A visit can feel routine and still support 99214. A prescription decision may still support 99213. A stressful, lengthy visit does not automatically become 99215.
When you select office/outpatient E/M by medical decision making (MDM), the code depends on two of three elements:
At least two elements must meet or exceed the level being billed. One moderate element is not 99214. One high element is not 99215.
| 99213 | 99214 | 99215 | |
|---|---|---|---|
| Overall MDM | Low | Moderate | High |
| Problems example | One stable chronic illness at treatment goal | Two stable chronic illnesses; one with exacerbation, progression, or side effects of treatment; or an undiagnosed new problem with uncertain prognosis | Severe exacerbation or threat to life or bodily function |
| Data example | Two qualifying items or a necessary independent historian | Three Category 1 items (a necessary independent historian can be one of them), qualifying independent interpretation, or external management discussion | Two of three Data categories |
| Risk example | Low-risk management | Prescription drug management | Hospitalization decision or therapy requiring intensive monitoring for toxicity |
Those are examples of individual elements, not code assignments. You still need two qualifying elements for the overall MDM level.
This element concerns conditions evaluated or managed at this encounter, not every diagnosis in the chart.
Two diagnoses on a problem list do not automatically establish moderate Problems. Likewise, a severe diagnosis label does not establish high Problems without the clinical facts.
Not at treatment goal is not stable. CPT defines stability against the individual patient's treatment goals, not merely whether symptoms have changed since the previous visit. A chronic condition that remains poorly controlled, even without recent worsening, can support moderate Problems when that incomplete control is evaluated and managed. Think persistent ADHD symptoms with meaningful impairment despite treatment, or depression that has improved but remains short of the treatment goal. The assessment must establish the problem and how it was addressed, not just say "not at goal."
Do not confuse the risk of the illness (Problems) with the risk of your management decisions (Risk).
Limited Data (low) requires two qualifying items (unique external notes, unique tests reviewed or ordered), or an assessment requiring an independent historian.
Moderate Data requires one of:
Extensive Data (high) requires two of these three categories. Three or even several items in Category 1 alone cannot make Data high.
Psychiatry-specific pitfalls:
Count the defined activities, not the number of documents on your desk.
Risk is not a second way to count diagnoses. It concerns the consequences of management decisions made, including meaningful options considered but not selected.
Merely treating multiple conditions is not moderate Risk. Financial or psychosocial stressors are not automatically moderate Risk, although diagnosis or treatment significantly limited by social determinants of health may qualify.
What qualifies as intensive toxicity monitoring? The drug has the potential for serious morbidity or death; monitoring is performed to detect adverse effects, not primarily to measure efficacy. It ordinarily involves laboratory, physiologic, or imaging tests, not history or examination alone. Long-term intensive monitoring is performed at least quarterly, although a patient-specific or shorter-term schedule may differ. That monitoring must actually be considered in management at this encounter.
Examples that can meet the criterion include clozapine with appropriate ANC monitoring for severe neutropenia, or lithium with levels and related testing drawn at least quarterly to monitor for serious toxicity. A stable patient with levels every six months does not meet the long-term criterion. The actual drug regimen, purpose and frequency of testing matter; the drug's name alone is not sufficient. Clozapine's federal REMS program ended in 2025, but the FDA continues to recommend ANC monitoring according to the prescribing information.
The patient reports sustained benefit, ADHD symptoms at the agreed treatment goal, and no important adverse effects. You assess response, tolerability, and safety and document a decision to continue prescribed treatment. No other condition is addressed; no qualifying Data is present.
Problems: Low (one stable illness). Data: Minimal. Risk: Moderate (prescription drug management).
Overall: Low MDM → 99213.
One moderate element is not enough for 99214. The prescription decision does not upgrade Problems.
Same established patient and active prescription management, but ADHD symptoms remain poorly controlled relative to the patient's agreed treatment goals, with ongoing impairment. You assess the partial response, discuss appropriate next steps, and make a documented management decision addressing the incomplete response. Data is still minimal.
Problems: Moderate (one poorly controlled chronic illness actively managed). Risk: Moderate (prescription drug management). Data: Minimal.
Overall: Moderate MDM → 99214.
The medication does not have to change for moderate Risk if a genuine management decision is made. The difference is that the condition is not at its treatment goal, so Problems can also be moderate. This isn't a rule that every patient who reports residual symptoms automatically gets 99214; the individualized goal and clinical management have to support the classification.
At follow-up you evaluate both ADHD and MDD. Each remains at its treatment goal. You assess benefits, adverse effects, and ongoing indication for the prescribed regimen and decide to continue it unchanged. Data is minimal.
Problems: Moderate (two stable illnesses addressed). Risk: Moderate (prescription drug management). Data: Minimal.
Overall: Moderate MDM → 99214.
No medication change is necessary. The difference from Case A is the second moderate element, not a longer note.
Another 99214 pathway: a worsening chronic depressive condition plus a documented antidepressant dose adjustment may establish moderate Problems and moderate Risk.
An established patient presents with severe acute mania, dangerous behavior, and a serious near-term safety threat. You evaluate the risk and alternatives and make and document a decision for emergency hospitalization, coordinating the transfer. Assume no additional qualifying Data.
Problems: High (condition threatening life or bodily function). Risk: High (hospitalization decision). Data: Minimal.
Overall: High MDM → 99215.
It is the actual severity and management decision that matter, not the diagnostic label, length of discussion, or safety-plan template. A single high element would not suffice.
Time is a separate E/M selection method. When no psychotherapy add-on is billed, office/outpatient E/M may generally be selected by qualifying total time or supported MDM. Time does not tell you how high the MDM is. For when each pathway fits, see Choosing Time vs MDM: A Practical Billing Guide.
Psychotherapy add-ons are different. With 90833, 90836, or 90838, the accompanying E/M level must be selected by MDM, not time. Psychotherapy must be actually performed, separately identifiable, and documented with its own minutes: 90833: 16–37; 90836: 38–52; 90838: 53 or more. Ordinary medication counseling or a difficult visit does not create a psychotherapy add-on.
None of the three cases above establishes psychotherapy, so none supports an automatic add-on code.
Ask three questions:
Now find the two elements supporting your proposed level.
Don't undercode because the visit was routine. Don't upcode because it was hard. Code the decisions and facts.
To run this framework in real time while you chart, use The 60-Second Code Check. For more on stable follow-ups, read Stable Does Not Automatically Mean 99213.
To practice the decisions, open Practice Lab: Billing & Coding and choose MDM Foundations.
Educational guidance for office/outpatient E/M. Actual code selection depends on the encounter, current CPT rules, applicable payer policy, and the documented service.