UC17: Intelligent Patient Triage (Urgent Care/ED)

The Problem

Emergency departments and urgent care centres are overwhelmed by non-urgent walk-ins, creating bottlenecks that delay critical patients. Manual triage by nursing staff is inconsistent; red-flag symptoms (chest pain, stroke signs, severe trauma) are sometimes missed during peak hours. Waiting times extend 4 to 6 hours even for low-acuity patients, driving poor satisfaction and LWBS (left without being seen) rates.

What the Agent Does

Conducts conversational patient interviews on intake tablets, asking about chief complaint, symptom onset, and red-flag symptoms (chest pain, shortness of breath, neurological changes)
Applies Emergency Severity Index (ESI) protocol logic to categorise patients into acuity levels (ESI-1 to ESI-5)
Immediately flags ESI-1 and ESI-2 patients for nurse/physician review; routes ESI-4 and ESI-5 to fast-track or self-care pathways
Estimates realistic wait time based on current ED census and acuity mix
Identifies patients requiring immediate interventions (e.g., cardiac monitoring, stroke protocol activation)

Data Requirements

Data Sources:

Epic ADT / Meditech ED Module , Real-time ED census, acuity-weighted bed status, current wait times
Protocol Library , ESI triage guidelines, facility-specific red-flag rules, STEMI/stroke alert criteria
Patient Chief Complaint , Real-time user input via tablet interface
Historical Acuity Data , Past triage outcomes to validate ESI assignment accuracy

Data Classification:

Patient input & acuity data: Sensitive, Unstructured (conversational)
ED census: Internal, Structured
Protocol rules: Internal, Structured

Data Quality Requirements:

ED census data real-time (updated every 5 minutes during peak hours)
ESI protocol rules validated annually against American College of Emergency Physicians (ACEP) standards
Red-flag symptom library reviewed quarterly with ED physician staff

Integration Complexity: Medium

Real-time ADT feed from Epic/Meditech (HL7 v2 or FHIR-based)
Conversational UI must handle diverse, non-medical patient language
Integration with nurse alerting systems (Epic Hyperspace or SMS)
Validation against ESI guidelines requires clinical oversight

Score Breakdown

Criterion	Weight	Score (1-5)	Weighted
Time Recaptured	15%	4	0.60
Error Reduction	10%	4	0.40
Cost Avoidance	10%	3	0.30
Strategic Leverage	5%	3	0.15
Data Availability	15%	4	0.60
Process Clarity	15%	4	0.60
Ease of Implementation	10%	3	0.30
Fallback Available	10%	3	0.30
Audience (Int/Ext)	10%	3	0.30
Composite	100%		3.60

Why It Scores Well

Intelligent triage directly improves patient safety (red-flag symptoms caught earlier), reduces LWBS rates, and improves ED throughput. ESI triage is an evidence-based, nationally standardised protocol; the agent applies deterministic logic to a well-defined problem. The external audience (patients) benefits immediately from shorter wait estimates and faster critical care.

Regulatory Alignment

EMTALA (Emergency Medical Treatment & Labor Act): All patients must be screened and stabilised regardless of ability to pay; AI triage must not be used to deny care or circumvent EMTALA screening
State Scope-of-Practice: Some states restrict AI-driven medical decision-making; triage must be validated by licensed nurse/physician
ADA (Americans with Disabilities Act): Tablet-based interface must be accessible (audio, large text, alternative input)
Joint Commission: ED operations must maintain clinically validated triage protocols

Sprint Factory Fit

Sprint 0 (2 weeks) + 4 build sprints (8 weeks)

Triage is event-driven: every patient arrival triggers triage logic. The initial 2-week sprint focuses on ESI protocol implementation and tablet UI; subsequent sprints add real-time ED census integration, red-flag symptom detection, and outcome tracking. This use case is lower-complexity than denial management but requires clinical validation.

Comparable Implementations

Deploy This Use Case with the Sprint Factory

From zero to a governed, production agent in 6 weeks.

Sprint Factory Schedule a Briefing

Related Use Cases

Real-Time Clinical Operations

Bed Management "Traffic Controller"

Score: 4.15/5.0

Real-Time Clinical Operations

Supply Chain & Surgical Kit Inventory

Score: 3.7/5.0

Real-Time Clinical Operations

After-Hours Triage (Urgent Care)

Score: 3.55/5.0

Governance Risks to Consider

Before deploying this use case, review these agentic AI risks from the Corvair Risk Catalogue. Each is scored on the DAMAGE framework and mapped to regulatory expectations.