Complete Guide to CPT Codes 99211–99215: E/M Billing, Documentation, and Reimbursement Explained
- Sirius solutions global
- 20 hours ago
- 5 min read

You Are Probably Leaving $80,000 a Year on the Table
Not from fraud. Not from complex billing failures. From one habit.
The habit of selecting 99213 because the visit felt like a 99213 and moving on without checking whether the documentation supports something higher.
Primary care practices that audit their E/M distribution find the same thing consistently: 30 to 45 percent of visits coded as 99213 should have been 99214. Medicare revenue difference between those two codes: $48 per visit. Commercial: $55 to $70. Across 18 patients a day, five days a week, that becomes a number that is hard to ignore.
The inverse is equally real. Practices that default upward, billing 99214 or 99215 without documentation support, are building quiet audit exposure that arrives as a recoupment request, not a denial.
This guide covers what these codes actually require, how the 2021 rule changes work, what documentation supports each level, and how to audit your own charts today.
The Code Family: One Table, No Confusion
These are established patient codes. New patients use 99202–99205, same MDM logic, higher reimbursement, different range.
Before 2021: point-counting. History elements, exam components, review of systems, level came from those counts.
That system is gone.
Since January 2021, office E/M visits code on one of two pathways:
Pathway A — Medical Decision Making: Complexity of clinical thinking. How many problems, what data reviewed, what risk. Needs two of three MDM components at the same level.
Pathway B — Total Time: Total time on the encounter date, pre-visit chart review, face-to-face, post-visit documentation, ordering and reviewing results.
The exam does not drive the level anymore. Head-to-toe physical with simple MDM = 99212. Brief physical with moderate MDM = 99214. Providers still using the old mental model are almost certainly undercoding complex visits and potentially overcoding visits where the documentation does not match the complexity they managed.
MDM has three parts. Two of the three must hit the same level for that code to be billable.
Part 1 — Problems Addressed
Part 2 — Data Reviewed and Analyzed
Where most practices leave MDM credit unclaimed. "Data" is not just lab orders:
Reviewing results of prior tests or imaging
Reviewing external records or specialist notes
Independently interpreting a test (not just seeing the result)
Discussing the case with another provider
Using an independent historian (parent, caregiver, interpreter)
Part 3 — Risk
The Grid: Two of Three Wins
Code | To Bill It, You Need... |
99212 | 2 of 3 MDM components at minimal/low |
99213 | 2 of 3 MDM components at low |
99214 | 2 of 3 MDM components at moderate |
99215 | 2 of 3 MDM components at high |
Code | Minimum Total Time |
99212 | 10 minutes |
99213 | 20 minutes |
99214 | 30 minutes |
99215 | 40 minutes |
Counts toward time: pre-visit chart review, face-to-face time, post-visit documentation, ordering/reviewing results that date. Does not count: work on other days, time on other patients, unrelated administrative tasks.
One sentence makes time billing valid: "Total time for this encounter: 38 minutes." Backed by a note that reflects the work, that line supports 99214. Most providers who spend 38 minutes on a complex patient walk out and bill 99213 by habit.
99211 — The Nurse Visit
No physician, NP, or PA required. BP check, prescription pickup, nurse education. The provider does not need to be present. Billing it under a provider who was not involved is a coding error.
99212 — The Genuinely Simple Visit
One minor, self-limited problem. No complex thinking required.
✔ Wart removal, healthy established patient. ✔ Medication refill, stable uncomplicated condition, nothing new. ✗ Diabetic patient with elevated A1C, that visit is not simple.
99213 — The Most Overclaimed Code in Medicine
The default. The comfort zone. The code that quietly drains revenue.
99213 is correct for stable chronic disease managed without complications, or a minor new problem that is straightforward. It is not correct for:
Hypertensive patient, BP 158/96 at today's visit
Anxiety medication not working
Chest pain that needs EKG and labs
Any visit where you are genuinely uncertain what is happening
Those visits have moderate complexity. They qualify for 99214, if the note shows it.
99214 — The Most Under-Billed Code in Medicine
The visits that deserve this code are not rare. They are Tuesday afternoon.
99214 applies when:
A chronic condition is not controlled, exacerbating, or presenting with a new symptom
A new problem requires workup before it can be managed
You reviewed a specialist's note, independently interpreted a test, or discussed the case with a colleague
Management involves a medication requiring monitoring
The note needs: moderate problem complexity + moderate data reviewed OR moderate risk management. Two of three.
Hypertension and diabetes, elevated readings on both, two medications adjusted, that is 99214. Bill it. But if the note says "BP elevated, adjusted dose, RTC 1 month" and nothing else. It does not support 99214 even if the visit justified it.
Documentation drives the level. Not the visit.
99215 — The Code Providers Fear
Most providers talk themselves out of 99215 before they leave the room. They do not want the scrutiny.
It is appropriate and fully defensible when the documentation reflects it:
Multiple chronic conditions with exacerbation
New problem with significant risk requiring immediate workup
High-risk management decision or hospitalization discussion
Drug therapy requiring toxicity monitoring (warfarin, lithium, chemotherapy)
Visit runs 45+ minutes? Document total time, bill on time. That is a defensible 99215, nothing more required than an accurate time note and a chart that reflects the work.
Bell curve that never moves — 85 percent 99213 across all patients for three years. No panel is that uniformly stable. Payers flag statistical uniformity as a documentation red flag.
99215 above specialty benchmark — High volume gets reviewed. Not because it is wrong because it is statistically uncommon. Every 99215 stands on its own documentation.
Copy-paste notes — Identical findings across dozens of visits. Payers run similarity scoring. 90 percent match rate across encounters signals template-driven documentation, not actual clinical work.
MDM complexity without matching data — 99214 on problem complexity alone, no data or risk element documented. MDM needs two of three. Problem alone is one of three.
Reimbursement Reality in 2026
Code | Medicare | Commercial (TX avg.) | Medicaid (TX avg.) |
99211 | ~$24 | $20 – $35 | $15 – $25 |
99212 | ~$58 | $55 – $85 | $35 – $55 |
99213 | ~$97 | $90 – $140 | $60 – $85 |
99214 | ~$145 | $135 – $210 | $90 – $120 |
99215 | ~$207 | $190 – $290 | $120 – $160 |
Medicare gap between 99213 and 99214: $48. Commercial gap: $45 to $70. At 20 visits per day, 30 percent miscoded one level down: $105,000+ annually.
That number is recoverable. It does not require seeing more patients. It requires documenting the visits you are already having.
Pull 20 random charts from the last 30 days. Three questions per chart:
What MDM does this note actually support? Not what you intended to bill what the documentation would support if a payer pulled it.
Is total time documented if time was the pathway? "Total time: X minutes" needs to appear in the note.
Is there a pattern? Sixteen of 20 coded 99213, that conversation needs to happen.
E/M coding accuracy is not a one-time project. It requires chart audits, documentation coaching, and a billing workflow that surfaces patterns before payers do.
At Sirius Solutions Global, E/M levels are reviewed against documentation on every claim, not spot-checked. Undercoding patterns get specific feedback: which MDM elements are missing, where time documentation is absent, which visit types are consistently coded below what the complexity supports. Quarterly audits catch the patterns before they become audit letters.
If your practice has not reviewed E/M distribution against documentation in the past year, request a free coding review from Sirius Solutions Global. Most practices find measurable, recoverable gaps in the first session.

