Finding a Fair Price for Data

Stakeholders in the healthcare ecosystem that own or manage large data sets, notably integrated delivery systems, payors, and governments, face the same dilemmas: in a new “data as gold,” era, how to quantify the value of an intangible asset in the context of a healthcare setting?

The exercise has important legal implications because putting a specific quantitative value on data helps protect companies from violating the Federal Anti-Kickback Statute (AKS), which prohibits knowingly offering, paying, or receiving renumeration to induce referrals or items or services paid for by a federal healthcare program, according to Roger Cohen, a partner at Goodwin Law, who gave a presentation, “The Fair Market Value of Electronic Patient Health Data,” at the American Bar Association Healthcare Delivery and Innovation Conference in Chicago in September.

Along with fellow panelist Christopher W. David, a managing partner and specialist in healthcare data valuation at HealthValueGroup, a healthcare valuation and transactions advisory, Cohen noted that medical device companies—and others—are taking a legal risk if they substantially pay more than fair market value for data or other assets to healthcare systems with the goal of improving or sustaining sales of their goods and services to those systems.

“If I was a healthcare provider doing business with a device company [that has a significant data component] I’d want to be thoughtful about structuring any arrangement where data is a significant part of the arrangement and particularly thoughtful about anti-kickback compliance,” Cohen said. Ideally, he added, “You should have a valuation for every transaction.” As the value of the data component of deals grows, valuations become more important and the use of AI exponentially increases the need for access to data, David noted. “Healthcare data is going to be in tremendous demand because of the results it can produce from the algorithms and AI models.”

To comply with the law, compensation has to be consistent with “fair market value and not tied to volume or value of referrals,” Cohen said. Many transactions, particularly those involving SaaS providers, include services that are not reimbursed by the federal government and therefore do not run up against the anti-kickback provisions. Still, commercial reasons also drive the need for valuations and can provide a level of comfort that the transaction price is appropriate, and that a corporate board is meeting its fiduciary responsibilities. 

To know whether they are overstepping, however, dealmakers need to have an accurate assessment of the value of the data sets involved. But determining fair market value presents healthcare-specific challenges, which are becoming pressing as “leveraging data becomes an increasingly key part of business strategy and data rights are incorporated into commercial transactions, and financings, and becoming a potential point of friction,” Cohen continued.

Healthcare data is traditionally unstructured, fragmented, multimodal, and hard to integrate. Extracting, organizing, curating, and building the infrastructure necessary to market data is extremely expensive and, until recently, almost impossible to collect in a comprehensive way, although advances in data science and the introduction of large language models and artificial intelligence are facilitating the ability to capture unstructured data and integrate different formats, bringing down costs. Moreover, data transactions involving federal renumeration of any kind must comply with HIPAA (the Healthcare Information and Portability Assurance Act of 1996), which protects sensitive patient health information from disclosure without consent and establishes federal standards for securing that information.

Data ownership also remains a confusing area of law, and one that requires clarification for valuation and transaction purposes. The federal government does not regulate data ownership, but states are increasingly interested in doing so, with most presuming patients own their data. Ultimately, proprietorship is on a case-by-case basis, depending on the nature of the services, Cohen said in a follow-up interview with MedTech Strategist. “Ownership is a bundle of sticks—and an oversimplistic way to think about rights to data,” because it doesn’t preclude healthcare providers, manufacturers, or others from retaining custody of data, licensing data, or analyzing it, and obtaining intellectual property rights protecting any new products or services that arise from that work. “The question,” he posed, “is, do you have rights to de-identified or identifiable data, and additional rights to use that data to improve your products and services and commercialize those improvements?”

Continue reading here.