Part 2 of 2
By Michael Solomon, PhD, MBA, Practice Lead, eCare Management
As I conveyed in the last article of this series, artificial intelligence took center stage in the early weeks of our COVID-19 battle but is limited in its effectiveness because of the lack of data. In this article, I’ll examine what we need to reduce the “noise” and improve the liquidity of data required for online, AI-based applications operating in a complex, rapidly changing environment like COVID-19: the rapid scaling of a national health information exchange.
Health information exchange today is mostly regional, with just a few statewide networks, connecting a minority of health care providers let alone public health agencies or payers in the vast majority of the country. The ONC recently reported slightly more than half of U.S. hospitals used data received electronically from outside the organization in 2018. The number is far less if you move outside health systems to independent medical practices, post-acute care, community health centers, behavioral health, ancillary providers, payers, and consumers. No more than one in four hospitals participate in a multi-EHR network with a national footprint (e.g., e-Health Exchange). Considering the potential of the AI applications I’ve described, not to mention surveillance systems, and the outcomes analysis needed to inform clinical research in the battle with this pandemic – the lack of a national health information exchange warrants being treated as a crisis of its own.
TEFCA and innovation to achieve a national health information exchange
Fortunately, the architects of the 21st Century Cures Act recognized the need for nationwide HIE. The HHS Office of the National Coordinator for Health Information Technology (ONC) is leading government efforts to achieve nationwide connectivity and a scalable network that enables patients’ electronic health information to flow when and where it’s needed. The three pillars of the Trusted Exchange Framework and Common Agreement (TEFCA) -- ONC's model for a nationwide HIN – reflect a “network of networks” approach to leverage existing HINs across the country. A select number of “qualified” health information networks (QHINs) will connect their participants (e.g., regional HIEs, public health agencies) and will also connect to other QHINs, enabling exchange of information among all entities and ultimately nationwide connectivity. QHINs will be governed by a set of universal data exchange principles (the Trusted Exchange Framework), and terms and conditions for data sharing (Common Agreement), which are anticipated to foster participation.
TEFCA is still in the development stage, with specifications in draft form and an active Common Agreement work group. The model for TEFCA was developed before the COVID-19 outbreak. While the fundamental structure remains sound, response to the pandemic (and the inevitable public health crises in the future) demands a bolder, more innovative rollout.
First, the lengthy process of soliciting applications from HINs meeting specified criteria and the voluntary nature of participation risks perpetuates the fragmented, uneven coverage and quality of health information exchange that exists today. Onboarding to a health information network should not be optional for payers and providers. “Walled gardens” where community-level HIEs are comprised of health organizations that do not compete with one another must become a thing of the past.
A more assertive strategy is needed that focuses on scaling up and maximizing participation nationwide as soon as possible. The current HIE landscape features HIN entities – commercial as well as public – that clearly demonstrate the ability to deploy the infrastructure and resources needed to quickly stand up a national footprint, some with highly specialized and innovative data exchange services that could be leveraged across the entire network. ONC can accelerate the selection and implementation cycle by going on the offensive and pursuing these HINs, selecting from the best, providing them incentives to deploy QHINs and achieving pre-determined participation goals.
Voluntary participation of HIEs, payers and all provider sectors should be replaced with a model of incentives and penalties similar to what was used to stimulate EHR adoption a decade ago. Surely a Congress that has recently appropriated $2 trillion for COVID-19 relief can provide ONC with the funding needed to ensure rapid and comprehensive build-out of the national network of networks called for in TEFCA.
Second, the use of legacy standards and implementation variations is also a limitation to health information exchange that should and can be in our rear-view mirror, thanks to the rapid acceleration of the Fast Healthcare Interoperability Resources (FHIR®) standard. The eCQI resource center, sponsored by CMS, sums up why architects of a national HIN should embrace FHIR rather than sinking more money into legacy standards: “FHIR combines the best features of previous standards…while being flexible enough to meet needs of a wide variety of use cases within the healthcare ecosystem. FHIR focuses on implementation and uses the latest web technologies to aid rapid adoption.”
Almost every executive of a leading HIN that I’ve spoken with over the past few months supports the FHIR standard and is currently investing in apps to support use cases of interest to their participants, and for good reason. According to my colleague Jocelyn Keegan, who serves as program manager for the Da Vinci Project, “the shift to API from transactions moves us toward a truer services-oriented architecture supporting real-time interactions that are specific to a purpose and not ‘all the data at once.’ This transforms the way data flows, away from silos of information and toward more closely coupled and real-time APIs for interactions between stakeholders.” Exactly what’s needed to reduce the “noise” and improve the liquidity of data required for online, AI-based applications operating in a complex, rapidly changing environment like the current pandemic.
The potential for AI in public health and recent HHS efforts offer us the chance to look ahead and plan for future epidemics. Let’s be sure to seize the opportunity and build a national health information exchange that will boost the value of AI, and not limit this powerful and mission-critical technology. If you have questions about how and/or when you or your organization can participate and make a difference, contact me at michael.solomon@pocp.com.
Related Posts:
Three Cutting Edge Areas to Consider as AI Turns to More Practical Matters
Beyond the AI Hype: Practical Applications to Augment Medication Management Decisions (Part 2)
