Searching for Awkwardness

Is there an intersection between healthcare, blockchain & crypto? Or is it just a Yeti - only imagined and not real.

Today marks a new phase in my life. As of July 2021, I’ve stepped down from my position as Chief of Digital Therapeutics at Holmusk to focus my time on other opportunities in tech which I’m deeply excited about. 

Holmusk remains very dear to my heart - as employee #2, I've been a core part of the company from its inception over 6 years ago. Today we are a team of over 100+ people working together to solve highly complex challenges in mental health using analytics & technology. The past years have been a wonderful, yet often bumpy journey of designing, developing and bringing our digital solutions to market and real-world impact. I am constantly amazed by my colleagues every time we accomplished something special and I have no doubt in my mind that Holmusk will play an integral part in shaping the mental healthcare landscape. Super proud to be a Holmusketeer and I will continue to support the team in an advisory capacity for the next few months.

I’ve spent quite a bit of my life straddling the gap between healthcare and technology. These experiences have fundamentally shaped my thinking about health. And I’ve gone full-time into crypto because I believe it is one of the most important things happening in the world right now. Looking ahead, I will be devoting the next years to flesh out and go deep into 2 core hypotheses:

Hypothesis 1: Software Eats Healthcare (... eventually).

It was 10 years ago when Marc Andreessen famously coined the phrase "Software is eating the world". Yet today, finance & healthcare are 2 industries which have yet to see software eating them up or reap the full benefits of technology, largely because they are heavily regulated and permission-ed.

Healthcare is a particularly tough cookie to digest and all healthcare tech founders know it's a slog to get a new product to actual clinical use by doctors / patients / hospitals, going beyond a pilot or clinical trial. Yet if you step back and look at the long term picture - we’re 100% heading towards a digital, data-driven future for health, which will lead to better quality of care, better health outcomes, better patient experiences and lower healthcare costs.

It’s what I’ve immersed myself in all these years and will not stop. So many problems (and to me, opportunities) to solve here - from prevention to disease management, oncology to radiology, patient-tools to clinical decision support features.

Hypothesis 2: Decentralised networks, Web 3.0 & Crypto are heralding the most revolutionary technological change for humanity since the internet.

I fell down the crypto rabbit hole 5 years ago, first learning about Bitcoin and Ethereum. At that time, things were highly speculative. Networks like Ethereum were still akin to science-lab experiments: limited throughput, difficult to scale, divisiveness in the communities over the direction of the protocol. Many people thought it was doomed to fail.

Today, millions of dollars of value is transacted every single day on the Ethereum network. This year, we will have scalability solutions that enable actual real-world products to be built on the technology (eg. Arbitrum, a layer 2 protocol in developer release which I'm psyched about). As an observer, the speed of development is staggering and only accelerating. It's drawing in the smartest, brightest young talents globally, nothing like i've ever seen before.

Above: Number of Ethereum transactions/day - Up only. (Source: Glassnode)

Plus, a16z just announced a $2.2 billion crypto fund to invest in crypto networks and teams, the largest in the space. They're smart folks... right? There are still huge technological risks to surmount, but history lessons have taught me never to go short on innovation and human ingenuity.

"Crypto is not only the future of finance but, as with the internet in the early days, is poised to transform all aspects of our lives." - Chris Dixon, Katie Haun, Ali Yahya (a16z)

Searching for Awkwardness

Soon I started to ask myself: Is there a central ground where these 2 hypotheses can and will intersect? I like circles so I’m going to represent them with circles. It seems a little awkward for sure, like when you go on a blind date with a girl and then realise you actually don't speak each other's language. Healthcare is the antithesis of everything that the decentralised world stands for: dominated by large centralised entities who retain control on access, power & data, huge amount of regulation, high barriers to innovation.

(Fun Fact: I learnt that the region of intersection of 3 circles in a Venn diagram is known as a Reuleaux triangle. Later, Dr Tan Teck Jack kindly shared that 2-circle intersection is called Vesica Piscis. Trivia ftw).

I believe there is. Let me briefly explore some of the fundamental characteristics of blockchain and crypto-economic networks, use cases, how they may potentially solve problems in healthcare and lay out a few ideas for anyone to build upon - ie. why i'm excited.

Why I'm excited, Reason #1 - Today, we can build scalable decentralised, open-source systems.

Self-organising, distributed systems are more equitable, more efficient and more capable of repair at points of local failure (vs centralised entities). Decentralised systems can democratize and redistribute power/authority and potentially solve human coordination failures: which are unsurprisingly quite common in healthcare, such as overtreatment, low value care, pricing discrepancies.

What's wrong today:

In the current Web 2.0 era, the reality is that Silicon Valley and big tech companies have a financial monopoly on almost all the data and its value in the system. For example,

• Instagram has the rights to use every well-taken photo that you've ever posted (after spending hours conceptualising + trying different filters + optimizing for likes), and monetise it in anyway it can. You don't have rights to any revenue that's generated from it despite being the content creator. What a wonderful business model!

• Let's also consider why Google purchased Nest in 2014: does Google really want to pivot into the thermostat business? I daresay it is to plug a gap in their data harvesting - once you get home, you leave your phone on the table. Nest allows them to know exactly when you’re home and what you’re doing at home.

A Real Use Case for Decentralisation: Finance 

Decentralised Finance (DeFi) is evolving at breakneck speed. We're seeing glimpses of an entirely new, more equitable financial system with less friction points being designed from grounds-up by smart young folks frustrated by the status quo. In the permissioned TradFi world, the best investment opportunities are often only available to those with power and influence within the system. For example, if you happen to be employed as a partner at Goldman Sachs.

Decentralised financial systems like Aave (a borrow-lend protocol) cut out the centralised intermediaries - the banks - and allow anyone to take out a loan in seconds without filling in any paperwork, and with much lower fees. Entirely based on immutable smart contracts and code. You can even earn yields on your assets, >4% on USD stablecoins vs 0.1% in a bank (if you’re lucky) 

The rise of DeFi is an extremely fascinating discourse on its own, but this essay is really about healthcare so..

What we can do in Healthcare: 

In a similar fashion, the vast majority of healthcare data today is held in custody by centralised entities (hospitals, universities, governments, big tech). Often, the data is sitting around in silos and doing absolutely nothing because they don't know how/what to do with it: a heinous crime if you ask me.

There's an absolutely massive value for humanity when you pull together huge amounts of health data in one place: think about how we would be able to speed up medical research through "virtual" trials / decentralized trials / more efficient clinical trials or to improve clinical management through predictive analytics. A recent example: the Apple Watch heart study enrolled 419,000+ people in less than 4 months, distributed through out the globe and capturing racial diversity. One of the reasons this was doable was because the trial was fully decentralized.

Yet at a large enough scale, it would be foolish for all the sensitive health data to be held by a single centralised entity, which is vulnerable to security breaches, human errors, poor access controls, organisation failure.

De-identified medical data should be made widely available on highly-secure distributed networks, in an access-controlled way that protects individual data rights/privacy, while remaining open-sourced: a good blockchain enables all of these. Imagine how much quicker medical science could advance, if this was available.

Idea 1: Can we build a fully-decentralised network specifically designed for the efficient, secure handling and standardisation of various types of health data (eg. clinical data, user-generated data from apps/wearables, social determinants of health)? While having a high level of composability to accelerate innovation.

Taking this on would be basically playing in hard-mode: building good infrastructure is an extremely challenging technical feat. Not to mention, getting entrenched healthcare institutions to use 'blockchain', when they are already baulking at the idea of putting their data on the cloud. A few teams have tried previously (see Medilot, a Singapore-based startup) and apparently given up for reasons not clear to me yet. Perhaps it was too early. But I'm certain something like this will exist in 10 years time.

And composability...what's that? Which is a great segue to...

Why I'm excited, Reason #2 - Composability = Lego Blocks

Highly-composable platforms like blockchain networks enable interoperability, and allows products to be selected and assembled in various combinations with ease. Just like lego blocks: you can stack them on top of one another, or put together together in various ways to to build something amazing that would not have been in the instruction manual. Plus, unlike lego blocks, they can even talk to each other.

What's wrong today:

There are 100,000s of health apps, clinical decision support tools, electronic medical record software out there, and none of them communicate with each other. Anyone trying to be a data aggregator in the space faces a steep uphill battle and clearly enjoys eating glass. Most Web 2 products aren't built to be composable. And trying to pull data from others (eg via APIs) is difficult because you have to seek permission from the data provider/product owner to gain access. And often permission is granted only if it generates business value for the provider.

A more immediate example: Now that the gyms are open again here in Singapore, I've restarted a regular gym habit, trying to put on more lean muscle mass. My Fitness First app knows that I went to the gym because I had to make a booking in the app. But my FitBit app doesn't know this. My health data is spread across multiple apps which makes it difficult for me to get a complete picture of my health.

Apple has helped by providing easy access to data aggregating tools like Apple Health, but ultimately Apple controls the platform and it extracts a heavy toll for building on it (30% commission on app revenues). Far from ideal.

Potential Use Case in Healthcare:

Composability is important because it unlocks the compounding effect in innovation. New innovators can build and improve upon what has been previously built, without having to start from scratch. Do more with less. I have experienced - multiple times - the pain of building a new digital health app with a great concept but has to first begin by re-creating baseline functionality of something that already exists. This often wastes at least a year of development time.

Blockchain networks are permission-less and stateful, which are key enablers for composability. So you can build on top of them without worrying about your app failing when lower-level dependencies change. (This article by Jesse Warden from 2018 is a great read on the evolution of composable blockchain networks)

Imagine if you’re able to build health apps on a platform that can all pull data from each other, knowing what I eat, how much I’m sleeping, what my heart rate is. And killer idea here : automatically book me to a HIIT class if I've been slacking off on my gym routine for a week (ok maybe not). This is the only way we can realise the truly personalised future that we've been promised.

We are already seeing health legos being built: I'm heartened to see new companies like Zus Health, which just closed a $34M Series A round led by a16z. Zus Health is focused on creating the industry's first shared development platform backed by a shared data record, allowing digital health builders to create technologies and services without the usual blockers. As more people build on the platform, the value of the platform increases exponentially (through data capture and network, compounding effects). It is not using blockchain technology AFAIK.

Why I'm excited, Reason #3: Tokenisation of health can drive adoption and behaviour change

I've shared a couple of big picture ideas that can transform the digital health sector. And now we're getting to something I'm super stoked about, something more immediate, more real and can be done right now..

Blockchain networks make it easy to create, distribute and keep track of digital tokens in a permission-less, trustless manner. On Ethereum, you can create and launch a ERC-20 token on the network in under an hour.

And now think of digital tokens as a potential unit of health accounting, similar to incentive points issued by insurers (eg. AIA Vitality points) or government bodies (Singapore Health Promotion Board's Healthpoints) to reward users for healthy behaviours - but with additional superpowers. By nature, they can be interchangeable (similar to how you might swap airline miles among different rewards programmes), transferrable to other people, exchanged for goods, sold on exchanges, staked for more rewards etc. With more utility, these 'healthpoints' will have a greater value in the eyes of the consumer.

What's wrong today:

This is something I think is broken in our society. For the large majority of products and companies, their users/customers have no interest in the success of the company. The relationship is very transactional: users pay (in time, money or data) and get something in exchange.

Take Mcdonalds as an example, I would guess that <1% of people who eat at Mcdonalds owns shares in Mcdonalds. Most people are there to get their burger & fries and want the cheapest deal. Myself included - my body regularly craves for a McSpicy meal. I don't own any MCD shares. Sorry i'm not interested in helping McDonalds become the best fast food chain in the world.

What we can do in Healthcare:

In my past 14 years spent as a practising clinician and digital health technologist, I believe the biggest, top-level problem in health is changing health behaviours. It's the holy grail of healthcare. If we can get people to eat better, exercise regularly, sleep more and practice mindfulness, we would prevent huge amount of health problems downstream.

For digital health products / digital therapeutics, how do we get users to engage and continue using the products we create? ie. User engagement and retention. That's the biggest challenge I've always faced - it's gargantuan effort that requires a lot of design thinking, great UI/UX, incentive models. Stickiness is a tough nut to crack in digital health because the feedback loops and habit-forming cycles are long and sometimes not apparent. Yet it is even more critical because only users who are engaged will reap the health benefits.

Now we have another tool, a powerful one: If I (as a user) own a piece of the health app and get a share of rewards from engaging with the health app - eg. through data monetisation, licensing fees, subscription fees - I will be much more motivated to keep using it, and tell other people about it. I might even be able to vote and have a say in the direction of the product, deciding on which new features to build. This can be enabled by digital tokens. It's the greatest growth marketing tool available. The success of this governance / dividend token model in DeFi is promising and could be an example for many other industries to follow.

"Give a man $5 and he'll help you for a day. Give a man a stake in your organisation, and he'll help you for a lifetime" - An anonymous wise man

Idea 2: Through well-designed token-economics, can digital tokens incentivise users towards higher engagement, retention and eventually behaviour change? By empowering them to be stakeholders of the health product they're using.

And...we haven't gotten to NFTs yet and the tokenisation of health data. Anyone up for putting your genomic data on a token? This deserves a separate article on its own and is next on my list.

An ending note on the beginning

There are clearly many challenges to overcome: infrastructure is still being developed, major friction points in onboarding non-crypto-native users onto the blockchain, limited integrations on mobile apps, transaction fees etc. It may seem even a little crazy, but that just shows how early we are. I do think that some of these ideas will also not materialise in the traditional 'startup' way of incorporating a company and raising VC equity funding to build a product.

In the end, for these ideas to work, all of the complexity of the technology needs to be abstracted away at the backend. So all the user sees is a simplified, magical, superpower-ed healthcare experience.