new care pathways, clinical trials, publish papers. Our goal is to connect both those groups,” Rao explained.
Many healthcare researchers funded by the National Institutes of Health have had their projects terminated or their funds frozen since President Donald Trump took office, especially if their work related to race, HIV or COVID-19. Rao sees a role for OMNY to connect researchers to private funding streams through the organizations that are partnered with the company.
One of OMNY’s core principles is letting organizations retain ownership of their data. The platform is also a secure environment in which to use the data and can protect organizations from the risks of outsiders operating directly within their systems.
“I remember sitting at Northwestern, at Stanford, I got pitched by these platforms who were before OMNY, who'd come in and they'd say, 'OK, here's a cool platform with analytics, and you can do your research on here, and you can get access to stuff,' and on the back end, in the contract, they take the rights to the data and they'd sail off to the sunset," Rao said. "I wanted a much more democratic approach, where everybody benefited.”
Healthcare data are also difficult to parse because much of the context about patients is buried within unstructured fields of text. It’s valuable for researchers, for example, to know whether a treatment failed because a patient could not access reliable transportation to a pharmacy.
OMNY links the data from structured and unstructured fields in the electronic health record to give a fuller picture of patients’ circumstances. It also provides up to eight years of historical data for patients.
“We use … more than six and a half billion unstructured notes, to really understand what a patient is going through and help align that into the research, because that voice of the patient gets lost. If you just look at a medical record without the unstructured data, you don't understand the patient's context.”
Rao said the platform is “infinitely scalable” and could eventually support the data of all U.S. patients. For now, OMNY continues to onboard health systems with the goal of doubling its number of partnerships by the end of the year.