Pulse: ADLM Signals, Cyclospora Surges, and CLIA Redraws the Lab

# ADLM Signals, Cyclospora Surges, and CLIA Redraws the Lab## SummaryChris and Aakash break down three major signals from ADLM 2026: simultaneous centralization of high-throughput systems and decentralization of point-of-care testing, AI shifting from demos to implementation reality with data infrastructure as the real bottleneck, and the Cyclospora outbreak exploding to 4,200 confirmed cases across 41 states. They also unpack CMS and CDC's request for information on modernizing CLIA (last updated 1988) for AI-assisted diagnosis, remote oversight, and data-only facilities, plus the FDA-approved Elecsys sFlt-1/PlGF ratio test for pre-eclampsia risk stratification in hospitalized pregnant patients 23–34 weeks gestation.## Key takeaways- **Centralization and decentralization are happening simultaneously** — vendors are building massive high-throughput analyzers for core labs while pushing smaller instruments closer to patients, creating a "mismatch" in product strategy.- **AI implementation has moved past the demo phase** — the real questions now: who monitors data drift, who owns the data, who tracks software versions, and how performance varies across patient populations.- **Data infrastructure is the bottleneck, not algorithms** — companies gaining traction (like Bunker Hill) function as infrastructure plays, solving how to connect to health system databases; CAP is prioritizing AI inspection frameworks.- **Cyclospora outbreak is not slowing** — 4,200 lab-confirmed domestic cases, 7,400+ probable cases under investigation, 300+ hospitalizations, 41 states; CDC warns reporting lags illness onset by ~6 weeks; linked to Taylor Farms iceberg lettuce recall (~2,000 confirmed cases).- **Testing gaps obscure true outbreak scope** — Cyclospora isn't on routine GI panels; genetic tracking is harder than Salmonella (no whole-genome sequencing); labs may need to spin up testing if this becomes seasonal.- **CLIA modernization RFI is a critical inflection point** — CMS/CDC asking whether AI-assisted interpretation constitutes a "test," whether data-only facilities count as labs, and how remote competency assessment works for 90+ clinic locations.- **FDA-approved pre-eclampsia test (Elecsys sFlt-1/PlGF ratio)** — 91% sensitivity, 77% specificity for predicting severe pre-eclampsia within 2 weeks in singleton pregnancies 23–34 weeks; intended to support (not replace) clinical judgment; education needed to prevent off-label confirmatory use.## Chapters- Opening and Personal Catch Up- ADLM Signals Overview- Centralization and Decentralization Trends- AI Implementation Beyond Demos- Data Infrastructure as AI Bottleneck- Cyclospora Outbreak Escalation- CLIA Modernization Request for Information- FDA Approved Pre-eclampsia Test## Mentioned- **ADLM** (Association for Diagnostics & Laboratory Medicine) — 2026 conference- **Bunker Hill** — health AI company focused on data infrastructure/connectivity- **CAP** (College of American Pathologists) — prioritizing AI inspection frameworks- **CDC** — Cyclospora outbreak tracking, reporting lag warnings- **CMS** — CLIA modernization request for information (with CDC)- **Taylor Farms** — iceberg lettuce recall linked to Cyclospora- **Elecsys sFlt-1/PlGF ratio** (Roche) — FDA-approved pre-eclampsia risk test- **CLIA** (Clinical Laboratory Improvement Amendments) — 1988 legislation under review## Quotes> "The data infrastructure they call it a layer... is going to be as important as anything else as far as AI implementation... a lot of the health AI companies that are actually gaining real traction are almost like infrastructure companies."> "The core important questions still need to be asked as, like who monitors the data drift, who owns that data? Who tracks these software versions and like who checks the performances differences amongst different patient populations."> "It used to just be build the walls higher and thicker, right? That's pretty much what we thought of and now it's a very different world."(00:00) - Opening and Personal Catch Up (01:32) - ADLM Signals Overview (02:41) - Centralization and Decentralization Trends (04:00) - AI Implementation Beyond Demos (05:44) - Data Infrastructure as AI Bottleneck (08:07) - Cyclospora Outbreak Escalation (13:59) - CLIA Modernization Request for Information (19:29) - FDA Approved Pre-eclampsia Test