How do I know if I am undercoding?

Run your typical notes through an AI gap analysis tool. If the AI consistently identifies documentation supporting a higher code than what you would have billed, you are undercoding. Common signs include defaulting to 99213 for most established patients, never using time-based coding, and ignoring independent interpretation of diagnostic tests.

Does undercoding protect me from audits?

No. An atypically low E/M distribution can trigger an audit just as easily as a high one — it signals that documentation practices are inconsistent with the complexity of care being delivered. The real audit protection is coding accurately with documented rationale.

What is the revenue difference between 99213 and 99214?

The reimbursement gap is approximately $45-$65 per claim based on 2024 Medicare national averages. If you undercode just 5 patients per day, that equals roughly $68,750 per year in lost revenue for a single physician.

What is an HCC coding opportunity and why does gap analysis flag it?

Hierarchical Condition Categories (HCCs) are diagnoses that affect Risk Adjustment Factor scores, which determine payments in Medicare Advantage and value-based contracts. AI gap analysis flags documented conditions that qualify as HCC codes but were not captured as codeable diagnoses — revenue you earned but did not bill.

Stop Undercoding: How AI Catches Revenue You're Missing

If you’re like most physicians, you’re leaving money on the table every single day — not from fraud, not from billing errors, but from undercoding.

Undercoding is the silent revenue leak in medical practices. And it’s far more common than you might think.

What Is Undercoding, Exactly?

Undercoding means billing a lower E/M code than your documentation actually supports. You did the work. You saw the complexity. But something — time pressure, habit, fear of audit, incomplete documentation — caused you to select a lower-level code.

Common undercoding patterns:

Defaulting to 99213 when 99214 is warranted. The 99213 is the comfort code. Physicians reach for it reflexively, especially for established patients. But if you reviewed external records, managed multiple chronic conditions, or spent 25+ minutes, 99214 (or even 99215) may be fully supported.

Ignoring time-based coding. The AMA 2021 guidelines allow you to code by total time OR MDM — whichever yields the higher code. Physicians who only think about MDM miss half their optimization opportunities.

Missing independent interpretation. Ordering an EKG is not the same as independently interpreting it. The latter qualifies as data review for MDM purposes. Physicians routinely fail to document the distinction.

Omitting HCC-relevant diagnoses. Hierarchical Condition Categories affect RAF scores, care coordination payments, and value-based contract performance. Many physicians document conditions clinically but don’t capture them as codeable diagnoses.

The Revenue Impact Is Real

The reimbursement difference between 99213 and 99214 is approximately $45–$65 (2024 Medicare national averages). If you see 20 patients/day and undercode just 5 of them:

5 missed upgrades × $55 avg = $275/day
250 working days = $68,750/year

In a group practice with 5 physicians, that’s $343,750 in annual revenue evaporation — from undercoding alone.

How AI Gap Analysis Works

AI medical coding tools don’t just assign a code — they analyze your note for coding gaps:

MDM extraction: The AI parses your note and identifies every element that contributes to medical decision making.
Dual-code comparison: It calculates both MDM-based and time-based codes, then flags which is higher and why.
Gap flagging: It identifies documentation that’s present but uncoded — e.g., you mention reviewing old records but didn’t explicitly note independent interpretation.
HCC opportunities: For applicable diagnoses, it surfaces HCC coding opportunities with MEAT criteria.
Audit trail generation: You get the full MDM rationale pre-written, which both justifies your code and protects you in an audit.

The Audit Fear Paradox

Here’s an irony most physicians don’t realize: undercoding doesn’t protect you from audits. An atypically low E/M distribution can be as much an audit trigger as an atypically high one — it suggests documentation practices are inconsistent with care complexity.

The protection isn’t coding lower. It’s coding accurately with documented rationale. That’s exactly what AI gap analysis provides.

What “Properly Coded” Revenue Looks Like

Practices that implement AI-assisted coding consistently report:

15–25% increase in average E/M code level within 90 days
Reduction in claim denials (better-documented claims hold up to payer scrutiny)
Faster billing cycles (AI-generated audit trails reduce coder review time)
Improved RAF scores for value-based contracts

This isn’t upcoding. It’s accurate coding — billing for what you actually did.

See Your Own Gap Analysis

The only way to know how much you’re undercoding is to run your own notes through an AI analysis engine. CodeItRight will show you side-by-side: what you would have coded vs. what your documentation actually supports.

For most physicians, the gap analysis is the wake-up call.