Stream: blockchain
Topic: Myth-busting
Josh Mandel (Nov 01 2017 at 21:53):
Really nice myth-busting write-up at: https://blog.andreacoravos.com/myth-busting-can-a-blockchain-save-healthcare-d398cdebf0c1 (I know this here is hardly the crowd for myth-busting, but I couldn't not share :-))
Doug Bulleit (Nov 02 2017 at 22:07):
Josh. Thanks for posting this. FWIW, IMO most all of the arguments behind these myths are easily debunked; I'd be happy to take swings at knocking down any you'd care to name except #4 (which, I suspect, you and the Seek-For-Science crowd could demolish even quicker). Indeed, while benevolent skepticism is always useful, the whole tenor of these "blockchain over-hyped" pieces are reminiscent of the now infamous Amazon Dot Bomb piece that was widely circulated in the early days of Web 1.0. `just one man's opinion, but Web 3.0-- i.e., the Web of Trust-- is likely to make more and bigger fortunes than Mr Bezos has yet to imagine.
Grahame Grieve (Nov 02 2017 at 23:31):
I don't think they are easily debunked. Other than #1. So perhaps you should try here
Doug Bulleit (Nov 03 2017 at 02:44):
OK then...I'll try to be succinct...inviting further rejoinders
Myth #2 [PHI not on blockchain] half answered in BLOG itself; but FHIR Resources don't need to be "stored" anywhere (other than within their native EHR servers). A certain kind of blockchain, however, could comprise nothing more, nor less, than a decentralized repository of URI/locations, smart contract-created access tokens, self sovereign identity attestations, provenance and other irrefutable authentication and authorization information. Multiple layers of encryption, hashed trees, and digital wallet access control are, by their nature, more secure than standalone UID/PW's operating over TLS.
Myth #3 This one's complicated by larger business objectives (having to do with a well functioning ecosystem) as well as more secure PII and a dramatically-narrowed attack surface. But the long and short of it is this: a permissioned blockchain can both better protect decentralized resources as well support a multilateral business model--one in which no peer-node player exerts control over another.
Myth #4 I'll leave this one to Josh
Myth #5 There is no IT case that I know of where a decentralized system is more efficient than traditional hierarchical alternatives. That said, referring back to Myth #2, centralization inevitably connotes the opportunity for the central player to restrict others access, and/or exact monopoly rents, and/or censor data, etc. Thus, no one "really needs" a blockchain; but in the situations where an intermediary control point is unwanted, it introduces new business model options. And, while all of the "tools" may not be ready for prime time, their pace of evolution is breathtaking.
Myth #6 "Technology [doesn't] always make life better" But, over time, it usually does; and, is this one we really want to bet against?
Lloyd McKenzie (Nov 03 2017 at 03:18):
If the data continues to live in the EHR repositories, then the EHR maintains final control over who has access to the data - regardless of what a block chain smart contract might suggest. So I don't see how blockchain allows the data to flow more freely. In theory, it might provide the patient with a single interface point to declare their wish to share - should all relevant EHR systems choose to register their data on the block chain. But a patient-managed PHR could do the same - and provide a single point of data access as well.
Doug Bulleit (Nov 03 2017 at 03:29):
Yes, but..., if an expanding group of health system/IDN's confederate with one another, within a permissioned blockchain, each agreeing to honor and validate one another's subtending app's and registered users' credentials, couldn't a new kind of network effect come into view?
Lloyd McKenzie (Nov 03 2017 at 03:41):
I'm not clear why blockchain is necessary there. The work there is formation of the group, agreement of policies around access (and consent and security tags, etc.) If they can get the agreement in place and recognize credentials, why bother with block chain at all?
Doug Bulleit (Nov 03 2017 at 03:56):
It's not necessary, just elegant--i.e., a permissioned blockchain can be described as nothing more, nor less, than a P2P arrangement of algorithmicaly enforced Trust. Moreover, as inter-nodal Trust ( a.k.a., the rules and policies that are encoded into smart contracts within the blockchain) become established and proven, the new ecosystem could expand dramatically ( sans the need for complex bilateral agreements, 3rd-party HIE organizations, etc)
In fact, IMO, one of the more interesting question being debated within permissioned blockchain circles is not technical at all; rather, the discussion involve minimal but essential governance, admission, token pricing and other policy. And, this is an area where the FHIR community could exert extraordinarily useful influence
Last updated: Apr 12 2022 at 19:14 UTC
