How to Code for MEAT Criteria Without Missing Revenue or Failing Audits

Risk adjustment coders face an impossible task every day: find evidence in a clinical note that the provider may not have documented clearly, then make a coding decision that could cost your organization thousands of dollars if you get it wrong.

Welcome to MEAT criteria coding, where the stakes are high and the documentation is often unclear.

MEAT stands for Monitor, Evaluate, Assess, and Treat. For every HCC you code, you need to find at least one of these four elements in the provider’s note. No MEAT evidence means no HCC code, even if you know the patient has the condition. CMS doesn’t care what you know. They care what you can prove from the documentation.

Here’s how to code for MEAT criteria without leaving money on the table or exposing your organization to audit risk.

The Coder’s Dilemma

You’re reviewing a note. The provider lists “CHF” in the assessment section. Nothing else. No mention of symptoms, no exam findings related to heart failure, no medication adjustments, no patient education. Just the diagnosis on the problem list.

Do you code it?

If you do, you’re betting that CMS will accept a diagnosis code listed in the assessment as sufficient evidence. They won’t. If you don’t code it, you’re leaving legitimate revenue unclaimed if the patient truly has active heart failure that the provider simply didn’t document well.

This is the daily reality of MEAT criteria coding. You’re making judgment calls based on incomplete information, and both overcoding and undercoding have consequences.

What Qualifies as MEAT Evidence

Before you can code correctly, you need to recognize MEAT when you see it. Here’s what actually counts:

Monitoring includes any tracking of the condition’s status. Lab values (“A1C 7.8”), vital signs tied to the condition (“BP 145/92 despite current antihypertensive regimen”), patient-reported symptoms (“Patient reports increased shortness of breath this week”), or physical exam findings (“Bilateral lower extremity edema, 2+ pitting”) all demonstrate monitoring.

Evaluation means the provider is thinking about the condition and documenting their clinical judgment. Phrases like “condition is well-controlled,” “symptoms are worsening,” “patient is compensated today,” or “disease appears stable” show evaluation. The provider is assessing the current state of the condition.

Assessment appears in the assessment and plan section but needs substance. “Continue current management” tied to a specific condition shows assessment. “Patient’s diabetes is at goal” shows assessment. Just listing the condition name without context doesn’t.

Treatment is usually the easiest to identify. Medication orders, prescription renewals, dosage adjustments, therapy referrals, diagnostic test orders, patient education, or lifestyle counseling all count as treatment. “Renewed metformin,” “ordered echocardiogram,” or “counseled on low sodium diet” are clear treatment documentation.

The Coding Workflow

Start by reading the entire note. Don’t jump straight to the assessment section. MEAT evidence often appears in the history of present illness, review of systems, physical exam, or plan sections rather than the problem list.

For each chronic condition listed in the assessment, scan the note for corresponding MEAT criteria. Use a systematic approach. Check the HPI for monitoring or evaluation language. Review the exam for objective findings. Look at the plan for treatment documentation.

When you find MEAT evidence, link it clearly to the specific condition. If the note says “A1C 7.8, continue current diabetes medications,” you have monitoring (the A1C) and treatment (medication continuation) for diabetes. Code it with confidence.

When you don’t find clear MEAT evidence, you have three choices: don’t code it, query the provider, or accept the audit risk. Your decision depends on your organization’s risk tolerance and query protocols.

The Query Decision

Queries slow down your workflow. Nobody wants to send a query for every borderline documentation issue. But queries also protect your organization from audit risk and recover legitimate revenue.

Query when the clinical context strongly suggests the condition was addressed but the documentation is vague. If a patient with known CHF is on Lasix, has recent weight monitoring, and comes in for dyspnea, the provider almost certainly evaluated the heart failure. If the note doesn’t explicitly say so, query.

Don’t query for conditions where there’s no clinical context. If the note is about an ankle sprain and diabetes is only mentioned on the problem list with no related documentation, querying won’t help. The provider didn’t address it during that encounter.

Use queries as education opportunities. Don’t just ask “Did you address this condition?” Explain what documentation you need: “To support coding for CHF, I need evidence of monitoring (such as weight, symptoms, exam findings), evaluation (your clinical judgment about the condition), or treatment (medication management or counseling). Can you clarify whether you addressed the CHF during this visit?”

Common Coding Mistakes

The biggest coding error is accepting a problem list as sufficient evidence. Just because a condition appears in the assessment section doesn’t mean it meets MEAT criteria. You need documentation in the body of the note.

Another common mistake is coding based on treatment from a previous visit. “Continue current medications” without specifying which medications for which conditions doesn’t support HCC coding. You need explicit linkage between the treatment and the condition.

Coders also sometimes miss embedded MEAT evidence. A note might not mention diabetes in the assessment but document “checked patient’s blood sugar today, reading is 142, acceptable for this patient.” That’s monitoring for diabetes. The HCC can be coded even if the provider didn’t list it in their assessment.

Conversely, some coders code too aggressively on thin documentation. “Patient has history of stroke” without current symptoms, current medications, or current limitations doesn’t support an HCC. Historical conditions need present relevance to be coded.

Condition-Specific Challenges

Certain conditions create recurring coding dilemmas. Chronic kidney disease requires a current GFR or creatinine level plus stage documentation. A note that mentions “CKD” without recent lab values or staging doesn’t meet MEAT criteria for a specific CKD HCC.

Cancer coding requires evidence of current treatment or active disease. A patient who completed treatment three years ago and is disease-free doesn’t support a cancer HCC unless the note specifically documents continued medication, monitoring, or effects.

Heart failure needs evidence of current symptom management or ongoing treatment. “Patient had heart failure in 2020” isn’t enough. “Patient continues on Lasix 40mg daily for heart failure management” supports the HCC.

Mental health conditions require documented symptoms or treatment during the encounter. A patient on standing psychiatric medications needs evidence that the provider addressed the mental health condition, not just renewed a prescription without evaluation.

Building Coding Confidence

Good MEAT criteria coding requires pattern recognition developed through experience. The more notes you review, the faster you identify MEAT evidence and recognize when it’s missing.

Create personal coding references for conditions you code frequently. Document what MEAT evidence you typically see for diabetes, CHF, COPD, and other common chronic conditions. This speeds up your review process and ensures consistency.

When you’re uncertain, consult with other coders or your coding supervisor. “I see this phrase used often, does it count as MEAT evidence?” Building team consensus on borderline documentation creates consistency across your coding operation.

Track your audit results. When charts get audited and HCCs are rejected, understand why. Was the MEAT criteria truly missing, or did you miss evidence that was there? This feedback loop improves your coding accuracy over time.

The Bottom Line

MEAT criteria coding isn’t about finding loopholes or making aggressive assumptions. It’s about accurately identifying documented evidence that supports the codes you assign. When the evidence is there, code it. When it’s not, don’t guess.

Your coding accuracy directly impacts your organization’s revenue integrity and audit defensibility. Code too conservatively and you leave legitimate money on the table. Code too aggressively and you create compliance exposure. The balance point is coding exactly what the documentation supports, nothing more and nothing less.

That requires strong clinical knowledge, attention to detail, and the judgment to know when documentation meets the standard and when it falls short. It’s not easy work, but it’s the foundation of defensible risk adjustment coding.