Network & Attribution Management

Intelligent Referral Routing Agent

ai

Detects referral orders, matches patients to optimal in-network specialists, checks prior auth, transmits referrals, tracks completion, and flags network leakage.

Referral CoordinatorPrimary Care PhysicianPopulation Health
32%
Reduction in network leakage
80%
Referrals completed vs. 55% baseline
$240K–$900K
Annual retained revenue
Use Case 1

Closed-Loop Referral Management

Trigger: Provider places referral order

1.Detect referral order in EHR
2.Match to optimal in-network specialist on clinical, geographic, quality, and patient-preference factors
3.Verify insurance network and prior auth requirements
4.Transmit referral to specialist
5.Send scheduling link to patient
6.Track scheduling and completion
7.Flag if patient goes out-of-network
8.Return specialist note to PCP

Referrals route in-network, patients schedule easily, PCPs see completion—leakage drops measurably.

Use Case 2

Network Leakage Analytics

Trigger: Monthly network utilization review

1.Aggregate referral volume in- vs. out-of-network
2.Classify leakage by root cause
3.Identify capacity, coverage, and patient-preference drivers
4.Recommend network additions or interventions
5.Quantify retained and leaked revenue

Network strategy driven by data, not anecdote.

Integrations

EHR Referral Orders (Epic, Cerner)
Provider Directory & Network Data
Payer Network Verification
Prior Auth Database
Patient Messaging
HIE / Specialist Notes Returns

Industries

healthcare

See It In Action

A sample conversation with this agent

Dr. Chen just ordered a cardiology referral for Maria Santos.
Referral detected. Matching Maria to optimal in-network cardiologists using: clinical specialty (general cardiology, CHF focus given her history), geographic proximity to her ZIP, insurance network participation, quality scores, wait time, language capability (she prefers Spanish), and continuity with prior referrals. Top match: Dr. Reyes at Penn Cardiology — 3.2 miles from her home, Spanish-speaking, 11-day first-available, in-network for her plan, 4.7 quality score. Prior auth not required for initial consult. I'll transmit the referral and send Maria a scheduling link. Updating Dr. Chen's EHR with referral status.
Why did we lose 8 cardiology referrals to out-of-network last month?
Analyzing the 8 leakage events: 4 were scheduled by patients directly through urgent symptoms before referral workflow completed (ED cardiology consults); 2 had prior relationships with out-of-network cardiologists the patient chose to continue; 1 was routed in-network but the specialist's wait time was 6+ weeks so patient self-referred elsewhere; 1 was a geographic fit issue—no in-network cardiologist within 15 miles of patient's home. Recommendation: add a shorter-wait cardiology group to your network and set up a warm handoff for ED-discovered cardiology needs.
Show me our network utilization for Q2.
Q2 referral volume: 2,341. In-network: 1,896 (81%). Out-of-network: 445 (19%). Of the OON: 287 were urgent/emergent (acceptable leakage), 118 were patient-preference driven, 40 were capacity-driven leakage (in-network wait times too long). Top leaking specialties: Orthopedics (wait time), Dermatology (panel coverage gap), Behavioral Health (capacity). Retained revenue from 1,896 in-network referrals: estimated $2.4M. Actionable OON loss: $540K addressable through network additions and wait-time management.
Intelligent Referral Routing Agent | AI Workflow Library | actAVA