Assess a patient and recommend management
See the requirement-level assessment and its sources.
SCIENCE UNDER AUTOMATION
AI in Clinical medicine: capabilities and evidence
Clinical work connects diagnosis to treatment, follow-up and responses to complications in real patients.
Autonomy · up to 4.4 / 10
Model interval: 2.1–4.4. Assessed 21 October 2026.
Two equally weighted workflow families: a bounded operational task and an investigation or deployment beyond that task. Inherited subfields use this parent scenario.
Calculation and work mix
Each requirement uses the operational scale from 0 (independent execution unestablished) to 4 (reliable independent execution in the stated setting). A workflow takes its lowest requirement level. The two workflows receive equal weight; the headline is 1 + 9 × weighted level / 4.
These levels are source-linked model assignments, not measurements reported by the papers. “Up to” is the upper model bound, not a probability or a forecast date.
Assess a patient and recommend management: 1–2 / 4.
Deliver treatment and manage an evolving clinical course: 0–1 / 4.
Upper bound with 75:25 workflow weights: 4.9. With 25:75 weights: 3.8.
What has AI demonstrated?
An agentic AMIE system was compared with 21 primary-care physicians across 100 simulated multivisit cases and was non-inferior on specialist-rated management reasoning.
- Test
- Peer-reviewed randomized blinded virtual clinical evaluation
- Comparison
- Comparators vary by study; see the linked sources.
- System and version
- AMIE disease management · 2026-06-17
- Evidence checked
- 2026-10-21
Virtual consultations and medication questions do not test treatment delivery, physical examination or longitudinal outcomes in patients.
What does the field require?
Assess a patient and recommend management
- Gather relevant information and reason about management
Autonomy level: 2–3 / 4.
AMIE supports multivisit management reasoning in the simulated clinical evaluation.
AMIE disease management - Integrate examinations and clinical context
Autonomy level: 1–2 / 4.
Virtual cases do not supply autonomous physical examinations; real clinical officers retain responsibility in the trial.
AMIE disease management · Generative AI decision support in primary care - Verify the plan before acting on it
Autonomy level: 1–2 / 4.
Guideline grounding supports recommendations, while clinicians execute and check patient care.
AMIE disease management · Generative AI decision support in primary care
Deliver treatment and manage an evolving clinical course
- Execute treatment safely in a real patient
Autonomy level: 0–1 / 4.
Ex-vivo robotic steps are demonstrated, while the real-patient trial retains clinical officers for treatment.
SRT-H, v3 · Generative AI decision support in primary care - Recognize complications and adapt care
Autonomy level: 1–2 / 4.
Multivisit reasoning has been simulated; responsibility for real complications remains with clinicians.
AMIE disease management · Generative AI decision support in primary care - Verify sustained clinical benefit
Autonomy level: 0–1 / 4.
The randomized trial measured outcomes under clinical supervision and found no significant primary-outcome benefit.
Generative AI decision support in primary care
Research tasks
Deliver treatment and manage an evolving clinical course
See the requirement-level assessment and its sources.
Can AI replace the whole workflow?
The results above measure specific tasks. The field assessment also requires completion of the connected workflow, including the requirements listed here. A field-wide replacement percentage has not been established.
Read the assessment method