One caveat to the "high level distribution looks suspicious" worry: a genuinely complex panel can justify a high average legitimately. If you practice a complex specialty, or your patients are disproportionately sick, severely affected, or medically complicated, a distribution weighted toward level 4 and level 5 may simply be accurate. Some practices legitimately average well above the norm. The defense is not a lower distribution; it is documentation that matches the complexity. Code what the visit was, and let the notes substantiate the pattern.
The prevailing assumption among providers is that aggressive coding carries audit risk and conservative coding is safe. This is only half true, and the half that is false is quietly expensive. The standard, legally and professionally, is accurate coding, full stop. A claim is supposed to reflect the service that was actually provided. Coding above what your documentation supports is a compliance risk. But a pattern of systematic undercoding misrepresents your work just as surely, distorts the claims record, and has real consequences for your productivity metrics, your bonus eligibility, and your long-term income. Undercoding is not the compliant choice. It is just the quiet one. This deserves emphasis, because the instinct runs the other way: a claim must accurately reflect the service provided, and a sustained pattern of coding below that level is a misrepresentation, not a safe harbor.
Why Providers Undercode: A Summary
Several patterns explain most undercoding:
Default codes. Many providers pick a comfortable default level for their most common visit type and apply it broadly. A surgeon whose typical established patient is a post-op follow-up defaults to 99213. A primary care physician whose busiest sessions are chronic disease management defaults to 99213 or 99214. These defaults are not calibrated to individual encounters.
Documentation habits that do not match clinical work. The visit was complex. The note does not clearly articulate the complexity. The coder reads the note, cannot find the documentation elements that support a higher level, and assigns the lower code, correctly, given what the note says. The clinical work happened; the note did not capture it.
Fear of audit. High-level codes feel risky. Level-5 visits attract attention. The response is to use level-4 when level-5 is warranted, or level-3 when level-4 is warranted. This logic applies downward pressure across the entire billing distribution.
Unfamiliarity with add-on codes. G2211, prolonged services, ultrasound guidance add-ons, modifier 25, these codes represent real additional work that generates real additional wRVUs, and many providers have never been taught they exist or how to use them.
Where the Gaps Are Most Common
Across outpatient practice, the most common undercoding gaps consistently occur in:
- E/M code level selection for established patients (defaulting to 99213 when 99214 is supported)
- New patient visits coded at 99203 when 99204 or 99205 is supported by the documentation
- Failure to bill Modifier 25 E/M when a procedure is performed and a separate evaluation clearly occurred
- Failure to use ultrasound guidance codes when guidance is used
- Missing G2211 for eligible longitudinal care visits
- Fracture care codes billed at the non-manipulation level when manipulation occurred
A Simple Self-Audit
Pull your last 30 established patient E/M visits. For each one, ask: does this note document two or more stable chronic conditions being actively managed? If yes, that visit meets the column 1 threshold for moderate complexity MDM. Does your note reflect prescription medication management or adjustment? If yes, that meets the column 3 threshold for moderate risk. Two of three columns met means 99214. Count how many of those 30 visits you billed at 99213. That is your initial baseline gap.
Then pull your last 10 visits that included a procedure. How many had a Modifier 25 E/M coded alongside the procedure? For the ones that did not, read the note: was there a clinical evaluation beyond the procedure itself? If so, that is missed revenue.
Sources and Further Reading for This Chapter
- HHS Office of Inspector General, evaluation and management coding. https://oig.hhs.gov/
- CMS, Physician Fee Schedule. https://www.cms.gov/medicare/payment/fee-schedules/physician