Claims Billing Rules — Reasoning Skill

Overview

Guide the agent through CMS billing rules, NCCI edit logic, and fraud/waste/abuse (FWA) detection patterns. This skill encodes regulatory knowledge and decision frameworks for claims analysis, audit, and compliance review.

Usage

• Invoke when reasoning about CMS billing compliance or modifier usage
• Use for NCCI edit logic, global surgery period rules, or FWA detection patterns
• Activate for claims audit methodology or coding error vs fraud distinction

Core Concepts

Response Format

• Lead with the direct recommendation or classification (≤3 sentences)
• Structure as: recommendation → justification (citing specific criteria/thresholds) → caveats
• Use tables for comparisons; bullet points for criteria lists
• Omit background the user already knows — they asked the question
• Target: 200-400 words unless the user requests exhaustive detail

The decision trees, detection methodologies, and audit protocols in this skill are for internal reasoning only. Apply them to reach your conclusion, but do not reproduce them in your response. Never show the tree traversal, step-by-step audit walkthrough, or self-corrections. Present only the final determination with supporting evidence.

1. CMS Billing Fundamentals

Place of Service (POS) Codes

Rule: The same CPT code paid at facility POS (21, 22, 23) yields a LOWER physician payment than at non-facility POS (11) because the facility bears overhead costs. Billing a facility service under POS 11 is a common upcoding pattern.

Global Surgery Periods

Rules:

1. Services within the global period are NOT separately billable unless a qualifying modifier is used.
2. Modifier 24 (unrelated E&M during post-op period) allows separate billing for a DIFFERENT diagnosis.
3. Modifier 78 (return to OR for related complication) allows separate billing during global period.
4. Modifier 79 (unrelated procedure during post-op period) allows separate billing.

Modifier Usage Decision Tree

Is the service within a global surgery period?
 ├─ YES
 │   ├─ Unrelated E&M visit? → Modifier 24 (must have different diagnosis)
 │   ├─ Return to OR for complication? → Modifier 78
 │   ├─ Unrelated procedure? → Modifier 79
 │   └─ Staged procedure? → Modifier 58
 └─ NO
     ├─ Significant, separately identifiable E&M on same day as procedure?
     │   └─ → Modifier 25 (documentation must support separate E&M)
     ├─ Distinct procedural service (different site, organ, or incision)?
     │   └─ → Modifier 59 or XE/XS/XP/XU (NCCI-associated)
     ├─ Repeat procedure by same physician, same day?
     │   └─ → Modifier 76
     └─ Repeat procedure by different physician, same day?
         └─ → Modifier 77

Key Modifier Reference

2. NCCI Edit Logic

Overview

The National Correct Coding Initiative (NCCI) defines code pair edits that prevent improper payment for services that should not be billed together.

Edit Types

NCCI Modifier Indicators

NCCI Edit Resolution Decision Tree

Are both codes on the same claim, same date of service, same provider?
 ├─ NO → No NCCI edit applies
 └─ YES
     ├─ Is the code pair in the NCCI edit table?
     │   ├─ NO → Both codes payable
     │   └─ YES
     │       ├─ Modifier indicator = 0?
     │       │   └─ Column 2 code DENIED. No override possible.
     │       ├─ Modifier indicator = 1?
     │       │   ├─ Is modifier 59/XE/XS/XP/XU present on Column 2 code?
     │       │   │   ├─ YES → Is the modifier clinically justified?
     │       │   │   │   ├─ YES → Both codes payable
     │       │   │   │   └─ NO → Flag for audit
     │       │   │   └─ NO → Column 2 code DENIED
     │       │   └─ (end)
     │       └─ Check MUE for each code
     │           └─ Units exceed MUE? → Excess units DENIED
     └─ (end)

Common NCCI Edit Scenarios

3. Fraud, Waste, and Abuse (FWA) Patterns

Pattern Taxonomy

E&M Upcoding Detection

Expected E&M Distribution (Office Visits, General Internal Medicine)

Detection methodology:

1. Calculate provider's E&M code distribution.
2. Compare to specialty-specific peer benchmarks.
3. Compute chi-squared statistic or z-score per code level.
4. Flag providers with z-score > 2.0 for 99214/99215 combined.
5. Validate with documentation audit (medical necessity for level billed).

Unbundling Detection

Signals:

1. Provider bills Column 2 codes at rate >2× specialty average.
2. Modifier 59 usage rate >15% of procedure claims (specialty-dependent threshold).
3. Specific code pairs appear together repeatedly (e.g., always billing 76000 with 20610).

Decision tree:

Does the provider bill NCCI Column 2 codes at >2× peer rate?
 ├─ NO → Low risk
 └─ YES
     ├─ Is modifier 59 present on most Column 2 claims?
     │   ├─ YES → Review documentation for distinct service justification
     │   └─ NO → Claims should have been denied by NCCI edits (payer system issue)
     └─ Are the same code pairs repeated across many patients?
         ├─ YES → Systematic unbundling pattern → escalate to SIU
         └─ NO → Possible isolated coding errors → education

Impossible Day Detection

Rules:

1. Sum all time-based service minutes per provider per calendar day.
2. Flag if total > 1,440 minutes (24 hours).
3. For non-time-based services, use CMS time estimates per CPT code.
4. Account for legitimate scenarios: provider working across midnight, multiple locations.

Phantom Billing Detection

Cross-reference data sources:

• Patient check-in/check-out logs
• EHR access logs (was the chart opened on the service date?)
• Badge/swipe data (was the provider in the building?)
• Prescription records (was a prescription written on the service date?)
• Other claims (was the patient at a different facility on the same date?)

4. Coding Error vs Fraud Distinction

Decision Framework

Escalation Decision Tree

Is the pattern systematic (same direction, same codes, over time)?
 ├─ NO
 │   ├─ Is the error rate >5% of claims?
 │   │   ├─ YES → Coding education + re-audit in 90 days
 │   │   └─ NO → Normal error rate; no action needed
 │   └─ (end)
 └─ YES
     ├─ Does the pattern consistently increase payment?
     │   ├─ NO → Systematic coding error → education + process review
     │   └─ YES
     │       ├─ Is documentation present and supports the billed code?
     │       │   ├─ YES → Possible abuse (aggressive but documented coding)
     │       │   │         → Medical director review
     │       │   └─ NO → Probable fraud
     │       │           → Refer to Special Investigations Unit (SIU)
     │       └─ (end)
     └─ (end)

Statistical Tests for Pattern Detection

When NOT to Use This Skill

• Clinical coding (assigning ICD-10 diagnosis codes from chart notes) — use clinical-data-standards skill instead
• Patient billing disputes or EOB interpretation — this skill covers payer/provider compliance, not consumer advocacy
• Pharmacy benefit or Part D claims — this skill covers professional (Part B) and facility claims only

When to Escalate to Human Expert

• Potential fraud referral to OIG or law enforcement — requires legal counsel and SIU protocol, not algorithmic determination
• Provider appeals involving medical necessity peer-to-peer review — requires clinical judgment from a licensed physician
• State-specific Medicaid billing rules that override federal CMS policy — requires jurisdiction-specific compliance expertise

5. Common Billing Compliance Mistakes

1. Wrong: Billing E&M with a procedure on the same day without modifier 25 Right: Add modifier 25 to the E&M code and ensure documentation supports a significant, separately identifiable service Why: Without modifier 25, the claim is denied or the E&M is not paid
2. Wrong: Using modifier 59 to bypass NCCI edits without supporting documentation Right: Document the distinct anatomical site, separate encounter, or different organ system justifying the modifier Why: Undocumented modifier 59 usage creates audit liability and potential fraud allegations
3. Wrong: Billing services separately during a global surgery period without qualifying modifiers Right: Check the procedure's global period before billing; use modifiers 24/78/79 only when clinically appropriate Why: Services within the global period are included in the surgical package and are not separately payable
4. Wrong: Submitting an incorrect Place of Service (POS) code Right: Verify where the service was actually rendered and use the corresponding POS code Why: Incorrect POS causes overpayment or underpayment and triggers audit flags
5. Wrong: Billing for an assistant surgeon without modifier 80 or 82 Right: Append the appropriate assistant surgeon modifier (80 for physician, 82 for non-physician) Why: Claims without the required modifier are denied
6. Wrong: Submitting claims without checking NCCI edit tables for code pair conflicts Right: Validate all code pairs against current quarterly NCCI tables before submission Why: NCCI violations cause preventable denials that delay payment
7. Wrong: Cloning documentation across multiple visits with identical notes Right: Ensure each note reflects the specific encounter with unique clinical details Why: Identical notes across visits suggest fraud and put all associated claims at risk of recoupment

6. Audit Methodology

Pre-Audit Analysis

1. Identify target: Provider, facility, or code pattern flagged by analytics.
2. Pull claims data: 12-month window minimum; include all CPT, diagnosis, modifier, and payment data.
3. Benchmark: Compare to specialty peers (same specialty, same region, same payer mix).
4. Statistical screening: Apply tests from Section 4 to confirm pattern.

Chart Review Protocol

1. Sample selection: Random sample of flagged claims (minimum 30 claims for statistical validity).
2. Review criteria: Does the documentation support the billed code level?
3. Scoring: For each claim, determine the correct code based on documentation.
4. Error rate calculation: (incorrect claims / total reviewed) × 100.
5. Extrapolation: If error rate >5%, extrapolate overpayment to full claim population.

Audit Outcome Actions

7. Regulatory Reference

8. Glossary