A timeline of web1, web2, web3 in healthcare
And why I think web3 for healthcare is inevitable—although years in the making
I’ve been wanting to write a piece about web3 and healthcare for months. I imagined a big hot take, and I spent a lot of time reading and researching (although I’ll confess that I have yet to set up a wallet).
…And my conclusion was the vaguely lame one that the field is just too early. There’s a lot of promise, and almost nothing specific (healthcare-wise) to get excited or antsy about. Am I missing something? Let me know.
Otherwise, here is a brief timeline of healthcare tied to the main stages of the web, and a reading list I’m working my way through as I wait for healthcare applications to explode.
From roughly the 1990s to the early 2000s, the average computing experience was limited to downloadable software and HTML pages. Interaction was with the information, rather than other users.
A painfully slow transition from paper records to electronic health records was eventually sped up by HITECH in 2009. EHRs represented a breakthrough in health IT but had serious issues: Lack of interoperability and gatekeeping of patient records by institutions, healthcare worker alert fatigue (EHRs generate an average of between 64 and 164 alerts per ICU patient per day, the majority of which are non-urgent), and necessary workflow changes that saw doctors spending increasing amounts of time staring at a screen.
As Dr. Emily Silverman put it in a New York Times op-ed published in 2019:
Using [the EHR] felt less like having a conversation with a helpful colleague and more like standing in a hellish echo chamber of dissonant voices. From the ’80s.
I debated whether EHRs should even count as web1; after all, they’re not really dpeendent on an Internet connection, and the software is accessible specifically to computers within a hospital network. On the other hand, EHRs are a centralized platform that allows users (within a hospital network) to share data to some extent. The sharing plus the locality of that sharing led me to categorize EHRs as web1, although I’m open to other interpretations.
Dominated by social media, this era began in the mid-2000s and continues today. The average computing experience was interaction with other users on platforms owned by large tech companies.
Perhaps one of the first people to try to evolve healthcare software from web1 to web2 was Jim Clark, the founder of Netscape.
In the mid-1990s, Clark hired a bunch of smart people, put them in a room, and called the new company Healtheon. With limited healthcare experience, the team was quickly overwhelmed, and the company eventually merged with WebMD. Although Healtheon had big ambitions to use software to connect healthcare entities, the venture collapsed into decidedly non-interactive HTML pages describing disease.
Other notable, more successful efforts have included platforms that connect patients with doctors, like ZocDoc; platforms that connect doctors with doctors, like Doximity; platforms that ferry data between institutions, like health information exchanges; commoditized telehealth platforms and more specialized virtual care pathways; and—less technical but very prevalent—patient forums and social media, where patients share (sometimes wacky) treatments and provider recommendations.
All of these web2 products have undoubtedly improved healthcare for providers, insurers, and patients alike. But the software is owned by big companies, updates are often centered around billing utility rather than user experience or simplicity, and the result is an expensive, fragmented, behind-the-scenes morass of software. (Plus a lot of disinformation that spreads even as tech companies worry over how to handle it.)
Beginning in the last few years, web3 is a combination of cryptocurrency, blockchain records, DAOs (decentralized autonomous organizations), and more, which (at least in the ideal form) devolve power from big tech companies to individual users, giving people more control over their interactions and transactions.
Packy McCormick recently wrote about how web3 attempts right now, silly as they may seem, are building skills and techniques. Much like building with legos allows for errors, the DAOs of 2021 will allow users to experiment, fail, and iterate. In his opinion, these are all necessary steps to come before real-world, widely adopted web3 technology can be built.
The most obvious application for web3 in healthcare is remaking EHRs, the keystone web1 technology. Shifting health records from siloed software to interoperable, patient-owned, immutable records could solve a lot of problems.
(The static nature of EHRs has become an increasing liability in an interconnected world. As covered on the Tradeoffs podcast, there has been a recent wave of ransomware attacks on U.S. hospitals and provider offices—a 61% increase in 2020. Many of the offices are forced to pay the ransom to access patient records without disrupting care for longer than a few hours or days.
There are other compelling reasons to transfer EHRs to the blockchain. A recent article in STAT covered how emergency medical professionals very often don’t have access to EHRs—because of the complexity of transitioning a static software to a dynamic environment. And because no one entity wants to pay to enable that.)
Besides EHRs, I’m sure there are a lot of places where decentralized information and transactions could improve healthcare—the problem is that the rest of the system remains centralized. Healthcare is one of the few industries that remains firmly IRL (in real life), and any kind of web3 application will have to contend with that.
In short, I think web3 could have transformational effects on healthcare—but it’ll take a lot of tinkering with legos before it even comes close.
What I’m reading to learn more
Not Boring, by Packy McCormick, especially:
Gaby Goldberg’s web3 reading list, especially:
Chris Dixon’s work at a16z, especially:
Also Why Decentralization Matters, by Chris Dixon
Product vs. Protocol: Finding a Balance in Web3, by Jesse Walden, Variant
This information shouldn’t be taken as investment advice (obviously), and the opinions expressed are entirely my own, not representative of my employer or anyone else.