🛡 Insurance

Real-time pre-auth, fraud detection & claims automation

Replace 72-hour manual pre-authorisation with 2.4-hour FHIR-structured API calls. UHIN connects your claims workflow directly to ABDM-verified hospital records — eliminating fraud, accelerating approvals, and transforming PM-JAY processing.

NHCX Integration ABDM HIU Real-time Pre-auth Fraud Detection PM-JAY

Key Features

Everything your team needs in one place

⚡

Real-time Pre-auth API

Submit a FHIR Claim resource and receive a structured decision in 2.4 hours. Clinical evidence auto-assembled from the patient's ABDM records — no fax, no phone calls, no paper.

IRDAI compliant

🔍

11-Signal Fraud Engine

Duplicate claims, procedure upcoding, phantom procedures, short admissions, drug billing mismatches, network outlier detection — each cross-referenced against live FHIR records.

ML-powered

📊

Risk-based Underwriting

5-year ABDM patient history enables accurate risk banding: medication adherence scores, chronic condition progression, preventive care completion, and hospitalisation risk ML model.

ABDM consent

💳

PM-JAY Integration

Ayushman Bharat beneficiary verification, package-based claim submission, empanelled hospital eligibility checks, and state health department reporting.

PM-JAY

📋

Live Claims Dashboard

Claims queue with ML risk scores, auto-approve thresholds (score <30), escalation workflows, settlement timelines, and full audit trail on every decision.

Live data

🧮

Portfolio Analytics

Population risk distribution, disease-wise claims incidence, seasonal surge forecasting, and network hospital performance benchmarking — all de-identified.

De-identified

Pre-Auth Flow

From 72 hours to 2.4 hours

UHIN auto-assembles the clinical evidence package from the patient's ABDM FHIR records and submits it alongside the pre-auth request. Your adjudicators review structured data, not photocopied documents.

Claim submitted

Hospital submits FHIR Claim resource via UHIN pre-auth API with procedure codes, diagnosis, and facility details.

Evidence assembled

UHIN pulls FHIR records — prior diagnoses, medications, lab results, prior claims — via ABDM consent artefact.

ML risk score

Model scores: ICD-10 vs procedure match, billing vs peer benchmark, medication cross-check, admission pattern.

Structured decision

Auto-approve (<score 30) or escalation with specific reason codes returned in <2.4 hours on average.

Fraud Detection Library

11 fraud signal types

✓

Duplicate claim detection — same procedure at multiple facilities within claim period

✓

Procedure upcoding flag — procedure billed vs FHIR Encounter content mismatch

✓

Phantom procedure detection — no ABDM care context registered for the claimed event

✓

Short admission pattern — IPD billing with sub-clinical discharge timeline

✓

Drug billing vs Rx cross-check — dispensed medication vs FHIR MedicationRequest mismatch

✓

Network provider outlier — facility billing rate vs peer deviation >2σ