From the J.P. Morgan Healthcare Conference, the healthcare system shares details about balancing innovation with the need to protect patient privacy.
As the healthcare industry struggles to find a balance between the desire for innovation and the need to protect patient data, Mayo Clinic has figured out a novel formula that it plans to employ with its first third-party partnership under the Mayo Clinic Platform.
Today, from the J.P. Morgan Healthcare Conference, the Rochester, Minnesota-based healthcare system announces a partnership with nference to accelerate drug discovery and development across the biopharmaceutical ecosystem. Yet before embarking on this initiative, Mayo Clinic hammered out a multi-layered plan designed to create a vault around patient identity, says Andy Danielsen, chair of Mayo Clinic Ventures.
"Mayo's mission statement is the needs of the patient come first," says Danielsen. "That's even reflected in our business deals."
Mayo Clinic Platform President John Halamka, MD, earlier told HealthLeaders that the health system's Platform will involve partnerships with multiple outside companies.
To ensure that patient data is protected in all of these endeavors, Danielsen says the organization determined the "model and structure that is best for patient privacy" and then would build "the best business we can around that." He describes the approach as "nontraditional."
"The hope," says Danielsen, "is that we've got a unique model that others are going to follow."
Mayo's data privacy model includes the following practices, according to Danielsen:
- Use only deidentified patient data that strips away nearly two dozen patient identifiers.
- Use independent experts to certify that the data has been fully deidentified.
- Eliminate the possibility of reidentification by employing a federated learning model. While others can query the data set, Mayo Clinic retains full possession of it and only sends participants the results.
- All parties are prohibited from pooling Mayo Clinic data with other data sources, such as consumer financial or geolocation data.
- Require contractual agreements with all involved entities, mandating they will make no attempt to reidentify patients.
"This is not the easiest way to run this business," says Danielsen.
The details of the data protection process were refined as the organization worked through details of its partnership with nference, a Cambridge, Massachusetts, augmented intelligence company, which Mayo Clinic selected as its first Clinical Data Analytics Platform—the Mayo Clinic Platform's first venture to improve healthcare through insights and knowledge derived from data.
The partnership will focus on identifying targets and biomarkers for new drugs, optimal matching of patients with therapeutic regimen, and real-world data and evidence applications, such as label expansion, postmarketing surveillance and drug purposing, according to a news release.
"The whole point of doing this is [to develop] new therapies faster, and get those therapies targeted to the right people," says Danielsen. These efficiencies, he says, should enable drugs to be produced at a lower cost by reducing failures in clinical trials.
Mayo Clinic's role in the process will be to piece together disease phenotypes and determine how to treat subpopulations of patients differently, which will provide clinical decision support benefits, says Danielsen. "It's a big vision," he says.
Last year Mayo Clinic launched the Mayo Clinic Platform, a coordinated portfolio approach to create new platform ventures and take advantage of emerging technologies, such as artificial intelligence, connected health care devices and natural language processing, according to the organization. The Mayo Clinic Platform is designed to include an ecosystem of partners that will complement Mayo Clinic’s clinical capabilities and provide access to scalable solutions.
“The hope is that we've got a unique [data privacy] model that others are going to follow.”
Andy Danielsen, Mayo Clinic
Mandy Roth is the innovations editor at HealthLeaders.
Patient's privacy comes first in all business ventures.
Novel model deidentifies data, employs a federated learning model to retain data control.
Reidentification is prohibited, and Mayo data cannot be combined with data from other sources.