Payment Variance Analysis

Detect and resolve discrepancies between expected and actual payments.

Overview

Payment variances occur when payers pay different amounts than expected. AI helps identify underpayments, overpayments, and patterns indicating contract compliance issues.

Tools Used

Tool	Purpose
create_payment_variance	Track underpayments/overpayments
update_payment_variance	Update resolution, link to appeals
list_payment_variances	Query variances with filters
get_payment_variance_analytics	Payer patterns and contract compliance

Example: Seeded Payment Variance

The database contains 1 seeded payment variance (Medicare underpayment):

Variance	Claim	Payer	Expected	Actual	Variance	Type
b50e8400...001	CLM-2024-001890	Medicare	$500	$425	-$75 (-15%)	underpayment

Check Payment Variances

User prompt:

Do we have any underpayments we should look at?

What happens:

The list_payment_variances tool retrieves variances:

{
  "variances": [
    {
      "id": "b50e8400-e29b-41d4-a716-446655440001",
      "claim_id": "650e8400-e29b-41d4-a716-446655440006",
      "payer_id": "550e8400-e29b-41d4-a716-446655440003",
      "expected_amount": 500.0,
      "actual_amount": 425.0,
      "variance_amount": -75.0,
      "variance_percentage": -15.0,
      "variance_type": "underpayment",
      "variance_reason": "contract_adjustment",
      "payment_date": "2024-02-25",
      "reason_notes": "Payer applied incorrect fee schedule. Expected payment based on 2024 Medicare rates, but 2023 rates were applied.",
      "requires_appeal": true,
      "resolved": false
    }
  ],
  "total": 1,
  "total_variance": -75.0
}

Expected Response:

Found 1 underpayment requiring attention:

• CLM-2024-001890 (Medicare)
  • Expected: $500, Received: $425
  • Variance: -$75 (-15%)
  • Reason: Incorrect 2023 fee schedule applied instead of 2024 rates
  • Status: Requires appeal

Recommendation: Appeal this variance - it's a contract compliance issue with high success probability.

Update Variance Resolution

User prompt:

We appealed the variance and got an additional $50. Mark it resolved.

What happens:

The update_payment_variance tool updates the record:

{
  "variance_id": "b50e8400-e29b-41d4-a716-446655440001",
  "resolved": true,
  "resolution_date": "2024-03-15",
  "resolution_notes": "Appeal successful. Additional payment of $50 received. Remaining $25 variance accepted as contract adjustment."
}

Variance Analytics

User prompt:

What are our payment variance patterns?

The get_payment_variance_analytics tool returns:

{
  "total_variances": 1,
  "total_variance_amount": -75.0,
  "underpayment_total": 75.0,
  "overpayment_total": 0,
  "by_type": {
    "underpayment": { "count": 1, "amount": -75.0 }
  },
  "by_reason": {
    "contract_adjustment": { "count": 1, "amount": -75.0 }
  },
  "by_payer": {
    "Medicare": { "count": 1, "amount": -75.0, "avg_variance_pct": -15.0 }
  },
  "unresolved_count": 1,
  "insights": [
    "1 unresolved variance totaling $75",
    "Medicare shows -15% average variance - verify fee schedule",
    "100% of variances are contract_adjustment - review contract terms"
  ]
}

Workflow Diagram

Variance Types (Auto-Calculated)

Type	Condition	Action
underpayment	Actual less than Expected	Review for appeal
overpayment	Actual greater than Expected	Return or adjust
expected	Within 1 cent	No action needed

Variance Reasons

Reason	Description	Typical Recovery
contract_adjustment	Rate differs from contract	High (95%)
bundling	Services bundled/downcoded	Medium (70%)
non_covered_service	Partial denial	Low (30%)
incorrect_coding	Downcoded by payer	High (85%)
coordination_of_benefits	COB adjustment	Varies

Key Analytics

• Payer patterns - Identify systematic underpayers
• Variance percentages - Track for contract negotiation
• Appeal linkage - Connect variances to appeals
• Unresolved monitoring - Prioritize open variances

Next Steps

• Appeals Management - Recover underpayments
• Cash Leakage - Identify variance patterns
• Denial Triage - Handle related denials