99213 vs 99214: How to Know When Your Documentation Supports the Higher Code

If there is one question that defines modern E/M coding, it is this: “Should this visit be a 99213 or a 99214?”

It is the most searched medical coding question on the internet, and for good reason. The reimbursement difference between these two codes is $55–$65 per visit under 2024 National Medicare averages. Multiply that across 20 patients a day, 250 working days a year, and even a modest shift in coding accuracy can mean $50,000–$80,000 in annual revenue.

The problem is not that physicians lack clinical skill. It is that the documentation habits separating a 99213 from a 99214 are subtle — and under daily volume pressure, most providers default to the safer, lower code.

The Core Difference: MDM Complexity

Under the AMA 2021 guidelines, both 99213 and 99214 are established patient office/outpatient visits. The difference comes down to Medical Decision Making (MDM) complexity:

• 99213 = Low complexity MDM
• 99214 = Moderate complexity MDM

MDM is scored across three elements. You need to meet the threshold in at least 2 of the 3 elements to qualify for a given level:

1. Number and Complexity of Problems Addressed
2. Amount and/or Complexity of Data Reviewed and Analyzed
3. Risk of Complications, Morbidity, or Mortality

Let’s walk through each element and see exactly where the line falls.

Element 1: Number and Complexity of Problems

99213 (Low): Two or more self-limited or minor problems (e.g., acute sinusitis, mild rash). OR one chronic illness that is stable and under adequate control (e.g., well-controlled hypertension at target blood pressure).

99214 (Moderate): One or more chronic illnesses with mild exacerbation, progression, or side effects of treatment (e.g., diabetes with A1C rising from 7.1 to 7.8). OR two or more stable chronic illnesses (e.g., hypertension + type 2 diabetes + hyperlipidemia, all stable but being actively managed). OR one undiagnosed new problem with uncertain prognosis (e.g., unexplained weight loss requiring diagnostic workup).

The critical distinction: a single stable chronic condition is low complexity. The moment that condition worsens, or you add a second chronic condition to the encounter, you cross into moderate territory.

Element 2: Data Reviewed and Analyzed

99213 (Limited): Reviewing or ordering limited data — for example, ordering a basic lab panel or reviewing a single test result. One category of data reviewed or ordered.

99214 (Moderate): Requires meeting any one of these data thresholds:

• Ordering and reviewing a test (labs, imaging) from two categories of data — e.g., ordering labs AND reviewing prior imaging
• Independent interpretation of a test you did not order (e.g., reviewing and independently interpreting an EKG or X-ray performed by another provider)
• Review of external records from a different facility or provider with a summarized documentation of that review
• Discussion of management or test interpretation with an external physician (not within your own group)

This is where most physicians leave money on the table. You may routinely review outside records, independently interpret imaging, or coordinate with specialists — but if you do not explicitly document those activities, they do not count toward your MDM data score.

Element 3: Risk (CMS Table of Risk)

99213 (Low Risk): Prescription drug management of a non-controlled substance (e.g., refilling lisinopril). Minor surgery without identified risk factors. Physical therapy or occupational therapy.

99214 (Moderate Risk): Prescription drug management requiring intensive monitoring for toxicity (e.g., warfarin, methotrexate, insulin adjustment). A decision regarding hospitalization. A decision regarding minor surgery with identified risk factors. Diagnosis or treatment significantly limited by social determinants of health.

The jump from low to moderate risk often comes down to one word in your documentation. “Continue lisinopril” is low risk. “Adjust insulin dosage from 20 to 25 units, monitoring for hypoglycemia” is moderate risk. The clinical work may be identical — the documentation is what separates the code levels.

Documentation Examples: Side by Side

This note supports 99213:

“Patient presents for follow-up of hypertension. Blood pressure 128/78, at target. Continue lisinopril 10mg daily. Return in 6 months.”

One stable chronic condition. No external data reviewed. Low-risk medication management. This is solidly a 99213.

This note supports 99214:

“Patient presents for follow-up of hypertension and newly elevated A1C (7.6%, up from 6.9% six months ago). Reviewed outside lab results from endocrinology referral. Blood pressure 138/86 — above target. Increased lisinopril to 20mg, added metformin 500mg BID with monitoring for GI side effects and renal function recheck in 3 months. Discussed diet modification and referral to diabetes education.”

Two chronic conditions (one worsening). External data reviewed. Prescription drug management requiring monitoring. This meets moderate complexity in all three MDM elements.

The Revenue Math

The 2024 National Medicare average reimbursement:

• 99213: ~$110
• 99214: ~$165
• Difference: ~$55 per visit

Commercial payers typically reimburse 120–180% of Medicare rates, pushing the per-visit difference to $65–$100+.

If you undercode just 5 visits per day from 99214 to 99213:

• 5 visits × $55 = $275/day
• 250 working days = $68,750/year in lost revenue

For a 4-physician group, that is $275,000 annually — revenue that was earned and documented, but never billed.

Time-Based Coding: The Alternative Path to 99214

Under AMA 2021 guidelines, you can code by total time on the date of encounter instead of MDM — whichever yields the higher code:

• 99213: 20–29 minutes total time
• 99214: 30–39 minutes total time

Total time includes chart review, the encounter itself, ordering tests, care coordination, and documentation — all on the date of service. Many physicians who would code 99213 by MDM actually spend 30+ minutes when all activities are counted. Telehealth visits are especially prone to this pattern, where significant pre-visit and post-visit work pushes total time above 30 minutes.

Common Mistakes That Lock You Into 99213

1. Defaulting to “stable” language. Writing “condition stable” when the patient actually has a mild exacerbation locks you out of moderate complexity on Element 1. If A1C rose, blood pressure is above target, or symptoms changed — document the change.

2. Not specifying data reviewed. “Reviewed labs” is vague and may not count. “Reviewed CBC and CMP from 4/10; independently interpreted ECG” explicitly meets the moderate data threshold.

3. Omitting risk language. “Continue metformin” does not convey risk. “Continue metformin with renal function monitoring due to CKD stage 2” establishes moderate risk.

4. Ignoring time-based coding. If you spent 32 minutes on the encounter and your MDM only supports 99213, the time-based code is 99214. Many physicians do not track or document total time, losing this upgrade path entirely.

How AI Identifies the Right Code

The challenge with manual coding is that you are making this determination 20+ times per day under time pressure. Undercoding from cognitive fatigue is not a character flaw — it is a systems problem.

CodeItRight’s AI analyzer extracts every MDM element from your clinical note automatically. It maps each element to the AMA 2021 framework, calculates both the MDM-based and time-based codes, and shows you exactly which elements pushed the code to 99213 or 99214. If your documentation is one sentence away from supporting the higher code, it flags the specific gap.

The result is not “the AI says 99214.” It is a full MDM breakdown with audit-defensible rationale — the same documentation you would need if a payer downcodes your claim.

FAQ: 99213 vs 99214

Q: Is billing 99214 instead of 99213 considered upcoding?
A: Only if your documentation does not support moderate complexity MDM or 30+ minutes of total time. Billing the code that your documentation legitimately supports is accurate coding, not upcoding. The risk is actually the reverse — undercoding is far more common and costs practices far more revenue.

Q: Can I bill 99214 for a single chronic condition?
A: Yes, if that condition has a mild exacerbation, progression, or side effects of treatment. A stable, well-controlled single chronic condition supports 99213. A worsening or complicated single chronic condition can support 99214.

Q: What if my MDM supports 99213 but I spent 35 minutes?
A: Bill 99214 using time-based coding. AMA 2021 guidelines explicitly allow you to choose the higher of MDM-based or time-based codes. Document total time and a brief description of how it was spent.

Q: Do payers flag 99214 claims more than 99213?
A: Some payers have automated edits that flag high rates of 99214 relative to specialty peers. The defense is documentation, not avoidance. A well-documented 99214 with clear MDM rationale survives audit. An undercoded 99213 just costs you money.