You Can't Regulate a Market Into Existence: Interoperability Depends on Multi-Sided Platforms
Updated: Feb 10, 2020
Mandated by the by the 21st Century Cures Act, the Trusted Exchange Framework and Common Agreement (TEFCA) is the government’s latest guidebook intended to make nationwide patient data exchange and healthcare interoperability a reality.
It’s an admirable goal, but TEFCA and its sister rule on information blocking (now in review with the Office of Management and Budget) may be to the healthcare industry’s detriment. Why? The policies put a number of undue constraints around a complex ecosystem of networks that need flexibility, variability, and time in order to optimize.
In addition, the guidelines ignore the fundamentals of network development, the importance of network effects, the need for both time and constraints in their development, and the critical role that network operators play in facilitating interoperability.
Why are network effects and network operators so vital when it comes to achieving nationwide healthcare information exchange? And, what steps should the Office of the National Coordinator for Health IT (ONC) take to ensure the guidelines are best positioned to achieve their intended goals going forward?
TEFCA’s goal vs. market demands
After the ONC’s release of the second draft of TEFCA in April 2019, the public comments submitted underscored the industry’s extreme trepidation about the framework’s ability to achieve its goals.
Stakeholders called out the proposal for being too broad and vague, in addition to its unrealistic timelines, lacking standards, and unclear incentive structures. Moreover, because requirements for API-based exchange (read: modern architecture for data exchange) and Fast Health Information Resources (FHIR) have been stripped, the primary focus will be on improving document-based exchange – two steps forward, one step back.
Through the twin policy levers of TEFCA and the proposed information blocking rule, it appears the ONC is attempting to regulate a market into existence. Outside of the growing patient desire for access to their own health data, sufficient industry demand for this market is lacking, as evidenced by the industry’s (un)willingness to pay.
Case in point: Almost a decade after the U.S. government invested more than $35 billion in the creation of the meaningful use program to spur EHR adoption, seamless exchange of health information between providers is a rarity -- despite the fact that the majority of providers are utilizing an EHR.
Perhaps the biggest barrier to TEFCA’s success, however, is a lack of understanding about how multisided networks work (e.g., health information networks that connect EHRs, health systems and physicians), which is a critical aspect of achieving healthcare interoperability.
Understanding network effects: the key to achieving nationwide interoperability
Unlike traditional software or technology companies, multisided networks and platform organizations don’t produce anything new or novel. Rather, networks and platforms create value by bringing different types of users together to facilitate the efficient exchange of information, goods, or services. (Think Amazon, LinkedIn or eBay.)
Their success is also reliant on a phenomenon known as network effects -- that is, the more users using a given network, the more value there is to the network. Similarly, as users drop off, there is less value to those remaining, who are more likely to drop off as well.
In its current form, TEFCA is not set up to achieve its objective of enabling nationwide interoperability because it does not take into consideration the importance of network effects, the difficulty of achieving network effects, nor the underlying economics that drive how multisided network and platforms make decisions.
Where TEFCA misses the interoperability boat
Two of the biggest disconnects between TEFCA’s design and the problem it’s supposed to be solving map back to time and constraints – necessary ingredients for network optimization.
Time: ONC plans to finalize TEFCA and the overlapping information blocking rule by early 2020, with the first group of Qualified Health Information Networks (QHINs) up and operating by next August. This accelerated timeframe is not long enough for proper network set up and development, let alone network growth and optimization. For instance, e-prescribing between doctors and pharmacies is now routine (85 percent of all prescriptions were e-prescribed in 2018), but Surescripts, the health information exchange that routes the majority of e-prescriptions in the country, was founded in 2001.
Also, e-prescribing has benefitted from various federal programs promoting adoption and use of health technology. Beyond this, Surescripts reports in 2019 that it has worked with pharmacies and EHRs to improve prescription accuracy markedly, a process that has taken years. Likewise, CoverMyMeds reported that in 2018 (seven years after its founding) only 24 percent of physicians used an electronic prior authorization solution exclusively to manage prior authorizations.
Constraints: ONC’s ‘information blocking’ proposed rule is intended to force health systems, EHRs, and health information networks to “open up” their systems to improve the exchange of health data. However, networks compete on all sorts of measures, including price, business model, quality, governance, and how they uniquely facilitate the exchanges between users.
By attempting to commoditize and regulate the means and methods of health information exchange (e.g., network fees), ONC is establishing a network value “ceiling,” and short-circuiting the actual way networks develop and compete.
Furthermore, by removing economic upside for network developers/operators, rather than having the “best network win” in the market, the industry may be left with: (1) a handful of “zombie” networks (unable to invest in improving the quality and value of exchange), or (2). a monopoly or duopoly situation, as smaller regional health information networks find themselves unable to compete with larger, national networks.
Nationwide interoperability depends on multisided network development
Healthcare organizations that provide legal and connectivity frameworks (such as CareQuality and CommonWell) and those that both set and maintain standards (such as HL7 and NCPDP) play a very important role in consensus-building and network governance functions. However, these organizations do not operate or maintain actual network infrastructures, nor do they provide operational or technical support.
Only multisided network and platform operators are uniquely positioned to play this role and address the myriad challenges that exist when trying to facilitate health information exchange in a way that creates value for network users, including:
Offering global, end-to-end technical support and coordination protocols.
Going beyond network governance to focus on improving network operations and optimization, including user experience, transaction processes, value delivered, etc.
Providing version control services such as ensuring that two participants using different versions of a standard can still “talk”.
Certifying technology and ensuring ongoing participant compliance with technology and processes.
Providing quality oversight and a continuous quality improvement program.
These services that occur on top of the actual platform may seem extraneous for local hospital-to-hospital information exchange but are critical when health information exchange starts to scale and becomes many-to-many.
Consider something as simple as a hospital sending a patient discharge summary to a doctor that may practice in multiple offices and work with different EHRs. If the doctor cannot easily locate the record, who should be responsible for figuring out where the discharge summary is? Without a centralized health information network operator connecting all parties, this simple exercise becomes a labyrinthine, trust-eroding, bureaucratic nightmare.
In healthcare, there are few established national health information networks (Surescripts and CoverMyMeds), and several innovative firms building multisided networks to help make the healthcare system more accessible, efficient, agile and affordable. Examples of these firms include Redox (a platform connecting health systems with innovators), Swift Medical (a digital wound management platform), and Sempre Health (a platform matching payers and pharma brands to improve access to prescription cost savings).
Five areas of focus: Reframing TEFCA with network effects in mind
There are five areas that ONC should focus on to ensure that the guidelines being put forth are setting the industry up for collective success, including the following:
Continue the trend towards value-based payments: Healthcare stakeholders will be more motivated to use technology and interoperability solutions when they solve a business problem. For instance, putting health systems at financial risk for Medicare beneficiary readmissions within 30 days encourages those health systems to coordinate care with primary care providers, home health agencies, and other post-acute providers. Coordination done well and at scale will require these various stakeholders to work together to find effective interoperability solutions for individual use cases.
Narrow the focus from interoperability’ to ‘individual interoperability use cases’: Successful multisided networks recognize the importance of solving for a specific problem or “core transaction”. Part of the challenge of CMS and ONC’s current efforts is that the definition of ‘interoperability’ is too broad and encompasses too many different things; so long as interoperability efforts are trying to solve every stakeholder’s myriad problems, they won’t do any of them well. This one-size-fits all approach also ignores the lessons learned from successful networks in health care such as Surescripts and CoverMyMeds, which focused on individual use cases and core transactions.
Look to what has worked in the past with e-prescribing: The federal government does have a tremendous success story with interoperability in e-prescribing. In 2008, following public and private efforts to advance and pilot e-prescribing standards to demonstrate industry technical readiness, Congress enacted the Medicare Improvement for Patients and Providers Act (MIPPA), which provided for a sliding scale of financial incentives (carrots and sticks) over time for physicians to adopt and use certified e-prescribing software. The impact of this modest program was enormous in terms of advancing e-prescribing adoption and has contributed to an estimated $140 billion to $240 billion in savings for the healthcare system. Importantly, Congress’ incentives came after federal and industry efforts had proven out the maturity and value of e-prescribing, and there were viable network operators (including Surescripts, RelayHealth, and Change Healthcare) in place to provide ecosystem support. Rather than continue to spread its effort and resources across unproven interoperability use cases (and demanding industry participants do likewise), CMS and ONC would be wise to advise Congress that further efforts might benefit from narrowing their respective roles.
Focus on providing answers to intersystem and internetwork technical issues: Federal agencies have been barred from spending any funds to investigate or create a universal patient identifier. Yet the seemingly simple task of matching patients from different systems has proven an incredibly vexing issue that, unresolved, disallows (or limits in large part) system-to-system interoperations. Industry-led efforts (including for profit and non-profit actors) have not yet led to a commonly accepted standard, and many industry participants have voiced support and desire for the federal government to play a role in the conversation. Issues such as patient matching, security, and privacy need to be addressed across interoperability use cases and would benefit from federal leadership.
Allow networks to compete (and on business model and price, too): While it may be undiplomatic to assert that health information networks should charge fees and be allowed to profit from enabling interoperability use cases, the ability of networks to compete freely -- and for the market to signal which networks it prefers for which use cases -- is important to realizing robust adoption of interoperability use cases.
Creating a single onramp for nationwide health information exchange is an aspirational and altruistic goal. A rushed and “one size fits all” approach to interoperability may result in a set of networks that can indeed “talk to one another” and are widely adopted but are not widely used because they do not create enough value for users.
Which brings us back to time and constraints: two necessities for network growth and optimization. If given the time and the ability to operate without constraints (with the opportunity to fail quickly, iterate, and change policies accordingly), multisided networks and platforms could be the prescription that nationwide interoperability needs to succeed. Only time -- and network effects -- will tell.