The E&M coding curve is where urgent care revenue and compliance risk meet. Code a visit too low and you give away earned revenue; code it too high and you invite a payer audit. This dashboard puts the whole 99202–99215 distribution on screen by provider and by center, so the medical director can see who codes outside the expected bell curve in either direction before a payer points it out.
What this dashboard answers
The coding-curve page answers the shape question: how visits distribute across the E&M levels for new and established patients, total RVUs, high-acuity share, and RVUs per visit — the headline intensity number. The provider-patterns page answers the outlier question: the E&M-level distribution by provider, high-acuity share by provider, and RVUs per visit by center, so a clinician who skews consistently low (lost revenue) or consistently high (audit risk) stands out immediately.
The metrics that matter
High-acuity share — the percentage at level 4 or 5 — is the curve's center of gravity, and the right number is a clinical-documentation question, not a target to maximize. Low-acuity visits (the level-2 count) is the under-coding flag, the quiet revenue leak where the documentation supported more than was billed. RVUs per visit summarizes documented intensity in a single comparable number, which is what makes provider-to-provider and center-to-center comparison meaningful.
Why the data is trapped
The coded visit and its E&M level live in Experity, eClinicalWorks, or athenahealth, but the analysis that matters — the full distribution by provider, the new-versus-established split, RVUs mapped from each code — isn't a standard report. RVU values have to be joined to each code, and provider-level distributions have to be normalized so a high-volume provider doesn't dominate the chart. Most groups never build this view, so coding drift goes unnoticed until an audit or a revenue dip surfaces it.
How to read it
Read high-acuity share by provider against the group — a provider far above the pack is an audit-risk conversation, one far below is a documentation-and-revenue conversation. The new-versus-established split matters because the two code families behave differently. Then read RVUs per visit by center to confirm whether a center's revenue gap is a coding-curve problem rather than a volume or mix one.
The sample on this page uses entirely synthetic coding data — anonymized providers and centers, standard E&M codes, no PHI.
Metrics it tracks
| Metric | What it means |
|---|---|
| Coded Visits | Count of coded visits in the period — one row per visit. |
| Total RVUs | Sum of work RVUs across all coded visits — documented clinical intensity. |
| High-Acuity Share | Share of visits coded at level 4 or 5 (99204/99205/99214/99215). |
| Low-Acuity Visits | Count of level-2 visits (99202/99212) — the under-coding flag. |
| High-Acuity Visits | Count of level-4/5 visits — the over-coding / audit-risk flag if outsized. |
| RVUs per Visit | Total work RVUs ÷ coded visits — average documented intensity per visit. |
Used by: Medical directors and compliance-minded owners auditing the coding curve
Frequently asked questions
What is a normal E&M coding distribution for urgent care?
Most urgent care visits land in the mid-level codes — 99203/99204 for new patients and 99213/99214 for established — forming a bell curve. The exact shape should reflect documented acuity, so the goal isn't a target percentage but spotting providers or centers whose curve sits far off the group's.
How does this catch under-coding and over-coding?
Under-coding shows up as an outsized share of low-acuity (level-2) visits — revenue left on the table where documentation supported more. Over-coding shows up as a high-acuity share well above peers, which is what triggers payer audits. Comparing each provider's curve to the group surfaces both before a payer does.
Can I build this on my own coding data?
Yes. Export coded visits with their E&M codes and rendering provider from your EMR/PM, and use this as a template. We map RVUs to each code and build the distribution; the sample here uses synthetic data so you can see the finished layout first.
Build this report on your own data
Clone this Urgent care template — describe it and we’ll generate sample data so you can try it free, or upload your own export. You get a fully modeled, branded Power BI project that opens in Power BI Desktop.
Use this as a template →