CPT 99213 vs 99214 documentation: what separates the two levels

What changed in 2021 and why it still trips people up

Before 2021, an established patient visit was scored on history, examination and medical decision making, and two of the three had to reach the level. Templates grew to capture review-of-systems bullets and exam elements that had nothing to do with the visit. The AMA rewrote the office visit codes 99202 to 99215 effective January 1, 2021. CMS adopted the same framework for Medicare. History and examination are now documented as medically appropriate and are not counted. The level is chosen by MDM alone, or by total time alone.

The problem is that many templates were never rebuilt. A note can still be three screens of normal findings and a one-line assessment, which supports nothing above 99213 no matter how sick the patient was. The fix is a note whose assessment and plan show the problems, the data and the risk in enough detail that a reviewer can score them.

The side-by-side table

Two of three is the whole rule. A visit with moderate problems and moderate risk is a 99214 even if no data was reviewed. A visit with moderate data and low problems and low risk is a 99213.

Element one: the problems addressed

"Addressed" is the operative word. A problem counts only if the clinician evaluated it or managed it at this visit. Listing six chronic diagnoses in the assessment does not make six problems addressed if the plan mentions two. The reverse also applies: a problem that was considered and deliberately left alone, with a note saying why, was addressed.

The distinction that decides most 99213 versus 99214 questions is "stable" versus "exacerbation, progression or side effects." In the CPT definition a chronic illness is stable only when the patient is at the treatment goal set for that condition. A patient with hypertension whose blood pressure is above goal has a chronic illness with progression even if the reading is the same as last visit, and that single problem is moderate. A patient at goal on the same medication is stable, and one stable chronic illness is low. Two stable chronic illnesses together are moderate. The note has to say which it is: "hypertension, controlled, continue lisinopril" reads as stable; "hypertension, above goal at 148/92, increase lisinopril" reads as progression.

Common problem-level mistakes

• Counting a problem that was only listed, not managed.
• Calling a chronic condition stable when the note shows it is not at goal.
• Treating a new problem with a clear diagnosis and a clear course as "undiagnosed with uncertain prognosis." That category is for genuine diagnostic uncertainty, such as a new breast lump awaiting imaging.
• Forgetting that an acute illness with systemic symptoms (fever, malaise, body aches with a diagnosis such as pyelonephritis) is moderate, not low.

Element two: data reviewed and analyzed

Data is counted per unique test or source, and only when it is not separately reported. Ordering a CBC and reviewing the result at the same visit is one test, not two. Reviewing three labs from an outside hospital is three unique tests. Reviewing the discharge summary from that hospital is one external note from one unique source. An independent historian is a parent, spouse or caregiver whose history is needed because the patient cannot give a reliable one, and it counts once.

Independent interpretation, the second moderate category, means the clinician looked at the image or tracing and recorded a reading, not that a radiology report was reviewed. Reading an X-ray in the office and writing "no acute fracture on my review" is independent interpretation if the practice is not billing the professional component. Discussion of management, the third category, is a real exchange with an outside clinician about this patient, documented with the name and the substance.

The failure mode here is unrecorded work. Clinicians routinely review outside records and results without a word in the note. If it was not documented, it was not counted, and a visit that was genuinely moderate on data drops to low.

Element three: risk

Risk is about the management chosen at the visit, not the patient's underlying condition. The examples in the CPT table are the working guide. Prescription drug management is moderate, and it includes starting, stopping, adjusting or deciding to continue a prescription medication after evaluating it. Continuing a drug counts only when the note shows the evaluation: the response, the side effects checked, the reason for no change. A refill line with no discussion is not management.

Over-the-counter medication and minor procedures without identified risk factors are low. A decision to refer for elective major surgery, or to proceed with a minor procedure in a patient with identified risk factors such as anticoagulation, is moderate. Social determinants of health count at moderate when they significantly limit diagnosis or treatment and the note says how, for example a patient who cannot afford the first-line drug and is switched to a lower-cost alternative.

Coding by time instead

Time is the alternative path, and it is a cleaner one when the visit was long but the decision making was simple: a stable diabetic who needed thirty minutes of counseling on diet and glucose monitoring, for example. Total time is all of the physician's or qualified health professional's own time on the date of the encounter, face to face and not, including chart preparation, history, examination, counseling, ordering, communicating with other professionals, documenting in the record, plus independently interpreting results and explaining them to the patient. It excludes clinical staff time, time on separately reported services, and travel.

Documentation needs one clear statement: total time, that it was on the date of the encounter, and roughly what it covered. Two cautions. Do not record time when the level was chosen by MDM, because a note that shows moderate MDM and 22 minutes invites a reviewer to argue. And do not reach for prolonged service codes. CPT 99417 and the Medicare code G2212 apply only when the 99215 threshold has been exceeded by at least 15 minutes, and Medicare requires the maximum time of 99215 to be passed rather than the minimum. Neither code ever attaches to a 99214. A 99214 that ran to 44 minutes is a 99215.

Building a note that supports the level

The practical answer to the CPT 99213 vs 99214 documentation question is an assessment and plan that scores itself. For each problem addressed, one line on status (stable at goal, worsening, new and uncertain), one on data (what was reviewed or ordered, from where), and one on management (the drug decision, the procedure decision, the referral). History and exam stay proportionate to the visit. Time appears only when time is the basis.

A short internal audit shows quickly whether the templates are helping or hurting. Pull ten established patient notes per clinician, score each on the three elements, and compare with the code billed. The usual finding in primary care is a cluster of 99213s where the plan shows prescription changes for conditions not at goal, which is undercoding, alongside a smaller set of 99214s where the note lists diagnoses but manages none of them, which is overcoding. Both are fixed by the same template change.

For the broader list of coding mistakes we see in audits, read our article on the top CPT coding errors. If your practice wants a second set of eyes on E/M levels before the payers provide one, DyBilling's coding audit service reviews a sample of notes against the current MDM table, and our medical coding team codes from the documentation rather than from the template. A free billing audit includes an E/M distribution report by clinician as a starting point.