🎙️ Voice is AI-generated. Inconsistencies may occur.
news article
Based on facts, either observed and verified firsthand by the reporter, or reported and verified from knowledgeable sources.
Billions of clinical insights—once buried in an unusable format—may now be used to train large language models and inform clinical research studies.
The data ecosystem OMNY Health has incorporated 4 billion unstructured clinical notes into its network—allowing health care, life sciences and research organizations to glean insights from previously unusable data.
OMNY collaborated with provider organizations, including academic medical centers, to gather unstructured clinical notes and turn them into accessible information.
About 80 percent of medical data is unstructured, according to a study published in the journal Healthcare Informatics Research. Most electronic health records cannot manage this type of data, so it is "ignored, unsaved or abandoned" by many medical centers, per the study.
Dr. Mitesh Rao, founder and CEO of OMNY Health, told Newsweek that looking for insights in unstructured data is like "looking for a needle in a haystack."
De-identifying this data and translating it into a clean, usable format can give health care organizations a richer understanding of patient care journeys, according to Rao.
"The reason [unstructured data] is so powerful is that, historically, it's always been an area that we've struggled to get information out of," he said.
OMNY is linking the unstructured notes with its broader clinical data network, which all of its partners across the industry can access—including health systems, specialty health networks and pharmaceutical, biotech and AI companies.
The notes contain information on disease progression that could contribute to a deeper understanding of certain illnesses. They also include insights on different treatments, such as patient responses, adherence patterns and adverse events.
This new information could be used to inform innovation, speed up recruitment for clinical trials and train large language models, according to Rao. Provider organizations can also use the insights to inform care decisions internally.
In addition, the unstructured notes contain more information on social determinants of health. These could illuminate care gaps and inform health equity research.
Historically, unstructured data has been challenging to tackle. Rao compared making sense of data to making cookies—free-text clinical notes aren't flour, but unprocessed wheat.
"You have to do a lot to actually transform it into something useful," he said. "Cleaning it, structuring it, de-identifying it, using natural processing to pull out key insights—those are all pieces of work that can be done on the note itself."
"But turning [that note] into something that is research- and regulatory-grade, that a researcher can leverage, that AI can train on?" he added. "That's an entire process in itself."
OMNY used large language models and its own natural processing systems to complete the process, according to Rao.
Dr. Mark Townsend, chief clinical digital ventures officer at Bon Secours Mercy Health and Accrete Health Partners, the health system's digital holding company, said his organization will use the information to "move beyond surface-level analytics," personalize care and improve decision making.
"Unstructured data, like that found in free-text clinical notes, represents a treasure trove of untapped insights," he said. "By leveraging unstructured data, we can improve patient outcomes, drive operational efficiencies and foster innovation in health care."
Data management is one of the health care industry's biggest bottlenecks, and Rao believes this move is a step in the right direction.
"This is the first time that this type of data has really been unlocked from the traditional IT systems which are designed to lock data down," he said. "We see this data as a way to improve connectivity and collaboration across the industry. We see this as a way—all across health care—to be able to actually work together through that common language of data."
OMNY Health's data ecosystem currently includes clinical notes from more than 500,000 providers across more than 200 specialties. There are 43 provider organizations on the platform, whose total network covers all 50 states and more than 85 million patients.
