Frank Font

This is absolutely brilliant and one of the few (if only) examples I have seen of declarative UI component design that takes advantage of the power of XSLT since the work we did on SemanticDB way back in the day (we built an entirely declarative architecture with an XML-based UI component library that was transformed to an XForms application executed entirely in the browser). I have forever felt that we have been sleeping on the power of using UI component vocabularies with XSLT to facilitate "separation [that] keeps designs stable and predictable" as stated in the included Claude skill. https://lnkd.in/gK83eVUF #XSLTPattern #AI #DeclarativeUI #Web3.0+

13

For years, government efforts at modernizing the National Directories have been sort of stuck in Web 1.0 models, myopically focused on traditional client-server approaches. This past weekend, 8 years of R&D culminated in something that could make a lot of those architectural assumptions obsolete: FHIR over Interplanetary File System (IPFS). I've been noodling on this since 2017, reverse engineering the Ethereum stack, and was finally able to put all the pieces together this weekend into a full utility service that will likely be deployed in all of my future projects.  IPFS is a P2P distributed file system, similar to BitTorrent or WebTorrent protocols; but that uses DAG and merkel trees to address content, provide load balance, and allows clients to download files from multiple peers. Initial use case is directory services, config files, terminologies, reference data sets, and data dictionaries. So, say you have a FHIR endpoint directory; or a copy of the NPPES National Directory.... something that is a few gigabytes large. Client A seeds the file and creates a share across the IPFS gateways; clients B, C, D, and E all download a copy of the directory. Then Client A goes offline. Client E can still download the file from B, C, and D. And what's even better, there are still the same type of bandwidth gains as torrents and other P2P systems. If client E is on fiber at gigabit speeds; and A, B, C, and D are all on 100MB/s, the incoming data stream will be 300MB/s. There will obviously be bottlenecks in the network; but all else being equal, once a large enough swarm is established, the data will both distribute evenly, reduce load on any individual node, and scales to whatever bandwidth a new node has available to receive (or send) data. More importantly, this allows participating distributed apps (DApps) to all remain up-to-date and in sync with directory updates, terminology updates, etc who are participating in a QHIN. If you're running MongoDB as your FHIR data store, and are interested in early access or piloting this technology, reach out. No blockchain or Ethereum needed. We’re creating something similar, but different.  A FHIR data layer.  And for US participants, we may offer TEFCA compliance soon also. Good chance this gets released as the 'Swarm' module for Node on FHIR, but branding is still to be determined.

6

1 Comment

Most businesses have a default bias away from opensource approaches. That isn’t hard to understand, but is mostly about conditioning than carefully considered strategy. This provides more light and enables informed choices. #opensource #proprietary #tco #digitalhealth #digitalization #transformation

2

Assistant Secretary for Technology Policy (ASTP/ONC) is finally signaling that they are cracking down on information blocking. Big activity in Washington last week: on Sept 3, U.S. Department of Health and Human Services (HHS) announced a new push to “Crack Down on Health Data Blocking."(https://lnkd.in/evGW4mZj) The very next day, ASTP/ONC and the HHS Office of Inspector General convened third party vendors at their DC offices for an invite-only all-day session on how enforcement will actually play out. Medcurio was in the room. The recording of presenters will speak for itself, and here are the 5 signals I took away from DC: 1) ASTP/ONC wants info blocking claims. They asked for them repeatedly. Don’t agonize over the submission, tell the story and let them evaluate. The specifics matter, and so does volume. 2) Volume drives enforcement. ONC was candid that the nuances of info blocking matter to identify and address loopholes in the regulations. More claims can help them close gaps. 3) Info blocking scope is broad. Once a vendor certifies a module, all of the company’s behaviors are subject to info blocking rules. 4) Discouragement counts as blocking. Info blocking isn’t only a hard “no.” Subtle tactics (ex: whisper campaigns, fearmongering around “security/performance,” extra approval layers) can qualify. If it systematically chills legitimate use, it’s in scope. 5) Culture is the hidden bottleneck. Providers solely own the decisions on what IT solutions to adopt. But, too often they don’t use best in class third party solutions because their EHR vendor has created a climate of FUD (a form of info blocking) that deters action. DHHS Deputy Sec’y Jim O’Neil emphasized the implications of info blocking to AI and automation-driven innovation. “The future is already here, it’s just not equally distributed.” In other industries, AI and automation are scaling at speed. But in healthcare the distribution gap is glaring. And it's largely because EHR data sits in a locked filing cabinet, with ambiguity and EHR vendor behaviors keeping it closed. ASTP/ONC is now asking the market to surface those ambiguities so they can close the gaps. Want my full debrief? Send me a DM to connect.

22

2 Comments

This video by @InterSystems introduces the OMOP Platform, a new service by InterSystems designed to support researchers by transforming EHR data into the OMOP format. 👉 https://lnkd.in/gAq42nhm The platform features a FHIR-to-OMOP transformation pipeline and an OMOP repository hosted on #InterSystems #IRIS Cloud #SQL.

8