Envisionings of a crypto-enabled healthcare platform

Intricacies and architectures

Control, access, and benefit with our personal health data should be inherent in our healthcare system. Why don’t we have a comprehensive healthcare record on our phone that we can easily share with our doctors? Why are our insurance and healthcare companies able to sell our data to make millions, with us not knowing or getting any care or fiscal benefit? Why do we not know when our data is sold or when it’s compromised? The ideals behind these outstanding questions are not unique to healthcare, and in fact encompass a lot of the same sentiments motivating the crypto community: self-sovereignty, transparency and traceability.

Despite the philosophical kinship between healthcare and crypto communities, we acknowledge that an extreme amount of care and caution must go into integrating these two fields. With this blog post, we talk about where healthcare solutions should start (e.g. where patients need it the most) and related forcing functions of patient onboarding/incentivization in a series of steps. For those that aren’t crypto native, we give an overview of crypto vs. decentralized compute, monetization, touch upon layers, and utility of tokenization. We conclude this piece by talking about our essential elements and suggestions of an ideal decentralized healthcare platform.

Step 1: Identify high need/low support areas

High need/low support areas are the best pockets for layering in crypto in healthcare. Kudos to the MoleculeDAO team which has tapped into many of these such areas with sub-DAOs focused on longevity, hair loss and women's health for example. There’s many other opportunities however such as rare disease (Vibe Bio), chronic conditions (such as diabetes), organ transplantation, antibiotic resistant infections, mental health, autoimmune diseases, and others. The lack of support in the traditional health (TradHealth) system is a forcing function of folks onto these platforms and imagine this will continue to only increase in the upcoming years. The Cystic Fibrosis Foundation is an early pioneer of extreme patient advocacy which has/is currently funding 34 clinical trials. However, most groups don’t have such active patient advocacy investment so are compelled to make their own.  Crypto, admittedly, doesn’t solve these problems but gives the potential for funding, community & coordination which could be very valuable (in addition to serious privacy settings). With a grounding raison d'être (e.g. alleviating menstrual pain, stopping hair loss, doubling life span), companies will be able to offer biomarker tests, data storage, and/or participation in distributed clinical trials. We’re really big fans of what the HairDAO folks are doing and excited to see this model of data storage, treatment suggestion, and ultimate target discovery proliferate throughout different areas. 

HairDAO patient database for their decentralized clinical trials.

Step 2: Privacy as a forcing function; multimodality as a moat

However, we believe thematic areas are just a stepping stone towards platforms which encompass biomarker storage for EHR, biomarker and lifestyle data. Data ownership and privacy has important tailwinds with over 39M US citizens having their data compromised in 2023 so far and women of reproductive age are under greater surveillance where healthcare data can be used against them in court.  Moreover, the precipitously decreasing costs of biomarker data and the commensurate companies having DTC business models (Tally Health, Viome, 23andMe) gives us more access and agency of our health data than we’ve ever had before. However, TradHealth doesn’t have easy ways of integrating the new biomarker data into your EHR, or using it to improve your healthcare. Not only that, but it’s nearly impossible to access your EHR, which is based on location and network rather than your needs. Some companies try to wrap their wedges around new tech (e.g. Acorn Genetics), but think they’re missing the larger proposition of data aggregation of all the biomarkers and digital health data. Companies which layer in crypto around new biomarker analysis technology miss the ability to layer in new biomarkers that might be better in outcome/lifestyle predictions than DNA. On the flip side, data storage incumbents (such as Filecoin and Storj) aren’t systematically building out multimodal datasets, let alone ones around high need/low support communities so miss the opportunity for activating high needs patient communities and rich data around those indications. The DAO model in general is still in its infancy, with very primitive tooling and an unclear end-state for coordination & decision-making. Specific to healthcare-related DAOs the communities around patient diseases are more focused on specific indications and don't seem to be building out robust backend for the multimodal data layering. We think the communities will still accrue a lot of value for those communities but aren’t working toward the ultimate vision of being a breakaway for healthcare data aggregation, privacy and self-sovereignty. With this in mind, we want to highlight the backend definitions and mechanics which would be needed in a crypto//healthcare platform. 

Step 3: How and where to build with crypto (if you’re crypto-native, you can skip this section)

When we talk about cryptography, in its simplest form it’s just a method of securing information and communication in the presence of adversarial behavior. I have information that I want only certain specific other actors to access and process. That could be financial data, health data, education data, biometric data, behavior data, etc. We live in a world where data is extremely valuable and so it’s inevitable there will be adversarial actors trying to access the data I want to protect. Cryptography uses mathematical concepts to transform messages in ways that make it harder to decipher. 

While crypto is often (correctly) thought of as an improved layer of security, it can also be linked to networks for decentralized compute, storage, processing and bandwidth. For the non-crypto folks in the audience here’s a debrief on decentralized compute: 

• Individuals (or corporations) with spare compute resources become providers in the network. They list those resources on the network, making them available for rent.

• Customers seeking additional compute can rent these resources -- this could be storage space to host data, processing power to run an app, bandwidth to serve content. The idea being that as demand for computing resources continues to rise – especially as AI becomes increasingly ubiquitous – there are latent, unused resources distributed across the world that can be tapped into, on-demand.

• The network records all of this activity on-chain (available resources, resource rental transactions, rewards for providers, etc). Smart contracts run automatically to handle the resource allocation & payments between the two sides.

Tokenomics

• Node Operators, aka the nodes that make sure the blockchain is running properly. In their simplest form nodes run software that keeps a copy of the blockchain and broadcasts transactions across the network. They are rewarded in the native token for this activity. Ethereum node operators are rewarded in ETH for example. Having a network of distributed nodes provides all the benefits we talk about (privacy, decentralization, ownership, etc).

• Resource providers are those contributing compute resources and are rewarded with tokens (i.e. the more resources you contribute/provide, the more token rewards accrue to you). This obviously incentivizes participation and network growth. There are all kinds of added elements here you can use to encourage the types of resources you’d like to see on the network. You can weight rewards toward higher-performance resources for example.

• Customers are the ones in need of additional resources and typically pay for the compute/storage/etc with the native token. A common method for bootstrapping this type of network is to reward early customers in an outsized manner (i.e. the earliest customers and contributors receive added rewards for helping take the network from 0 to 1)

As the network grows and demand for resources increases, all these participants are able to transact with tokens on the secondary market to convert to fiat, stablecoins or other cryptocurrencies.

Where to build?

This is a common question for all crypto-enabled startups and the answer you hear varies widely depending on who you speak with. To-date, Ethereum is the dominant Layer 1 for smart contract applications and where the bulk of capital resides. Gas fees (aka transaction fees) are quite expensive on the base layer, a problem Ethereum is well-aware of and has been tackling through its roadmap. That’s led to a proliferation of Layer 2’s like Arbitrum, Optimism, zkSync and Starkware for example. This is a far more feasible design choice for most applications as L2’s are purposefully designed to improve scaling and handle larger transaction loads efficiently. Here’s a very crude, albeit directionally correct representation of just how many orders of magnitude cheaper it is to interact on L2’s compared to Ethereum.

Beyond Ethereum or some of the Layer 2’s built atop Ethereum, there are other options as well. Solana for example is another Layer 1 that optimizes for different properties than Ethereum does: philosophically Solana is built with the view that latency should be low, throughput must be high, fees should always be cheap and that everything should be on one monolithic layer. There are trade-offs here of course, most notably that hardware requirements are higher and as a result the network will inevitably be more centralized. But as with most things, the narrative often misses nuance and decentralization is not binary, it exists on a spectrum. We are more than happy to go deeper on our internal views across the different architecture options for anyone building and interested.

The other option for building networks is to launch your own Proof of Stake (PoS) network. When we described earlier what a token-model network would look like, that is how a separate PoS blockchain network would operate. The difficulty here is first finding validators to run nodes for your network. The validator bootstrapping dilemma is real and you need a sufficient number of nodes to run the software and broadcast transactions. Intuitively this is more difficult than tapping into the already existing network of node operators on an L1 or L2. The other added benefit of building on an existing L2 is liquidity sourcing; many of these L2’s have spent years building and attracting users and liquidity to their network. Tapping into this existing liquidity is a major advantage, especially at the earliest stages of projects. It’s also worth noting that while this initial decision of where to build is important, it’s not necessarily the end-all-be-all. Earlier this year we saw dydx (a perpetual exchange) move from Ethereum to its own app-chain on Cosmos.

Step 4: What we’re looking for

We think a company should be built around the high incentive patient groups in the first instance, layering in the ability to collect and encrypt all data modalities (biomarkers, EHR, wearables, DNA) of the patient and providing them a central truth for their health data. From here network effects can be created by layering in community, funding mechanisms, and means for crowdsourcing knowledge through activities such as decentralized clinical trials. Hopefully the results of knowledge crowdsourcing would lead to improved patient outcomes and new innovations for the entire community.  The important thing for companies here will be around sequencing of other communities onto their platform with the ultimate vision of being the healthcare storage and center for actionable insights. This has been a difficulty in companies such as  AI Gia Health and JennyCo which provide a centralizing platform but without an onboarding imperative, we struggle seeing these platforms achieving breakaway success. 

On the crypto side, we think building on L2’s would provide the lowest fees and highest liquidity upfront. Which L2 and the choice between L2 and PoS is relatively inconsequential as the network can always be changed. This leads us to our final point which is that healthcare needs to lead in any platform being built here, not crypto. This is because we’re all still living in an entrenched healthcare system with lots of inertia and new storage mechanisms won’t move the needle. A new envisioning of patient care and agency will. Aesthetically and to broaden support, we think backend discussions (e.g. talking about crypto) during raises and public-facing communications to patients should be kept at a minimum. This isn’t because a company here is hiding what they’re doing but a focus on security for the betterment of healthcare is a much more inclusive premise. We think community shouldn’t be based around Discord as it’s still out of the mainstream and think meeting patients where they are (Facebook, Slack, Instagram, Tiktok, even SMS) is the most important part. We want everyone to be part of a future where we can improve patient outcomes faster and more proactively.  This is part of a broader trend in crypto (through account abstraction, SSO integrations & UI/UX improvements) to make the space more accessible for all. 

As a whole, we don’t think the pushes in new and more accessible biomarkers will achieve maximal utility until they are combined with other health information. However, the obvious tradeoff of aggregating all of this health data over the coming years is data security and privacy. Our view is that a privacy-preserving system is both necessary & leads to better outcomes for the future of healthcare. Blockchains can undeniably solve some of these problems, though their limited adoption today means real nuance is needed in the build to allow for the highest distribution possible. 

We’d love to see more innovation happening here and would really enjoy speaking to people building in this area.