Curbside Consult with Dr. Jayne 10/11/21

Early in my informatics career, I worked on a health information exchange project. It was during the early days of HIEs, and many of the challenges were legal and operational as opposed to technical. We had to wade through the minefield of consent, debating opt-in versus opt-out models within the confines of the laws of multiple states. We also had to address access issues, decide when break-the-glass functionality could be used, and create policies and procedures around auditing access to the data and ensuring appropriate use. Only once those thorny issues were settled could we begin to define the clinical data sharing model and determine what information would be shared from what sources.

We then had to work through the technical issues. We had to decide whether we wanted ambulatory office visits to automatically query the HIE versus whether providers would have a manual trigger to prompt data sharing. We had to address hosting issues as well, along with the pure limitations of the product we had, since we had purchased our solution from a company whose strategy was still evolving. There were dozens of interfaces to evaluate and integrate, and we had to create a solution that would provide immediate value while not breaking the bank, buying ourselves time to bring up the rest of the data feeds. The big draw for our solution was its ability to allow providers to incorporate discrete data from the HIE into their charts so that they could use it instantly within the context of the patient encounter.

We didn’t necessarily see them coming, but many issues we faced though turned out to be political in nature. Unknown to us, the CIO of the health system with which our physician group was affiliated had his own HIE plans, and they didn’t involve us. He had secured funding for his own HIE and had crafted a strategy without any input from the thousands of ambulatory physicians who were clamoring to be connected. His solution was more of a viewable repository that was document based rather than enabling the exchange of discrete data. The last thing our physicians wanted was to have to sift through textual information and then perform data entry tasks in order to incorporate that information in their own records, so you can guess whose solution was more popular.

Needless to say, he spent a lot of his time trying to kill off our project. Not only would our HIE concept provide more value, but we were planning to deploy it for a fraction of the cost of what he had planned. He also wasn’t terribly fond of having to work with physician informaticists, let alone one who was relatively young and decidedly sassy.

Since we were technically independent despite the affiliation, we pressed ahead and implemented quickly, helping physicians from day one. Our most valued feature was assisting in reconciling medication lists from different sources and identifying patients who might be seeking controlled substances from multiple physicians. Other solid features involved supplying data for problem and diagnosis lists as well as laboratory and biometric data.

Although I moved on before our little HIE reached maturity, I still regard it as one of the best projects I ever worked on, and also the most educational for me as a clinical informaticist. I learned more about discrete data, interfaces, and interoperability in those months than I probably did in the first five years of my career. In the early days of data normalization, I also learned that laboratory directors don’t like it when outsiders find problems with their data, and if you’re going to question senior physicians who are twice your age, you had better come ready with plenty of facts and examples because it’s going to be difficult to convince them that their system isn’t perfect.

Since then, I’ve kept my eye out for interesting HIE stories and have enjoyed seeing how exchanges have evolved over time. Although many of the technology issues have stabilized, there are a host of challenges that are both operational and financial. A Brookings Institution blog post caught my attention last week. It reviewed some of the digital transformation that has occurred as a result of the COVID-19 pandemic, including increased adoption of telehealth and the rise of healthcare technology startups and retail healthcare.

The authors note that the transformation is also impacting the health information exchange world, raising questions about how HIEs fit into the larger healthcare ecosystem. Where traditional HIEs typically involve data exchange among physicians and hospitals, there is a growing need to incorporate data from a multitude of other sources. Since many of the newer players, including retail clinics, involve large national organizations, there is motivation for them to maintain their own medical records without necessarily having to integrate with traditional provider or hospital organizations.

Additionally, given functionality required by federal incentive programs, patients now have a greater ability to view, download, and transmit their own health information. The authors note that new features such as Apple’s iOS Health Records functionality allow patients to communicate more directly with their physicians. However Epic, was not included in the Apple implementation. They summarize, “Not only can these companies choose which HIEs to work with, but they disrupt the original purpose of HIEs, which was to centralize medical care for improved efficacy of patient care.” This means that HIEs may need to play a new role in the marketplace, and the authors list strategies for HIEs to try to remain relevant:

• Diversify network members and data types to stay relevant.
• Include knowledge discovery in their focus.
• Work horizontally and vertically to meet patients and providers where they are.

These are certainly important points. HIEs are going to need to widen their user base and make sure they stay current in understanding the needs of their constituents. HIE use cases have gone far beyond catching patients who are seeking duplicate prescriptions (most states have prescription drug monitoring programs for that now) to providing opportunities for analysis of broad aggregations of patient data that could provide valuable information for public health as opposed to being merely push/pull platforms. In the third point, the authors propose that HIEs consider mergers and acquisitions to expand in similar service lines, such as collaborating with HIEs in neighboring states, where vertical integration would allow them to better integrate with their current data suppliers and consumers or add stakeholders such retail healthcare providers.

The authors also note that further HIE growth may be limited by other factors. These include concerns over patient privacy and worries about increased regulation.

For the latter, there are concerns that charging a fee for data exchange might be construed as information blocking, so there are plenty of issues to resolve there. I’d also mention that they need to worry about cyberattacks and maintaining adequate financial resources to ensure solvency.

In our current environment, they also need to be wary of becoming embroiled in political controversies. For clinicians, sharing vaccine information through HIEs has been a tremendous benefit and allows us to have complete records on pediatric patients and avoid giving duplicate immunizations. In our polarized political climate, I wouldn’t be surprised to see certain states try to make it illegal for healthcare organizations to share COVID-19 vaccination data.

HIEs have always had tremendous potential, but the road to success has been a rocky one and there have been quite a few failures along the way. I’m hopeful that the current generation of HIE leaders understands the challenges and that those leaders are getting creative about ways to ensure longevity and a bright future.

What role do you see for HIEs in the coming years? Leave a comment or email me.

Email Dr. Jayne.