Documentation savings estimator
Enter your annual volume and the shares you know; keep the defaults for the ones you do not. Each default is labelled as a published figure or as our assumption, and the arithmetic is written out under the result so you can check every step.
Need the whole report too? Type a few findings and get it written ↓
| Per year | Low | Mid | High |
|---|---|---|---|
| Avoidable denial rework | $234,375 | $411,000 | $536,531 |
| Revenue lost to downcoding | $18,000 | $18,000 | $18,000 |
| Total | $252,375 | $429,000 | $554,531 |
| Per 10,000 studies | $10,095 | $17,160 | $22,181 |
A sample dictation is ready below. Press generate, or type your own findings and end with “otherwise normal” to have the rest written out. No signup.
100 free reports over 14 days. No card, nothing to install.
Start your free trial →Please leave out patient identifiers. Text is de-identified before the AI sees it and is not stored. Demo output is not for clinical use.
How the arithmetic works
Two cost streams, added together. Both are linear in your volume, so the "per 10,000 studies" line is the same number scaled; nothing is compounded.
- Avoidable denial rework. Annual volume × first-pass denial rate × share of denials attributable to indication or documentation gaps = denials a year that better documentation might have avoided. Each one costs the per-claim rework figure. The low / mid / high bands use the three published rework figures: about $25 (MGMA, as cited by AAFP), $43.84 (Premier, 2022 claims) and $57.23 (Premier, 2023 claims).
- Revenue lost to downcoding. Annual volume × share billed as complete multi-element exams × share of those reports missing a required element = exams a year that are billed at the limited code because the report did not document every element the complete code requires. Each one loses the professional-component gap between the two codes. This stream uses a single dollar input, so it is the same in every band.
Where each default comes from
| Input | Default | Basis |
|---|---|---|
| Annual study volume | 250,000 | A placeholder. Replace it with your own volume; nothing below is meaningful until you do. |
| Share billed as complete multi-element exams | 20% | Assumption. The share of your volume billed under a complete code whose documentation requirements can be missed (CPT 76700 complete abdominal ultrasound is the example used here). Your coding mix will differ. |
| Current first-pass denial rate | 15% | Published. Premier Inc. survey, March 2024: 516 hospitals, 2022 claims, about 15% of claims to private payers initially denied. Premier Inc. |
| Share of denials attributable to indication / documentation gaps | 25% | Assumption. We know of no published radiology-specific figure we are prepared to cite. Your denial reports, broken down by reason code, are the right source; enter what they show. |
| Share of complete-exam reports missing a required element | 3% | Assumption. Our backtest of 1,291 signed reports found 3.7% with a required element missing; the default is rounded down. One practice's reports, so treat it as an indication, not a benchmark. |
| Professional-component loss per downcoded complete exam | $12 | Assumption, chosen to sit inside the CPT 76700 vs 76705 professional-component gap of roughly $11-16; see the note below. |
| Cost to rework one denied claim | $25 / $43.84 / $57.23 | Published. MGMA estimate as cited by AAFP (2015); Premier, 2022 claims; Premier follow-up, 2023 claims. Linked in the sources table. |
The CPT 76700 vs 76705 note
A complete abdominal ultrasound (CPT 76700) is billable only when the report documents every element the code requires; the elements are set out in the ACR practice parameters and the CPT descriptor. If one is missing, the exam is billed as a limited study (CPT 76705). The professional component differs by roughly $11-16 and the global fee by roughly $28-55, depending on year and locality. These ranges are approximate, verify against the current Medicare Physician Fee Schedule for your locality: Medicare Physician Fee Schedule look-up. A practice that bills globally should enter the global gap rather than the default.
What this estimate leaves out
- Denials that are never overturned. Premier found 54.3% of denied claims were eventually overturned. Revenue written off on the rest is not counted here; only the rework cost is, which keeps the estimate conservative.
- Other complete-versus-limited pairs. Only the abdominal ultrasound pair is used as the worked example. If your mix includes other codes with documentation-dependent levels, the share and dollar inputs apply to whatever figures you enter, but the page does not supply them.
- Staff time on addenda and clarification requests that never become a denial.
- Anything about MyRadAgent. The page estimates the size of the problem from your inputs. It does not model how much of it any tool would remove.
Sources
| Citation | What it gives | Link |
|---|---|---|
| Premier Inc., March 2024 survey of 516 hospitals (2022 claims) | about 15% of claims to private payers initially denied; $43.84 average spent per claim fighting denials; 54.3% of denials eventually overturned | View the source |
| Premier Inc. follow-up on 2023 claims, reported by Healthcare Innovation | $57.23 average spent per claim fighting denials | View the source |
| MGMA estimate, as cited by AAFP Family Practice Management (March 2015) | about $25 to rework one denied claim | View the source |
| CMS Medicare Physician Fee Schedule look-up | CPT 76700 vs 76705: professional component differs by roughly $11-16, global by roughly $28-55, depending on year and locality (approximate, verify against the current Medicare Physician Fee Schedule for your locality) | View the source |
Links verified 2026-11-01. Every number on this page is either one of these published figures, our labelled assumption, or your own input.
Related
- Pricing — what MyRadAgent costs, to set against your own estimate
- All radiology calculators
Related calculators
MyRadAgent’s radiology AI drafts the full report and applies this guideline automatically, with the category and follow-up in the impression. Free trial runs in the browser: 100 reports over 14 days, no card, no IT ticket.
Try it on a report → or try it here first — no signup ↑BI-RADS, TI-RADS, LI-RADS, PI-RADS, O-RADS, Lung-RADS, CAD-RADS, NI-RADS and ACR are registered trademarks of the American College of Radiology. They are used here only to identify the published systems these tools implement; MyRadAgent is not affiliated with, sponsored by or endorsed by the ACR.