Number and Complexity of Problems Addressed: The First MDM Column

The first column of the MDM table looks like the easy one. It is a list of problem types, arranged from “one self-limited or minor problem” up to “an acute or chronic illness or injury that poses a threat to life or bodily function,” and a clinician reading it usually feels they already know where their patient sits.

That confidence is the problem. Almost every term in this column has a definition in the CPT guidelines that differs from ordinary clinical usage — sometimes mildly, and in the case of “stable,” substantially and in a direction most clinicians find counter-intuitive. This page works through the column definition by definition, because the definitions are the column.

Verified against the American Medical Association’s CPT Evaluation and Management (E/M) Code and Guideline Changes, effective 1 January 2023 — the E/M Services Guidelines section “Number and Complexity of Problems Addressed at the Encounter,” Table 1, and the definitions of terms — and against CMS Medicare Learning Network booklet MLN006764, Evaluation and Management Services (May 2026 revision). CPT codes, descriptions and other data are copyright the American Medical Association.

The Threshold Question: Was the Problem “Addressed”?

Before any problem can be counted at any level, it has to clear a gate. The guidelines define the gate narrowly:

“A problem is addressed or managed when it is evaluated or treated at the encounter by the physician or other qualified health care professional reporting the service. This includes consideration of further testing or treatment that may not be elected by virtue of risk/benefit analysis or patient/parent/guardian/surrogate choice. Notation in the patient’s medical record that another professional is managing the problem without additional assessment or care coordination documented does not qualify as being addressed or managed by the physician or other qualified health care professional reporting the service. Referral without evaluation (by history, examination, or diagnostic study[ies]) or consideration of treatment does not qualify as being addressed or managed.”

Three tests are packed into that paragraph, and they are the tests a reviewer applies first.

• Evaluated or treated by you. A condition another clinician is managing does not count merely because it is listed. It counts if you assessed it or coordinated care for it, and documented that you did.
• Considered-but-rejected counts. Deciding against a test or treatment after a risk/benefit discussion is addressing the problem. This is generous and widely under-claimed.
• Bare referral does not count. “Refer to dermatology” with no evaluation and no consideration of treatment is not management of the lesion.

The problem itself is defined broadly — “a disease, condition, illness, injury, symptom, sign, finding, complaint, or other matter addressed at the encounter, with or without a diagnosis being established at the time of the encounter.” You do not need a diagnosis to have a problem. You need to have done something about it.

The problem-list trap

The most common single error in this column is treating the EHR problem list as the count. The guidelines shut that down explicitly:

“Comorbidities and underlying diseases, in and of themselves, are not considered in selecting a level of E/M services unless they are addressed, and their presence increases the amount and/or complexity of data to be reviewed and analyzed or the risk of complications and/or morbidity or mortality of patient management.”

Note the structure: a comorbidity earns its place by increasing data or risk. Carrying eight chronic diagnoses does not by itself make a visit complex. A note that pulls the full problem list forward and calls it “multiple chronic conditions” is asserting a count the guidelines do not recognise.

The symptom-cluster trap

The reverse error is splitting one condition into several. “Symptoms may cluster around a specific diagnosis and each symptom is not necessarily a unique condition.” Cough, fever and pleuritic pain in a patient with pneumonia are one problem, not three.

The Definitions, Level by Level

Minimal problem

“A problem that may not require the presence of the physician or other qualified health care professional, but the service is provided under the physician’s or other qualified health care professional’s supervision.” This tier maps to the nurse-visit level of service, where MDM is not used to select the level at all.

Self-limited or minor problem

“A problem that runs a definite and prescribed course, is transient in nature, and is not likely to permanently alter health status.” All three conditions have to hold. Definite course, transient, no lasting change.

Stable, chronic illness — the definition that surprises people

This is the single most misapplied definition in the whole table, and it has two independent parts.

First, chronicity: “A problem with an expected duration of at least one year or until the death of the patient. For the purpose of defining chronicity, conditions are treated as chronic whether or not stage or severity changes (eg, uncontrolled diabetes and controlled diabetes are a single chronic condition).”

Second, and this is the part that changes claim outcomes:

“‘Stable’ for the purposes of categorizing MDM is defined by the specific treatment goals for an individual patient. A patient who is not at his or her treatment goal is not stable, even if the condition has not changed and there is no short-term threat to life or function. For example, a patient with persistently poorly controlled blood pressure for whom better control is a goal is not stable, even if the pressures are not changing and the patient is asymptomatic. The risk of morbidity without treatment is significant.”

Read that again against the way clinicians normally speak. A patient whose blood pressure has been 158/94 at every visit for two years is, in ordinary usage, “stable” — nothing is changing. Under the MDM definition they are not stable, because they are not at goal. That single reclassification moves the patient from “one stable chronic illness” (low) toward “chronic illness with exacerbation, progression, or side effects of treatment” (moderate), provided the note shows an intent to control and additional supportive care or attention to treatment.

The discipline this demands is documentary, not clinical: the note has to state the treatment goal and whether the patient is at it. “BP 158/94, unchanged” says nothing about goal. “BP 158/94, above target of <140/90 despite current therapy; adding second agent” places the problem precisely.

Acute, uncomplicated illness or injury

“A recent or new short-term problem with low risk of morbidity for which treatment is considered. There is little to no risk of mortality with treatment, and full recovery without functional impairment is expected.” The guidelines add a useful escalator: “A problem that is normally self-limited or minor but is not resolving consistent with a definite and prescribed course is an acute, uncomplicated illness.” A cold that has not resolved in three weeks has moved up a tier by failing to behave.

Stable, acute illness

“A problem that is new or recent for which treatment has been initiated. The patient is improved and, while resolution may not be complete, is stable with respect to this condition.” This is the follow-up visit for something that is getting better.

Chronic illness with exacerbation, progression, or side effects of treatment

“A chronic illness that is acutely worsening, poorly controlled, or progressing with an intent to control progression and requiring additional supportive care or requiring attention to treatment for side effects.” Note “poorly controlled” sitting alongside “acutely worsening” — this is where the not-at-goal patient from the stability definition lands.

Undiagnosed new problem with uncertain prognosis

“A problem in the differential diagnosis that represents a condition likely to result in a high risk of morbidity without treatment.” The test is about the differential, not the eventual answer. This connects to a general principle stated earlier in the guidelines: “presenting symptoms that are likely to represent a highly morbid condition may ‘drive’ MDM even when the ultimate diagnosis is not highly morbid,” provided “the evaluation and/or treatment should be consistent with the likely nature of the condition.” You cannot claim a serious differential you did not work up.

Acute illness with systemic symptoms

“An illness that causes systemic symptoms and has a high risk of morbidity without treatment.” The guidelines immediately fence this off from ordinary viral illness: “For systemic general symptoms, such as fever, body aches, or fatigue in a minor illness that may be treated to alleviate symptoms, see the definitions for self-limited or minor problem or acute, uncomplicated illness or injury.” Fever alone does not make a cold into a moderate problem. They also note that “systemic symptoms may not be general but may be single system.”

Acute, complicated injury

“An injury which requires treatment that includes evaluation of body systems that are not directly part of the injured organ, the injury is extensive, or the treatment options are multiple and/or associated with risk of morbidity.” Any one of the three limbs qualifies.

Chronic illness with severe exacerbation

“The severe exacerbation or progression of a chronic illness or severe side effects of treatment that have significant risk of morbidity and may require escalation in level of care.” “May require escalation” is the discriminator between this high tier and the moderate exacerbation tier.

Acute or chronic illness or injury that poses a threat to life or bodily function

“An acute illness with systemic symptoms, an acute complicated injury, or a chronic illness or injury with exacerbation and/or progression or side effects of treatment, that poses a threat to life or bodily function in the near term without treatment.” The guidelines then extend it to strong suspicions: “Some symptoms may represent a condition that is significantly probable and poses a potential threat to life or bodily function. These may be included in this category when the evaluation and treatment are consistent with this degree of potential severity.”

Condition Risk Is Not Management Risk

One sentence in this section prevents a great deal of double-counting:

“The term ‘risk’ as used in these definitions relates to risk from the condition. While condition risk and management risk may often correlate, the risk from the condition is distinct from the risk of the management.”

The word “risk” appears in both column 1 and column 3 of the table and means different things in each. Column 1 asks how dangerous the illness is. Column 3 asks how dangerous your management of it is. A patient with a life-threatening condition managed conservatively can be high in column 1 and moderate in column 3; a patient with a benign condition for whom you elect major surgery can be the reverse. Treating them as one judgement is how a single clinical fact gets counted twice. The risk column guide works the other side of that line.

Two Worked Examples

Example 1: the deceptively simple follow-up

A 54-year-old man attends for review of hypertension and type 2 diabetes. Blood pressure is 152/92 against a target of below 140/90; HbA1c is 8.4% against a target of below 7%. Both have been at these levels for a year. The clinician documents the targets, states that neither is at goal, adds a second antihypertensive and increases metformin.

The instinct is “two stable chronic illnesses” — moderate, but only just. The definitions say otherwise. Neither condition is at its treatment goal, so neither is stable. Both are chronic illnesses with the clinician documenting an intent to control and initiating additional treatment. This is “one or more chronic illnesses with exacerbation, progression, or side effects of treatment,” which is the moderate row reached by a firmer route. The level did not change, but the reasoning did — and if only one condition had been off goal, the difference between the two routes would have mattered.

Example 2: the diagnosis that resolves downward

A 41-year-old woman presents with three days of central chest discomfort, worse lying flat. The differential documented includes acute coronary syndrome, pericarditis and reflux. An ECG is performed and interpreted, troponin is ordered, and she is observed in the department. Everything returns normal; the working diagnosis on discharge is gastro-oesophageal reflux.

Coded on the final diagnosis, this is a minor problem. Coded correctly, it is not. The guidelines state that “the final diagnosis for a condition does not, in and of itself, determine the complexity or risk, as extensive evaluation may be required to reach the conclusion that the signs or symptoms do not represent a highly morbid condition.” The presentation was an undiagnosed new problem with uncertain prognosis at minimum, and the documented differential plus the workup make the “significantly probable… potential threat to life” extension available. What makes it defensible is that the evaluation matched the claimed severity. A note listing acute coronary syndrome in the differential with no ECG and no troponin does not.

Frequently Asked Questions

Do I count a problem I addressed but did not diagnose?

Yes. A problem exists “with or without a diagnosis being established at the time of the encounter.” Undiagnosed problems have their own tier and it is not a low one.

The patient has ten chronic conditions. Is that automatically high?

No. Count only those addressed at this encounter, and remember that comorbidities not addressed count only if they increase the data or the risk. Two chronic illnesses genuinely managed at the visit reach moderate; ten copied forward reach nothing.

Is a controlled condition the same chronic illness as an uncontrolled one?

For chronicity, yes — the guidelines give exactly this example, that uncontrolled diabetes and controlled diabetes are a single chronic condition. For stability, no: control against the individual treatment goal is what decides the row.

Can a coder decide whether a problem was stable?

No. The AMA’s E/M guidance is that the physician, not the coder, determines whether problems are stable or worsening, and the documentation should reflect that clinical assessment. Where the note is ambiguous the correct route is a query, not a reclassification.

Does a problem addressed by telephone or portal message count here?

These guidelines govern services requiring a face-to-face encounter. Work done outside the encounter may count toward total time on the date of service, or may belong to a separate non-face-to-face service entirely. For telehealth encounters that are the visit, see the telehealth modifier guide for how the service itself is identified.

How does this column interact with a same-day procedure?

The problems column measures the E/M work. Whether that E/M work is separately billable alongside a procedure is a different test entirely, addressed in the modifier 25 guide. A problem can be genuinely addressed and still fail the separately-identifiable test if the work was inherent in the procedure.

What single change to my notes would help most?

State the treatment goal and whether the patient is at it, for every chronic condition you address. It is one clause, it is clinically useful independently of billing, and it is the fact on which the stability definition turns.

Once the problems column is settled, the other two follow their own rules — the data column with its counting thresholds, and the risk column with none at all. The overview guide shows how the three combine under the 2-out-of-3 rule.