Health Information Organizations: Who is the customer?

Health Information Organizations (HIOs) are a bit of an enigma for the uninitiated. To explain their purpose and function to someone outside the industry is a bit like describing the Hogwarts Castle in Harry Potter. There are hidden features, everchanging pathways to get to the same location, and an ominous feeling that something unexplainable will happen on any given day. Or maybe, it is hard to explain why collecting a lot of healthcare data, connecting several different service organizations, and providing a web-based viewer is important. Either way I ask...

Who is the customer?

This is a bit of an odd question to ask. But indeed it is hard to really know. Is an inactive onboarded organization a customer? What about those that pay versus those that do not pay for data services? This quandary is byproduct of a braided funding model that’s never needed to ensure all its participants are receiving value from the HIO. Now—let's be clear—HIOs do provide value, absolutely, and I have advocated for them in past presentations. However, my criticism is more about whether all participants receive value from a HIO, and to that I say—doubtful.

When you have a paying major health plan or health system in the HIO, they tend to have more influence than a CBO or small practice. Now, I am not naïve, and as someone that has managed a P&L before, I realize that bigger clients--particularly paying clients in this scenario—tend to have more influence over use cases and features than others. But this is not what bothers me, it is the fragility in this model that draws my ire.

Any business model built from the ground up tends to start with a core demographic and scales accordingly. Acquisition methods in the HIO community are a bit more about pursuing and onboarding any healthcare organization for greater network legitimacy. In California, it is broader than that—onboarding any healthcare or social organization. Now this is great for collecting large data sets and populations, but what percentage of the users find value in the HIO with this approach?

Surely large data organizations that have been stable in the market for over a decade are insulated from competition—right?

For some this may be a surprise, but at one point, a company called Yahoo was bigger than Google. And even more astounding, Google founders were trying to sell their company to Yahoo, but Yahoo declined. What’s the meme for I made a multi-trillion dollar mistake today?

So why did they fail?

Yahoo’s entire business model was based on flashy banner ads from other companies. Their homepage was cluttered with categories and new feeds. Nevermind the Internet user--or even the user traffic--the banner advertising was the sole source of revenue, so really paying companies were the customers.

So, how did Google beat them when Yahoo reigned supreme with a massive user base and revenue coming in quarterly? They did one thing great—search--and they focused on the casual Internet surfer. Google realized that attempting to categorize the Internet in a hierarchy was lunacy, but the activity of these Internet surfers could be valuable. Conversely, Yahoo (Which literally means: Yet Another Hierarchically Organized Oracle) focused on the advertising revenue. However, as history would show, you can easily find things using Google compared to Yahoo, and you don’t need a bunch of categories to help you.

Now, why discuss this business case when talking about HIOs? 

I believe that the near future holds a unique confluence of factors:

• More developer-friendly interoperability standards (FHIR).
• More robust data sharing practices and penalties for information blocking.
• More expansive financial coverage on social services and resources (state dependent).
• Larger data sets with more structured data.
• Small pool of infrastructure vendors (AWS, Azure, Salesforce).

Now the aforementioned factors seem to convey a sense of futuristic optimism, but it could also be an inflection point for many HIOs. Could we start to see more for-profit companies entering this space with more value offered to all customers? Additionally, are HIOs focused on the wrong types of data in their networks? What is more valuable than large, aggregated data sets?  How about identifying high demand data values within these data sets based on user activity, role, and geographical area.

Imagine the significance of this data. A HIO could know what specific users query or view by geographical area and frequency. This is valuable information to both government and commercial planners and developers. But most importantly, it serves as a benchmark for the HIO to promote other types of information around these users. This associative effect is what made Google so valuable. A user goes into a client’s record to view their housing status and the system recommends an assessment that other similar users access with that same housing data. This process is no different than when you are on YouTube or another social network tool, you started out watching a cooking video and ended up on a funny dog clip.

In a world where healthcare and community providers are at a point of technology fatigue, high turnover, and service capacity, the ability for HIOs to capture and leverage engagement activity will improve both HIO value to all users while solidifying a geographical competitive edge for themselves. Moreover, I realize that building these types of algorithmic models are extremely time consuming and complex, but simple data aggregation, mapping, and normalization is not a sufficient enough advantage over competitors these days.

Suggestions:

• Examine what data is already being leveraged by participants onboarded to the HIO and focus more support on these users. (These users are your community champions that already see the HIO value).
• If not present, build a backend feature to track engagement activity by user, data they viewed, and inactive time (another window over the viewer).
• Identify tangential populations to your community champions that require similar data or functionality to expand network engagement rates.
• Have a more efficient way for users to request data or a function, such as a call out box, in order to track high demand user needs.
• Build out semantic referential tables that map to common medically coded terminology to improve queries. (If I look up a disease name, I get LOINC codes associated with it.)

Naysayers may shun me and profess the Health Data Utility (HDU) concept is addressing this. A HDU is supposed to conduct stakeholder engagement across healthcare and social services to address broader needs in a given community. Said another way, a next generation HIO, one that serves a greater purpose to the community than just a data repository.

But I am not convinced that more workgroups will achieve different results. I am skeptical that the approach will generate more high-level use cases that don’t directly address the system user needs, just the organization's leadership. Furthermore, a reoccurring workgroup session is becoming an antiquated approach in the modern technology world. How can scheduling workgroup sessions be responsive to real world events or even efficient?  If we observe technology vendors and their approach to network optimization, they review engagement activity and reach out to power users or creators to better tailor their product.

And yes, I know, I am comparing healthcare technology organizations to for-profit non-medical technology companies without consent limitations and security requirements. But fundamentally the way people find value is the same, they stay logged in to products that meet their needs and delete the ones that don’t.

Now it is a bit conflicting when mission-driven nonprofit organizations get displaced by for-profit ones, it trades a vested company culture for more efficiency. But it doesn’t have to be this way, current HIOs have the first mover advantage. They have scale, networks, local program knowledge, and compliance. But without high engagement and perceived value from all users (customers) on the network, some may suffer the fate as Yahoo.