Robert Miller

Newsletter

About

Blockchain business networks in healthcare

- 5 minute read

I wrote the following in my 2019 review of blockchain and healthcare:

Blockchain technology is already changing healthcare by enabling competitors to collaborate to solve common problems. This alone is remarkable.

In the rest of healthcare you will seldom see fierce competitors working together, but it is commonplace with blockchain technology. Over the past three years we’ve seen the gradual rise of groups of businesses in healthcare, what I am calling business networks, coming together and using blockchain technology to solve their common problems. For example, the Synaptic Health Alliance is a group of 11 payers, providers, and technology partners collaborating to tackle the challenge of providing accurate and efficient provider data.

Don’t mistake this for blind optimism, the members of the Synaptic Health Alliance, and those of any business network, still compete vigorously with each other. But the shared infrastructure of business networks, blockchain technology, give competitors a a way to solve common problems without overly advantaging any one party. It is because of blockchain technology’s very unique property of neutrality that these businesses can come together in this way.

So what are these networks? How many are there, what problems are they solving, who is participating, and what’s working? Building on the work of John Bass and Hashed Health last year, I documented and described the major blockchain business networks in healthcare. In this post I discuss these and detail a few trends we are seeing.

Above is a table I compiled of business networks and a few details about them. There are 12 networks in this list. My criteria for this list was:

  • A focus on healthcare or the life sciences

  • Uses blockchain technology at its core

  • Composed of several members

  • Actively building in the space

  • Publicly announced

I declined to include a few networks because they weren’t focused on healthcare (e.g. TrustYourSupplier) or uncertainty over whether these are active networks (e.g. many DSCSA pilots). You can view this table here.

Instead of rehashing out the details of each network, which you can find in my commentaries linked in the table above, I want to highlight some trends across networks:

  • Almost every stakeholder group is represented in a business network

  • There are now competing business networks

  • Most activity remains anchored in the US

  • Most networks have one party that plays an outsized role

  • Use-case first business networks are progressing quicker than platform first

  • Blockchain-enabled federated learning networks are rising

Here are those trends expanded upon:

Almost every stakeholder group is represented in a business network

Every one of the top 10 pharma companies and almost all of the top 10 big payers are participating in at least one network. Some companies are a part of multiple networks now, particularly in pharma. There are a handful of health systems participating too: Synaptic, ProCredEx, and the Health Utility Network each have health systems as members, as do all of the the federated learning networks. However, adoption by health systems has generally lagged that of pharma and payers.

Some stakeholders, like PBMs and pharmacies, have a company represented in a network or two, but are lagging even further behind in adoption. Notably none of these networks directly engage patients today, which makes sense given the complexity of blockchain solutions and the difficulty of creating patient facing products in healthcare generally.

There are now competing business networks

Multiple networks are tackling the the same issues in pharmaceutical supply chains, which makes these business networks competitive with each other. Furthermore, the multitude of FDA pilot projects move from pilot to production this competition will only be heightened.

Although outside of pharma supply chains there are no explicitly competitive networks today, this is sure to change in the future. This is ironic given that business networks are ways for their members to collaborate, but having multiple networks serve the same purpose limits collaboration by creating fragmentation. In turn that diminishes the value of any given network.

Competition between business networks will be similar to that of platforms. The services that business network provide and their corresponding network effects will be the dominant dimensions of competition. In some cases business networks may decide, or be obliged, to interoperate. But this is a complex topic, as by interoperating with another network you lose some of the assurances of trust associated with blockchains, and different technical approaches may not be interoperable. Furthermore, larger networks may not have an incentive to interoperate. Some of these dynamics are discussed in my previous newsletter: Interoperability and blockchain networks in pharmaceutical supply chains.

Activity remains anchored in the US

In some ways it is unsurprising that most activity is anchored in the US. The US has a notoriously fragmented healthcare system, and it makes sense that more business networks would arise to help coordinate disparate entities with varying incentives. While a more centralized healthcare system could adopt blockchain technology more quickly, by definition there isn’t as great of a need for a blockchain to coordinate parties in more centralized systems.

Moreover, there are no networks documented for several other continents here. I suspect that there are business networks that I don’t know about, but blockchain scenes across the world are notoriously opaque and it is difficult to get accurate information. Trust me, I try. In particular, I get the sense that there is a great deal of blockchain and healthcare innovation in China and South Korea that is mostly undocumented in English.

Most networks have one party that plays an outsized role

All but two of these networks, the Synaptic Health Alliance and PharmaLedger, have a single party that has convened the network and generally plays an outsized role in moving the network forward compared to other members. For example, MediLedger has Chronicled and MELLODDY has Owkin.

The reality is that it is easier to advance a network this way, and generally this is acceptable to businesses. Someone needs to write the code, maintain it, be there to answer phone calls when things break, set roadmaps, onboard and offboard members, enforce compliance, etc. It is easier for one party to do this instead of a committee.

You may object and call this centralization that defeats the point. And to some extent you would be right, this is centralization of some functions of a business network. However, note that business processes can still happen in a decentralized way. Properly designed a party that had an outsized role in, say, product development, would only be 1 node in a many node network. Businesses can submit and verify transactions by themselves, interact independently, own their data, and so forth. Some parts of a business network can be decentralized while others are centralized.

But still, I think that there is an important bit of truth to the criticism that having a party with an outsized role in a business network defeats the purpose of decentralization. Business networks will have very high switching costs, and so the risk of being locked into a vendor is high. Businesses are right to worry about that, especially as these networks enter progressively higher value areas. Further, business networks with an outsized party will always be fighting Conway’s law, which states the organization of a software system will reflect the organization of the team that built it.

Less often observed is the incentives that having a dominant party creates. While a more decentralized business network will be keen to interoperate, networks with an outsized party may not have the incentives to do so, and instead seek to dominant a market. In turn this could lead to fierce competition between networks and fragmented systems.

Use case first vs platform first business networks

Use-case first business networks, which first convene around a particular use case, are both more common and making more progress than their platform first peers, which convene with the intention to create a platform that supports multiple use cases. In part this is selection bias: many of the platform first business network experiments of 2017 didn’t survive and thus aren’t on this list.

When making a platform you have to start somewhere, and in platform first business networks the key challenge is deciding where to start. This is a more difficult challenge than you would first think. A blockchain use case requires an appropriate business, governance, and technical model. These dimensions are each difficult to get right in isolation, and orders of magnitude more difficult when coordinating multiple parties with different visions.

Ironically it seems that MediLedger, a use-case first business network, will be the first network to have multiple production use cases. All business networks want to be platforms, but it might be those that start with a use case first approach that achieve this first.

The rise of federated learning networks

Lastly, there has been a rise in the number of federated learning networks. These use a blockchain to manage access to data, provide an audit trail, and decentralize control over what algorithms are trained. Not only are these literally networks of healthcare businesses using blockchains, I see them as closely aligned in purpose to non-federated learning networks: they also provide a neutral way for groups of likeminded organizations to collaborate and solve common problems.

Concluding thoughts

Today there are a dozen or so early business networks spanning across almost every part of healthcare. But these are still early days. No network has a system in production, though a few are close. Expect to see business networks expanding their members, adding new use cases, and moving into progressively more complex but higher value areas. As this happens business networks will increasingly compete against each other.

Building business networks is hard work, but it is necessary. As long as there are shared problems there will be a need for shared infrastructure. Our solutions to some of healthcare’s largest problems, like value-based care and truly patient centric systems, demand that companies work together. Forward thinking enterprises can use shared infrastructures and business networks to do so.





Sign up for my newsletter

Every Sunday I send out a free blockchain and healthcare newsletter with top stories and ideas that you need to know about. Join leaders from the Mayo Clinic, Kaiser Permante, Optum, Amerisource Bergen, ConsenSys, Hashed Health, Pfizer, Goldman Sachs, and more by subscribing today.