Excellent discussion with Kelly Watson + Tidepool team, informing #ArgonautProject on CGM-to-EHR writes. youtu.be/jMXY3vNOA3k covers:
- Reimbursement for remote monitoring
- Population dashboards
- Periodic & on-demand writes
- Provisioning devices/accounts
HL7 International #FHIR

More fun with Electronic Health Information Export! Quick demo exploring my data the hard way (manual inspection) + the increasingly easy way: with Large Language Models.

Making an appearance here: the whole Claude 3 family, and cameo from Mixtral 8x22B.

youtu.be/bnfVjvwi4T4

Great conversation with Mikael Rinnetmäki, founder Sensotrend about integrating CGM data into Finland's national PHR & many EHRs. Showcases pt-mediated access & the super-powers of keeping & using raw source data. HL7 International #ArgonautProject #FHIR #Interoperability

youtube.com/watch?v=wolGqS…

First cut of my Epic EHI export @ github.com/jmandel/my-hea… ! Still missing vital signs ('flowsheet') and some notes, but it's progress. See 'json' folder for my processed, merged, and slightly-easier-to-read annotated output. Let me know if you want to try on your own data.

Superb insights from Rimidi's Lucienne Ide on CGM data. We cover summary metrics vs 'raw' sensor data, schedules of push/pull, and EHR-integrated SMART Health IT on #FHIR for onboarding patients.

Critical background for HL7 International Argonaut 2024. Watch: youtube.com/watch?v=up_vdS…

Thanks to Tim Kessler & Redox team for sharing the real-world take on CGM integration into EHRs.

Challenges: identity matching, enrollment friction, data mapping, heterogeneous workflows. Input will inform HL7 International Argonaut 2024 CGM project.

Full interview: youtu.be/ukbmWOGd2H4

The UC San Diego Health sepsis AI study is the first to show improvement in clinical outcomes with a deep-learning AI model

Credit to the team for focusing on workflow, since healthcare AI is about more than algorithms – as the
editorial highlights! (5/5)

👉nature.com/articles/s4174…

This Change Healthcare hack causes real issues for those who rely on manufacturer coupons for expensive medications, especially rare disease patients. Just paid $1260 at the pharmacy because Change powers all the manufacturer coupons and it has been down for 10 days.

Writing converts your ideas from vague to bad. But once they're written down you can at least see and fix the badness.

It was such fun to chat with Jaansi Patel, a high school student on her LearnOn podcast. We discussed where does technology go wrong? Jaansi will help restore your faith in the next generation of scientists. ⁦HIStalk⁩ ⁦AMIA⁩ ⁦ podcasts.apple.com/us/podcast/lea…

Giant in the field of medical decision-making/decision-analysis has left us. legacy.com/us/obituaries/… This obituary is worth the read. He changed the way I think about maximizing utility. Not mentioned in obit is a life-long challenge Steve had because he did not menton either. 1/

EHI Export insight: over 100 individual users (i.e., provider org staff) contributed to some aspect of my EHI in the past 6 years of ambulatory care.

This diagram shows only users & tasks mentioned at least 10 times. (Task labels generated by GPT-4, based on Epic's EHI schema.)

Thanks to Epic team for reviewing with me! Sounds like my provider needs to update config before my EHI Export will include:

1. Vitals+flowsheet data
2. Notes
3. Data dictionary for many concept ids

The challenge: every provider may have *thousands* of EHI config knobs/dials :/

It's striking how little of my lived experienced 'dealing with health issues' over the past four years really shows up in the broad strokes of the data.

Though it looks like clinical notes are missing from my EHI export, so that obscures some of the story.