The Mayo Clinic will target drug discovery in its latest effort to improve healthcare through data insights, the Rochester, Minn.-based health system said Tuesday.
The clinical data analytics platform—Mayo Clinic Platform’s first project since launching last year—is a “mechanism by which the data of the past can be used to improve the care of the future,” said Dr. John Halamka, president of Mayo Clinic Platform. While drawing out insights from historical data isn’t new in healthcare, Mayo believes the technical architecture underpinning the project will make the process more efficient, particularly for collaborators from outside the system.
The platform’s mission is to create consistent sets of technology tools and policies for different types of collaborations.
For the new clinical data analytics platform, that meant setting up cloud containers that house de-identified patient data, which providers, payers and pharmaceutical companies outside of Mayo can link up to via application programming interfaces, as well as establishing standard templates for compliance and legal agreements.
It can take health systems more than a year to get a partnership up and running without these standardized tools and processes, according to Halamka.
Mayo Clinic Platform’s first partner on the clinical data analytics platform is Nference, an software startup that Mayo is an investor in.
Nference develops analytics, machine learning and natural language processing tools that “augment” the work of data scientists, in an effort to help pharmaceutical companies and research organizations conduct “research at scale,” said Venky Soundararajan, Nference’s founder and chief scientific officer.
Under the partnership, researchers from Mayo and Nference will aim to apply Nference’s tools to de-identified data held in the clinical data analytics platform to draw out insights, such as identifying biomarkers for new drugs.
Mayo Clinic said it hopes to work with pharmaceutical companies to commercialize new therapies, said Andy Danielsen, chair of Mayo Clinic Ventures, the system’s commercialization arm. Mayo itself wouldn’t commercialize those therapies, though the system could receive royalties from insights generated on the platform.
While efforts to discover new therapies is an initial focus of the clinical data analytics platform, Halamka suggested other uses could include allowing providers to query the platform to gain insight into patients exhibiting the same symptoms as a patient they’re treating, or insurers to query the platform to understand which patients respond best to an expensive therapy.
Mayo plans to only charge for-profit companies for accessing its clinical data analytics platform. (Mayo didn’t charge Nference for access because it is a partner on the clinical data analytics platform.)
Any revenue generated from the platform would be re-invested into Mayo’s clinical practice, research and education work, Danielsen said.
The clinical data analytics platform is hosted on Google Cloud, as part of a partnership with Google that Mayo Clinic announced in September. Danielsen stressed that Google, for now, is a “passive data-storage company” for the platform.
“Google doesn’t get to see any of our data, they don’t get to access any of our data there,” he said. “They’re like the bank vault.”
That could change in the future, as Mayo continues to strike agreements with new partners. Google could access the data if the system taps the tech giant to work on a specific project, Danielsen said.
Google, which has had a messy history with data privacy in the consumer world, has been embroiled in controversy over its healthcare work since late last year, when details emerged about a partnership it struck with Ascension.
To tackle privacy concerns, Halamka said Mayo Clinic Platform has instituted additional processes that cordon off third parties from being able to access patient data.
Rather than sharing de-identified patient data directly with a third-party partner, Mayo Clinic Platform will de-identify patient data in a cloud-based container—a type of technology that bundles an application’s code in a way that allows it to be deployed elsewhere—that it operates. Outside partners would only be able to access the de-identified data once it’s placed in a separate container.
That data doesn’t leave the container, since third parties would also run algorithms inside of it.
“The de-identified data never leaves,” Halamka said. “What leaves are answers.”