By VINCE KURAITIS, GIRISH MURALIDHARAN & JODY RANCK<\/p>\n
This entry is Section 2 of part 3 of 3 in the series <\/em>Platforming Healthcare \u2014 The Long View<\/em><\/a>. This essay is the next in the series entitled \u201cPlatforming Healthcare \u2014 the Long View\u201d. The series presents a 30-Year Framework for Platforming Healthcare. An updated v2.0 of a graphic depicting this 30-Year Framework is shown above.<\/em><\/p>\n Today\u2019s post is section 2 and will continue to describe and discuss a potential successor to the EHR era \u2014 Unified Digital Health Platforms (UDHPs). Here\u2019s an overview:<\/p>\n Mayo Clinic Platform<\/p>\n Business and Strategic Implications of UDHPs<\/p>\n APPENDIX: Additional Readings on UDHPs<\/p>\n The Mayo Clinic Platform (MCP)was launched several years ago with the goal of building the future Mayo Clinic business model that could move beyond the bricks-and-mortar approach to traditional healthcare and open up new avenues for products and services. The adoption of a platform business model was considered essential to serving patients beyond the traditional Mayo Clinic geography as well as a way to incentivize innovation in AI and decentralized care in the home.\u00a0<\/p>\n A large longitudinal database with both structured and unstructured data provides a foundation for the MCP, particularly in respect to catalyzing innovation in clinical applications of AI. The database called Mayo Clinic Platform Discover<\/a> has over 7.3 million de-identified patient records that can be used for training AI models as well as in research and discovery for early-stage startups in particular who wish to join the MCP ecosystem. The dataset is referred to as \u201cData behind glass\u201d for the privacy and security standards that are needed to create the bedrock for a collaborative ecosystem.<\/p>\n Source: Mayo Clinic Platform Playbook<\/p>\n MCP is a three-sided market that has solution developers, data providers, and clinicians composing the three sides. MCP acts as the orchestrator of the ecosystem and additional partners such as Mercy have joined as data contributors in the MCP component called Mayo Clinic Platform Connect<\/a>. These are the primary components of the platform.\u00a0<\/p>\n The Mayo Clinic Platform Playbook identifies six key success factors:<\/p>\n A privacy-protecting, secure collaborative environment with de-identified data from global sources. Longitudinal databases of patient records are vital to clinical research and the development of new clinical decision-support tools, therapeutics, and digital health solutions. Privacy and security need to be maintained to protect the trust of Mayo Clinic\u2019s patients.<\/p>\n Breadth and depth of patient data must be sufficient to generate insights for both rare and common diseases.<\/p>\n Seamless capabilities to both ingest diverse data sources (e.g. -omic (genome, metabolome, etc.), EHR, wearables, social determinants of health (SDoH)) and to deliver actionable insights at the point of care.<\/p>\n Cutting-edge data science analytic tools, robust computing power, and uniform data standards.<\/p>\n Strong governance to assume security, scientific validity, interoperability and validation of technologies fit for purpose.<\/p>\n Pathways for commercialization of effective, validated solutions.<\/p>\n MCP vets startups and more mature technology vendors through a process that can begin with providing access to the longitudinal database for developing and training models (early-stage startups) to scaling solutions across MCP and affiliated hospitals with Mayo Clinic. The robust governance structure and integration with MCP makes scaling into other systems much more feasible.<\/p>\n AI governance is a core component of the MCP and why they were one of the original sponsors of the Coalition for Health AI (CHAI) to bring together leading industry players to create the standards for responsible AI across validation, explainability, and transparency.\u00a0 Ensuring that AI tools have been rigorously validated is necessary for clinician adoption of AI as well as maintaining the trust of patients. These standards act as a kind of \u201crules of the road\u201d for technology solution providers on the MCP.<\/p>\n <\/span><\/p>\n MCP utilizes ServiceNow\u2019s Application Portfolio Management (APM) solution that enables users to create a comprehensive inventory of all applications used by a company. Many health systems struggle to deal with the accumulation of applications over time and this creates security issues, redundancies, and risks to hospitals. The APM solution facilitates integration of application portfolios with clearer lines for ownership, responsibility for maintenance and decommissioning, and administrative roles.\u00a0<\/p>\n For MCP this type of solution offers a myriad of advantages when orchestrating an ecosystem of technology vendors and applications that are integrated with Mayo Clinic\u2019s digital infrastructure (EHRs, pathology and radiology applications, etc.) and beyond when the MCP partners with both technology vendors on their platform and external hospitals and systems.<\/p>\n The challenge of maintaining trust in a UDHP is considerable. In the MCP case, one can find technology solutions that can have a platform business model that engages with multiple other technology vendors. This creates a \u201cplatform of platforms\u201d scenario with considerable complexity and can carry with it cybersecurity risks if collaborators of these external platforms have security vulnerabilities or challenges with interoperability. Therefore a rigorous vetting process with standards for not only validation of AI models but inclusive of integration into the MCP network.<\/p>\n UDHPs will reshape many competitive dynamics of healthcare.<\/p>\n In the short- to medium-term, UDHPs will be complementary to EHRs. EHRs are deeply entrenched in today\u2019s healthcare delivery systems.<\/p>\n But as noted earlier, the roots of today\u2019s EHRs are as systems of record<\/a> to document claims submissions. To be fair, Epic, Oracle, and other EHRs are attempting to move beyond their historic narrow scope.\u00a0<\/p>\n But, the long-term prize is to become the system of action that automates interoperability, data sharing, and workflows within and across healthcare organizations. The BOI Strategy Playbook<\/a> depicts the fundamental platform shift:<\/p>\n Health interoperability guru Brendan Keeler<\/a> has noted that it probably won\u2019t be another EHR that displaces Epic:<\/p>\n Disruptive software change for B2B SaaS takes decades across all industries for enterprise size clients. Meaningful disruption is not a pure play competitor but the end result of an entire platform change (Windows was disrupted by mobile, not by another PC OS or antitrust). What disrupts an incumbent EHR won\u2019t be another EHR.<\/p>\n Oracle\u2019s new, next generation EHR should be characterized more as a UDHP than as a traditional EHR. Writing in Cloud Wars<\/a>, Bob Evans makes a case for Oracle\u2019s strategy:<\/p>\n \u2026 the bromide about how \u201cwhat got you here won\u2019t get you there\u201d will provide a stark example of why niche providers that dominated the on-prem world will be overmatched and pushed to the periphery in the modern world of cloud and AI. Larry Ellison has committed the full resources of his entire company to ensuring that Oracle Health can address not only EHRs and adjacent segments but also the entire $16-trillion healthcare industry\u2026.Oracle has been dealing with everything from software integration to massive global deployments to complex governance issues for the world\u2019s largest and most-complex organizations for more than 30 years.<\/p>\n EHRs likely won\u2019t go away, but they\u2019re at risk of becoming commoditized as the value-creation and value-capture opportunity shifts toward UDHPs.<\/p>\n While private health systems are likely early customers, UDHPs can be adopted by or extended to many other sectors or subsectors of healthcare. Let\u2019s consider some examples:<\/p>\n Consulting firm Z\u00fchlke<\/a> described how UDHPs can be deployed by life sciences companies.<\/p>\n Because life sciences companies\u2026 are not tech companies, they will usually want to consider licensing a digital health technology platform in the form of a platform as a service (PaaS). This enables them to sidestep some of the complexity of building their own platform and focus on their core business.<\/p>\n Salesforce<\/a> described how its UDHP offering is suitable for Med Tech companies.\u00a0<\/p>\n Many MedTech companies rely on multiple technology systems that weren\u2019t designed with customer engagement or the ability to work with each other in mind. It\u2019s essential to integrate this into a unified experience. A single platform that is easy to manage, can grow with your business, and can integrate with existing data and homegrown systems can power your organization more effectively.<\/p>\n Government organizations are candidates for UDHP adoption. An Australian public hospital system<\/a> deployed ServiceNow\u2019s UDHP.<\/p>\n Better\u2019s UDHP<\/a> is the underlying technology supporting the OneLondon<\/a> shared care planning solution across the city of London. The\u00a0 \u201cend-of-life Universal Care Plan (UCP) \u2013 is supporting a population of approximately 10m people, across 5 Integrated Care Systems, over 40 NHS trusts, and 1,400 general practices.\u201d\u00a0<\/p>\n A report<\/a> for the Norwegian Directorate of e-health illustrates how a UDHP could go beyond EHRs to enable a broader digital care delivery system:<\/p>\n Finally, we see the potential for the creation and adoption of UDHPs that cater to particularly large and\/or complex subsectors of healthcare delivery. Medical imaging is one example where a UDHP might be deployed by an imaging center, a large radiology group, or even within a health system.\u00a0 According to NetApp<\/a>, \u201cmedical imaging comprises 90% of medical data [storage] worldwide.\u201c Of the over 1,000 FDA-authorized AI medical device algorithms<\/a>, approximately 75% relate to medical imaging.\u00a0<\/p>\n While we foresee that AI will play a key role in UDHP adoption and implementation, the specifics are unclear and up for grabs. We anticipate that over time AI will be embedded in every layer of healthcare IT and will be applied to a wide range of administrative and clinical workflows \u2013 both within and across organizations.\u00a0<\/p>\n An example \u2013 in the past, there have been huge battles over user interfaces:<\/p>\n In the 1980s \u2013 the GUI war between Apple and Microsoft<\/p>\n In the 1990s \u2013 browser wars between Netscape Navigator and Microsoft Internet Explorer.\u00a0<\/p>\n In the 2000s \u2013 phone operating system wars amongst a dozen contenders resulted in huge wins for Apple iOS and Android<\/p>\n Another user interface battle is shaping up for clinician mindshare. AI will play a major role.<\/p>\n Today\u2019s EHRs are renowned for their poor UI and UX. EHR incumbents are scurrying to build more user-friendly AI interfaces. Epic has 125 AI-enabled features<\/a> that are live or in development.<\/p>\n UDHPs themselves can compete with EHRs by incorporating more user-friendly AI interfaces. UDHPs will offer low code\/no code tools to manage workflow and processes across care teams.\u00a0<\/p>\n Could a startup digital health<\/a> company create a unique AI interface layer on top of EHRs?<\/p>\n As noted in TechCrunch<\/a>, \u201cthe interface is where the profit is.\u201d We can anticipate the battle\u2026but anticipating the winner(s) is much harder.<\/p>\n Incumbent healthcare organizations can extend UDHP technological infrastructure to reconceptualize themselves as having platform business models<\/a>. SAP makes the case<\/a> that \u201cthe future hospital is a platform.\u201d\u00a0<\/p>\n Mayo Clinic has been a poster child for innovative thinking with its Mayo Clinic platform. Dr. Gianrico Farrugia, President and CEO at Mayo Clinic, described<\/a> Mayo\u2019s transition from a traditional pipeline business model to an innovative platform model:<\/p>\n Our current healthcare model functions as a pipeline, which is linear, highly transactional, and inflexible. The failings of this model were clear during the pandemic, where collaboration was limited and stress to individual nodes led to points of failure. This, in turn, hampered and in some cases paralyzed the entire system.\u00a0<\/p>\n A platform model of health care and ecosystem is fundamentally different. A platform is built around secure, collective resources such as a powerful analytic engine and an open ecosystem of trusted collaborators and is highly dynamic, self-learning, inherently accessible, and scalable. This approach brings together providers, medical device companies, health tech startups, patients, and payers, among others, to design integrated end-to-end solutions that are supported by longitudinal clinical data and AI algorithms. A platform is purposefully collaborative and fuels innovation and access instead of limiting it.<\/p>\n Across industries, the IT landscape is experiencing a significant shift towards cloud computing<\/a>. Organizations are increasingly adopting multi-cloud and hybrid cloud strategies to optimize operations, enhance flexibility, and reduce vendor lock-in. This trend is driven by the need for scalability, cost-efficiency, and access to advanced services like AI and analytics.\u00a0<\/p>\n UDHPs align strongly with the cloud computing trend by leveraging its scalability, flexibility, and accessibility. Cloud infrastructure enables UDHPs to integrate disparate systems, facilitate data sharing, and support innovative applications.<\/p>\n Currently there are 20+ UDHPs. This market has many long-term characteristics of a highly concentrated, winner(s)-take-all marketplace. By \u201chighly concentrated\u201d, we mean a monopoly (one dominant firm), a duopoly (two dominant firms), or an oligopoly (a small group of dominant firms \u2014 typically three to five):<\/p>\n UDHP buyers will prefer to \u201csingle home\u201d, i.e., select a single UDHP platform<\/p>\n Switching costs will be very high<\/p>\n Network effects will be strong<\/p>\n We foresee the UDHP market as becoming a battle for applications and developers. Developers prefer to work on as few platforms as possible, ideally one or two.<\/p>\n UDHPs thus will have strong indirect network effects. Buyers will prefer UDHPs that offer a wide range of applications created by developers; developers will converge on platforms that have the greatest number of buyers.<\/p>\n The smartphone operating system<\/a> (OS) market provides an analogy. In 2006, there were 12 companies offering smartphone OSs; by 2015, the battle was over \u2014 Android and Apple iOS owned more than 99% of the market. Today these winners each boast over 2 million apps on their OS platforms.<\/p>\n Implications: As the UDHP market becomes more concentrated,\u00a0 there will be a few winners and many losers. Buyers should evaluate vendors carefully.<\/p>\n Unified Digital Health Platforms are a new category of enterprise software, one that Gartner predicts \u201cwill, over time, replace the dominant era of the monolithic electronic health record (EHR).\u201d<\/p>\n Today, the UDHP market is still very early \u2013 about 5% penetrated.<\/p>\n UDHPs offer compelling value propositions \u2013\u00a0 improving clinical care, overcoming the limitations of EHRs, offering modular architectures, enabling a common infrastructure, and providing competitive advantage.\u00a0\u00a0<\/p>\n UDHPs also promise to shake up the competitive landscape of healthcare by creating a major platform shift. In the short term, they must collaborate with EHRs; in the long term, they can commoditize EHRs. They enable new platform business models and facilitate the development of new ecosystems and partnerships.\u00a0<\/p>\n UDHPs will play a major role in the broader \u201cplatforming\u201d of healthcare. They\u2019re a trend worth watching.<\/p>\n Vince Kuraitis<\/em><\/a> JD\/MBA is an independent consultant with over 35 years\u2019 experience across 150+ healthcare & tech companies. He publishes <\/em>The Healthcare Platform Blog<\/em><\/a>, where this post first appeared.<\/em><\/p>\nMayo Clinic Platform: Healthcare platforms and AI<\/strong><\/h2>\n
Business and Strategic Implications of UDHPs<\/strong><\/h1>\n
A Platform Shift \u2013 Disrupting EHRs?<\/strong><\/h2>\n
UDHPs Enable New Ecosystems & Strategic Partnerships<\/strong><\/h2>\n
AI Will Play a Major Role in UDHP Adoption<\/strong><\/h2>\n
UDHPs Enable Platform Business Models<\/strong><\/h2>\n
UDHPs Advance the Movement Toward Cloud (CIO POV)<\/strong><\/h2>\n
Long-Term Potential for Winner(s)-Take-All?<\/strong><\/h2>\n
Summing Up<\/strong><\/h1>\n