HIE of One: Gropper's Healthcare Blockchain Vision

Let's use blockchain technology to shift the balance [of control] back to the physician-patient relationship. This is what Adrian Gropper proposes in ONC Blockchain Challenge paper #7. He compels us to look using the title "Powering the Physician-Patient Relationship with HIE of One Blockchain Health IT*".

Dr. Gropper asserts that physicians and patients are primarily responsible for care decisions and medical expenditures but that EHR and related technology is mostly controlled by large institutions (who can afford such assets). In turn, he reminds us, these institutions need to continue to maximize their own growth and control over the entire process. Dr. Gropper is concerned that these large organizations inhibit cost reductions and innovation due to inertial restraint. Blockchain, for Dr. Gropper is essentially a call to independence and a chance for physicians and patients to take more control over innovation from the large institutions.

I understand his concern about the inertia of large institutions. But, many of his assumptions do not resonate with me. For example, Dr. Gropper states that "hospitals are reluctant to show physicians the cost of a medication or procedure when they're about to order it. They are even less likely to present the physician with a list of alternatives".  Dr. Gropper, I would like you to meet my friend, the State of Maryland, where hospital cost is highly regulated and we would love to intervene in your cost choices at the hospital.  However, it is true that cost transparency has been a difficult process challenge. Dr. Gropper proposes that blockchain technology can provide for this cost transparency at the mobile device level during decision conversations with the patient. I'm in favor.

Gropper's first work in this blockchain realm is in the creation of identity standards that will address patient matching challenges and allow for identity federation. He would like to take the blockchain concept of the identity "wallet" and incorporate it into existing healthcare technology standards such as OpenID Connect and FHIR. This would support methods for patient and provider access as well as a way to link wallet identities to records. He has demonstrated this system in the context of standards created and discussed by or among participants in HL7 FHIR , OpenID Connect , Kantara UMA Standards and OpenID HEART workgroup.  These are all important standards bodies as they help define how individuals (patients, providers, and care teams) can access and exchange components of information in a secure, encrypted, identity authenticated environment over the world wide web.

Unlike other papers reviewed so far, Dr. Gropper makes it clear that PHI will not be stored on the blockchain. The HIE of One will only secure and manage access to PHI.  The data stays with the original creators and consumers and each one of these endpoints is accessed by others via the HIE of One when queried. In this way, he initially proposes that nothing will be added to the blockchain beyond the initial creation of identities and relationships. This makes this implementation cheaper than other blockchain proposals since adding data to a blockchain ledger incurs additional "proof-of-work" algorithm and infrastructure costs.

The author loses me when he discusses how the blockchain identity trust can be used to exclude hospital institutions from transactions. He gives an example of a prescription transaction. I agree that the ability to provide a trusted ledger that pharmacies can rely upon to verify that a licensed provider is prescribing to a legitimate patient is preferred.  However, I don't agree with his assertion that a "hospital trust" and hospital information system is required for these transactions to occur at a provider practice today.  Again, in Maryland, our goal is to keep the patient away from unnecessary hospital encounters in deference to community provider interventions. Most providers process these transactions without any involvement from hospitals.

Aside from that distraction, Dr. Gropper proposes that the physician to patient transactions can be mediated by a blockchain based decentralized identity or DID.  The benefit of the DID is that it does not rely upon a Health Information Exchange (HIE) or Enterprise Master Patient Index (EMPI) of centrally stored patient information in order to verify patients when they are involved in a transaction. This is setup, in my version of the truth, like this:

Step 1: Patient downloads trusted DID application to personal device.
Step 2: Patient creates identity wallet secured through various password or question and answer methods.
Step 3: Identity wallet submits encrypted identity onto the HIE of One blockchain.
Step 4: When patient presents at practice, the practice scans device output (let's call it a QR code).
Step 5: Transactions and EHR records are associated with this encrypted identity as verified via the world wide web and the HIE of One.

Gropper says' that the transactions will be recorded in both the provider and the patient's self-sovereign support technology (SSST). This confuses me a bit because he earlier maintained that the PHI would not be stored centrally and that there would not be expensive additions made to the blockchain.  Where is this SSST housed? Furthermore, an institutional or vendor-provided SSST is required, thereby re-introducing the potential need for institutional funding. It may be that I don't understand what he is proposing but what I am hearing is:

1.) Let's declare our independence from hospitals, EPIC, Cerner, Meditech, Allscripts and all the other large vendors and systems of the world who provide EHRs and patient portals, and
2.) Let's declare our allegiance to some new brood of vendors who will provide SSST systems for providers and patients.

I'm all in favor of disrupting inertia and decreasing price tags but I'm not sure Dr. Gropper's proposal is inherently a provider's or a patient's ticket to freedom. This new stuff will have a price tag that's paid for with money by someone who has it. Furthermore, if providers are to divorce from all these large entities then who will be providing the expensive services required for evidence-based and precision medicine, disease surveillance, case management, and the myriad of infrastructure required to manage the complexity of our populations?

One of his examples of a federated technology which is not inherently an SSST is called NOSH. It's an opensource charting application that a provider can install and implement without vendor intervention. I think it's great. I love opensource tools. Maybe one day I will have a NOSH jam session with the Linux system I only sort of understand. But, for most independent provider practices, chances are there won't be an abundance of Linux and MySQL techno-geeks and other expertise to install and maintain this system. This returns us to the money premise. Someone needs to spend it.

Make no mistake, Dr. Gropper and his cohort are making an important contribution to the brainstorming for the next generation of healthcare information solutions. I think they should continue this effort albeit without the distraction of using divorce from hospitals and other important team players as a premise for the work. We can do this better together.

And, speaking of federated opensource, Dr. Gropper has posted a user installable version of HIE of One.  All you need to do is run some sudo install commands on your Linux system equipped with PHP, MySQL, Apache, and CURL. Are you still with me? Stay awake!

_______________________________________
* Powering the Physician­Patient Relationship with HIE of One Blockchain Health IT Adrian Gropper, MD August 7, 2016