Scenario Architecture
One clinical prompt generates a structured case: opening vignette, evolving vitals, labs, imaging, decision points, and a management arc — all medically coherent.
How it works
Clinical reasoning training built on evidence you can trace.
Clinical Rounds generates structured medical cases with real decision points, knowledge-graph context, guideline-anchored feedback, and post-session analytics — designed for residents, attendings, and faculty teams.
Platform capabilities
Six systems, one clinical workflow.
Step-by-step
From prompt to structured review.
01
Generate
Enter a diagnosis, syndrome, or short clinical scenario. The engine constructs a coherent patient, a logical disease course, and the decision architecture for the session.
< 2 minutes to first case
02
Navigate
Work through the case as it evolves. Order investigations, reassess after new data, choose interventions. The platform tracks every decision and its timing.
Decision trail captured
03
Decide with Evidence
At each key decision point, see the guideline anchor, the knowledge graph constraint, and the counterfactual — what changes if the patient context changes.
Guideline-anchored feedback
04
Review the Outcome
After the case closes, the session generates a structured review: reasoning quality, stewardship score, evidence coverage, and where the trajectory diverged from the expert path.
Structured synthesis
Evidence at the decision point
Guideline excerpts surface when the decision matters.
Every significant management choice is anchored to the specific guideline excerpt that should inform it — not a bibliography at the end of the case, but the exact recommendation at the point of divergence. Source, strength, and applicability are shown in context.
- Guideline excerpt appears at the decision, not after the case
- Society, year, and evidence strength shown
- Counterfactual shown when management could change
- Linked citations — one click to full source
Evidence card · AHA/ACC 2021
Guideline anchor — AHA/ACC 2021
Low-risk acute chest pain → use structured risk assessment first; urgent testing not routinely required before evaluation completion.
✓ Evidence present at the moment you made the call — not discovered in the debrief.
Knowledge graph · Suspected PE workup
Default recommendation:CT Pulmonary Angiography
+ Pregnancy (32 wks)Prefer V/Q scan — reduces maternal breast radiation
+ CKD Stage 4 (eGFR 22)Avoid iodinated contrast — alternative pathway required
Context-aware guidance
Same case. Different patient. Different path.
The knowledge graph maps patient constraints — renal function, pregnancy, bleeding risk, drug interactions — to the recommendation space. When a constraint changes the optimal path, the platform shows you which factor changed and why the recommendation shifted.
- Constraints are traced to the recommendation, not assumed
- Factor that triggered the change is named explicitly
- Learner sees the "why" at the decision point, not in a footnote
- Adapts to RAAS safety, contrast risk, radiation risk, and more
Live interactive demo
Try a real clinical decision.
Work through an actual CKD + hypertension management decision. Evidence, knowledge graph cues, and coaching are live.
HTN + T2D + CKD Stage 4 — Rising Potassium
Interactive micro-case · Evidence-backed
Live demo
Clinical vignette
- 57-year-old with uncontrolled hypertension presenting to outpatient clinic
- PMH: T2D, obesity, CKD stage 4 (eGFR 22 mL/min/1.73m²), UACR 780 mg/g (severely elevated)
- Meds: amlodipine 5 mg daily; losartan 50 mg daily
- BP 164/92 mmHg (repeat confirms); asymptomatic. K⁺ 5.5 mEq/L (up from 4.9 last visit), creatinine stable
What is your next clinical move?
Clinical Rounds Coach
Evidence + knowledge graph feedback
Outpatient intensification can preserve kidney-protective therapy without unsafe K⁺ escalation. SGLT2i reduces cardiovascular and CKD progression risk at eGFR ≥20.
Post-case analytics
Know where reasoning broke down.
Each session closes with a structured report: reasoning quality scored against the expert pathway, stewardship patterns across similar cases, and evidence coverage — where you acted without guideline anchoring, and what should have changed the plan.
- Reasoning quality vs. expert path (not just right/wrong)
- Stewardship score: over/under-testing, unnecessary imaging flagged
- Evidence coverage: unanchored decisions get coaching notes
- Trend tracking across multiple sessions for faculty
Post-case report · Session summary
Reasoning qualitySolid
Stewardship scoreHigh-value
Evidence coverage90%
Stewardship insight: You avoided low-value imaging by using a pregnancy-adapted rule-out pathway before CT in a stable presentation.
Your report highlights not just what you did, but which patient constraints and evidence links should have shaped the path.
For your level