Clinical AI in practice
AI in day-to-day care: documentation, triage and decision support.
capabilityEmerging Open on the map →
Facts
- Lane
- Science and health
- Wave
- 6 of 12 (build order, not a date)
- Needs
- 11 boxes all the way back through hard links
- Unlocks
- 2 boxes downstream through hard links
- Leverage
- 0.2 points · stands in the way of 1 of 53 stages · rank 46 of 57
- Cluster
- Science and medicine
Evidence as of October 2026
In a study of about 1,800 clinicians, ambient AI scribes saved 16 minutes of documentation per eight hours of care and reduced burnout. Useful, but modest.
Detail boxes 2
Ambient scribes raise billed coding intensity and insurers answer with automated downcoding and audits, so payment acceptance becomes the constraint on clinical AI.
Scribe users billed 1.81 more RVUs a week (UCSF, JAMA Network Open), Cigna has auto-downcoded many E/M claims since 1 Oct 2025, and BCBSA estimates about $2.3B of added spending linked to AI-enabled coding (Mar 2026).
Payer billing codes that pay for AI software per use, which is what makes hospitals buy it.
CMS's CY2027 outpatient proposed rule (7 Jul 2026) would pay separately for 36 'Software as a Medical Service' codes through New Technology APCs; the final rule normally appears in early November.
Around it
Brakes
Loops and off-ramps
Openings researchers' judgement, October 2026 scouting
- Depth
- light
- Players
- Abridge ($5.3bn, ~$100m ARR); Microsoft DAX/Nuance; Ambience; Nabla; Freed; Heidi; Epic native scribe
- Wedge
- Outcome-level RCT evidence for clinical decision support; scribe products are commoditising. research