domains/health/AI scribes reached 92 percent provider adoption in under 3 years because documentation is the rare healthcare workflow where AI value is immediate unambiguous and low-risk.md

@@ -49,6 +49,12 @@ The 92% figure applies to 'deploying, implementing, or piloting' ambient AI as o
 
 WVU Medicine expanded Abridge ambient AI across 25 hospitals including rural facilities in March 2026, one month after Epic AI Charting launch. This rural expansion suggests ambient AI has passed from pilot phase to broad deployment phase, as enterprise technology typically enters academic medical centers first, then regional health systems, then rural/critical access hospitals last. The fact that a state academic health system serving one of the most rural and medically underserved states chose to expand Abridge post-Epic launch provides implicit market validation of Abridge's competitive position.
 
+
+### Additional Evidence (extend)
+*Source: [[2026-01-01-openevidence-clinical-ai-growth-12b-valuation]] | Added: 2026-03-18*
+
+OpenEvidence (clinical reasoning) and Abridge (documentation) represent two distinct clinical AI beachheads serving different workflows. OpenEvidence's 40%+ physician adoption for clinical decision support suggests that point-of-care reasoning may be reaching similar adoption velocity as documentation AI, but with different trust barriers: 44% of physicians remain concerned about accuracy/misinformation and 19% about lack of oversight, even among heavy users.
+
 ---
 
 Relevant Notes:

domains/health/OpenEvidence became the fastest-adopted clinical technology in history reaching 40 percent of US physicians daily within two years.md

@@ -17,6 +17,12 @@ What makes this significant is the adoption speed. Reaching 40% of US physicians
 
 The incumbent response is UpToDate ExpertAI (Wolters Kluwer, Q4 2025), leveraging its trusted brand and install base. The competitive dynamic -- startup vs incumbent in clinical decision support -- will determine whether AI clinical knowledge becomes a winner-take-all market or fragments.
 
+
+### Additional Evidence (extend)
+*Source: [[2026-01-01-openevidence-clinical-ai-growth-12b-valuation]] | Added: 2026-03-18*
+
+OpenEvidence reached 20M clinical consultations/month by January 2026 (2,000%+ YoY growth from 8.5M/month in 2025), achieved 1M consultations in a single day on March 10, 2026, and is now valued at $12B (up from $3.5B). Used across 10,000+ hospitals nationwide. First AI to score 100% on all parts of USMLE.
+
 ---
 
 Relevant Notes:

domains/health/medical LLM benchmark performance does not translate to clinical impact because physicians with and without AI access achieve similar diagnostic accuracy in randomized trials.md

@@ -17,6 +17,12 @@ A deeper finding from a Stanford/Harvard study challenges even the "similar accu
 
 The implication for AI deployment strategy: the highest-value clinical AI applications are not diagnostic augmentation but workflow automation (ambient documentation, administrative burden reduction) and safety netting (AI triage catching missed findings). The centaur model may still apply to medicine, but the interaction design must prevent physicians from overriding AI on tasks where AI demonstrably outperforms -- a politically and ethically charged constraint.
 
+
+### Additional Evidence (challenge)
+*Source: [[2026-01-01-openevidence-clinical-ai-growth-12b-valuation]] | Added: 2026-03-18*
+
+OpenEvidence's 100% USMLE score and scale to 20M monthly consultations with 40%+ of US physicians creates the first real-world test case. If benchmark performance doesn't translate to outcomes, this should be detectable at population scale — but no outcomes data has been published despite the massive deployment.
+
 ---
 
 Relevant Notes:

inbox/queue/2026-01-01-openevidence-clinical-ai-growth-12b-valuation.md

@@ -7,9 +7,13 @@ date: 2026-01-01
 domain: health
 secondary_domains: [ai-alignment]
 format: company-announcement
-status: unprocessed
+status: enrichment
 priority: medium
 tags: [openevidence, clinical-ai, decision-support, physician-adoption, clinical-decision-support, health-ai, trust]
+processed_by: vida
+processed_date: 2026-03-18
+enrichments_applied: ["OpenEvidence became the fastest-adopted clinical technology in history reaching 40 percent of US physicians daily within two years.md", "medical LLM benchmark performance does not translate to clinical impact because physicians with and without AI access achieve similar diagnostic accuracy in randomized trials.md", "AI scribes reached 92 percent provider adoption in under 3 years because documentation is the rare healthcare workflow where AI value is immediate unambiguous and low-risk.md"]
+extraction_model: "anthropic/claude-sonnet-4.5"
 ---
 
 ## Content
@@ -68,3 +72,14 @@ This creates a two-track clinical AI story: (1) Abridge/ambient scribes for docu
 PRIMARY CONNECTION: [[OpenEvidence became the fastest-adopted clinical technology in history reaching 40 percent of US physicians daily within two years]]
 WHY ARCHIVED: Significant scale update — the existing claim understates 2026 metrics by an order of magnitude. Also: USMLE 100% creates the benchmark vs. outcomes tension in practice, not theory.
 EXTRACTION HINT: Update the existing claim with scale metrics, but flag the benchmark-to-outcomes translation tension as a challenge to both the OpenEvidence claim and the benchmark performance claim
+
+
+## Key Facts
+- OpenEvidence valued at $12B as of January 2026, up from $6B three months earlier and $3.5B in 2025
+- 20M clinical consultations/month by January 2026, up from 8.5M/month in 2025 (2,000%+ YoY growth)
+- 1M clinical consultations in a single day on March 10, 2026
+- Used across 10,000+ hospitals and medical centers nationwide
+- 44% of physicians concerned about accuracy and risk of misinformation
+- 19% of physicians concerned about lack of physician oversight or explainability
+- Series D: $250M led by Thrive Capital and DST Global (January 2026)
+- First AI to score 100% on USMLE (all parts)