CorePulse · Coding Intelligence

Leveling E/M by MDM in 2026: what the engine weighs that a coder rushes

Under the 2026 AMA guidelines, an office visit level is decided by medical decision making, not by how many boxes a tired coder ticks at 4:47pm. CorePulse scores the same grid every time, in the same order.

The short answer: the level is the highest tier reached by 2 of the 3 MDM elements: problems addressed, data reviewed, and risk. CorePulse reads all three from the note, takes the second-highest, and then checks one separate question for the G2211 complexity add-on: is this a continuing, longitudinal relationship? Those are two different decisions, and rushing collapses them into one.
The rule
2 of 3
MDM elements set the level, not all three
Office E/M
4 tiers
Straightforward, Low, Moderate, High MDM
Add-on
G2211
Longitudinal-care complexity, level-independent
Consistency
1 order
Same grid, same sequence, every chart
The grid the engine scores

Three elements. Two of them decide.

The 2026 AMA E/M guidelines keep the same three MDM elements. A visit reaches a level when at least two of the three meet that tier. Here is the map CorePulse reads against.

01 / ELEMENT

Problems Addressed

The number and complexity of problems managed at the encounter, weighted by acuity and prognosis.

  • STRAIGHTFWD1 self-limited or minor problem
  • LOW2 minor, or 1 stable chronic, or 1 acute uncomplicated
  • MODERATEChronic exacerbation, undiagnosed new problem, or acute systemic illness
  • HIGHSevere exacerbation or a threat to life or bodily function
02 / ELEMENT

Data Reviewed & Analyzed

Notes, unique tests, independent interpretation, and discussion with other clinicians, scored across the three data categories.

  • STRAIGHTFWDMinimal or none
  • LOWLimited: one category partially met
  • MODERATEMeets one of the three data categories
  • HIGHExtensive: meets two of the three categories
03 / ELEMENT

Risk of Complications

Risk of morbidity or mortality from the problems, testing, and treatment discussed, including social determinants that raise risk.

  • STRAIGHTFWDMinimal risk
  • LOWLow risk of morbidity
  • MODERATEPrescription drug management, or SDOH-limited care
  • HIGHDrug therapy needing intensive monitoring, or hospitalization decision
Interactive · MDM leveler

Walk the grid the way CorePulse does

Set each element and CorePulse takes the level met by two of three. Then flip the longitudinal-care switch to see where G2211 legitimately attaches. Nothing here is a substitute for the note.

01 · Problems addressed
02 · Data reviewed & analyzed
03 · Risk of complications
CorePulse selects
99214Moderate MDM · established patient
PROBLEMS
3
DATA
3
RISK
3
Two of three elements reach Moderate, so the level is Moderate. The lowest element does not pull it down.
+ G2211 attaches — longitudinal relationship documented

Educational tool. Reflects the 2026 AMA CPT E/M guidelines for office/outpatient visits; it does not read your documentation and is not a coding determination. Final leveling depends on what the note supports.

Where G2211 legitimately attaches

The add-on is a relationship question, not a level question

HCPCS G2211 recognizes the visit complexity inherent to primary and longitudinal care. It is reported in addition to the office E/M, and the trap is treating it as a reward for a high level. It is not.

G2211 attaches when

  • The clinician is the continuing focal point for all of the patient's care, or is treating a single serious or complex condition over time.
  • It rides on an office/outpatient E/M (99202-99215), including lower levels when the relationship supports it.
  • The note reflects the ongoing relationship, not a one-time or unrelated encounter.

The rushed mistakes

  • ×Appending it to a discrete, one-off visit with no continuing relationship.
  • ×Assuming it only pairs with a 99214/99215 because those "feel" complex.
  • ×Reporting it without the documentation that shows the longitudinal focal-point role.
What a rush loses that the engine holds

Same note, two outcomes

The coder in a hurry

  • !Levels on the problem list length instead of the 2-of-3 rule, so a stack of stable chronics inflates the visit.
  • !Counts "prescription drug management" as automatic Moderate risk without confirming the note actually manages a drug.
  • !Misses independent interpretation and external-clinician discussion in the data element, under-leveling a real Moderate.
  • !Bolts G2211 to the top level and drops it from the legitimate lower-level longitudinal visits.

CorePulse

  • Scores each element independently, then applies 2-of-3 the same way on chart one and chart ten thousand.
  • Ties Moderate risk to a documented drug decision or SDOH factor, not a keyword.
  • Reads all three data categories so credit is neither missed nor invented.
  • Evaluates G2211 as a separate longitudinal-care check, decoupled from the E/M level.
Cited by name

The 2026 guidance this is built on

AMA CPT

2026 AMA E/M Guidelines — MDM elements

The three-element MDM framework (problems addressed, data reviewed and analyzed, risk) and the 2-of-3 leveling rule for office and outpatient visits, 99202-99215.

CMS

CY2026 Medicare Physician Fee Schedule Final Rule

Sets office/outpatient E/M valuation and payment policy for 2026, and the RVUs the selected level maps to under Medicare.

CMS / HCPCS

G2211 visit complexity add-on

The Medicare add-on for the inherent complexity of primary and longitudinal care, reported alongside an office/outpatient E/M when the continuing relationship is documented.

Documentation

Note-supported leveling

Every element score and the G2211 attachment must be supported by the record. The engine surfaces the driver; the documentation defends it.

Level from the note, not the clock

CorePulse scores the 2026 MDM grid the same way on every chart, flags the real level driver, and keeps G2211 as its own longitudinal-care decision. Fewer under-levels left on the table, fewer over-levels waiting to be clawed back. See it run on your own visit mix.

Book a CorePulse walkthrough → Talk to an ASP-RCM coding lead