Editorial, Corrections & Fact-Checking Policy
Every statistic on this site either carries a link to the source it came from, or is published as an explicit blank. We do not relay industry figures we cannot trace to a primary source, even when every competitor publishes them. When we get something wrong, we correct it on the page and log it here with a date. The build fails if an unsourced benchmark attribution ships.
How we source a number
A figure may appear on this site only if it satisfies one of four tests. There is no fifth route, and "it is widely reported" is not one of them.
- It is cited to a checkable primary source. The URL is fetched at the time of writing and the exact figure confirmed on the page — not recalled from memory, and not carried over because the site cited it before.
- It is an artifact we produced. A calculation we ran, with the inputs shown.
- It is a first-party operating fact. Our pricing, our contract terms, our process. These are labelled as ours, never as an industry standard.
- It was supplied by the business owner and recorded in writing, so the claim has a traceable origin rather than living in one person's memory.
Three rules govern how a sourced number is then used. The number we print is the number in the source — we do not round, derive, combine or reframe a figure into a new claim. The population travels with the figure: a hospital denial rate is not a physician-practice denial rate, and an ACA-Marketplace rate is neither. And a definition is not a benchmark — HFMA's MAP Keys standardise how days in A/R is calculated, which is genuinely useful and genuinely different from publishing a target value.
When no source exists, we publish the blank
This is the part that distinguishes a real sourcing policy from a stated one. Where an industry figure is universally quoted but has no traceable origin, the honest output is to say so — and the absence is often more useful than the number would have been.
Our benchmarks hub publishes several of these blanks explicitly. There is no free primary source for a physician-practice clean claim rate target, none for the share of A/R over 90 days, and no authoritative ranking of denial codes by frequency — X12 publishes no usage statistics and CMS publishes none, so every "top 10 denial codes" ranking in this market is vendor marketing with no methodology behind it. We publish those gaps as gaps.
What we will not publish
- Case studies. This site publishes none. Client results are confidential and unverifiable to a reader, which makes a published case study an assertion the reader has no way to test.
- Testimonials, client logos, client counts or star ratings — including aggregate rating markup, which would place an unverifiable claim into search results.
- Invented statistics or invented sources, including a real organisation's name attached to a figure it does not publish. The corrections log below shows this failure mode is real rather than hypothetical.
- Achieved-performance claims about our own service stated as a rate — "98% clean claims", "we increase collections by 30%". We state what we do and what we report against, not what we have achieved on accounts you cannot inspect.
- Fabricated author personas or credentials. Bylines describe who actually reviewed the page.
How the build enforces this
A policy that depends on everyone remembering it will fail. Ours runs as an automated gate: the site cannot be deployed unless a suite of data-integrity assertions passes. Among them:
- The attribution guard. A benchmark number attributed to a named body — MGMA, HFMA, AAPC, AHIMA, HBMA, AMA, KFF, CAQH and others — in bare prose, with no inline citation, fails the build. The rule it encodes is the one this page states: name a body, link it.
- Source resolution. Every registered figure must point to a real registered source, and every figure carries its population, its denominator and whether it is a measurement or a target. A dangling reference fails the build.
- Licensing guards. Republished X12 code description text fails the build, as does presenting a deactivated denial code as live.
- Arithmetic guards. Every figure in our illustrative sample deliverables is asserted, including an A/R roll-forward that must reconcile. The benefits-load error described in the corrections log now has a dedicated assertion so it cannot be reintroduced.
The attribution guard is deliberately built as a ratchet rather than a wall. When it was introduced it found a backlog far larger than the handful of pages we knew about, and failing the entire site would simply have meant switching the guard off. Instead the known backlog is recorded in a file, the build fails the moment a new page joins it, and a page removed from that list can never quietly return. 89 pages currently remain in that backlog, each awaiting the same treatment: fetch the source, or state the blank. That number is read directly from the enforcement file when this page is built, so it cannot drift from reality, and by construction it can only go down.
Illustrative examples and sample deliverables
Because we publish no case studies, we publish worked examples instead — a sample monthly report, a sample denial analysis, a sample A/R aging plan. Each is built on the same hypothetical practice so they can be read end to end, and each carries this disclaimer visibly, next to the first figure rather than in a footnote: "This is an illustrative example demonstrating MedPrecision's reporting methodology. It is not presented as the result of an actual client engagement." Every practice-level number inside them is hypothetical; every benchmark comparator beside it is real and cited. The samples deliberately show problems rather than a flawless month, including one where over-coding reduces revenue.
Third-party data and licensing
Some of the most-wanted data in this field is not ours to republish. X12 prohibits posting its copyrighted CARC and RARC descriptions on a third-party public website and prohibits paraphrasing them separately, so our denial code reference carries original commentary and deep-links to x12.org for the official text. MGMA's benchmark values sit inside licensed products and are not quoted here. Where a licensing constraint stops us publishing something, we say so on the page rather than quietly omitting it.
Who writes and reviews this content
Articles are written by our revenue cycle operations team and reviewed for coding accuracy before publication. Bylines on this site are collective team bylines rather than individual author personas, and they say so plainly — we would rather show an honest collective byline than invent a person. Pages carrying clinical coding detail are reviewed against the primary payer or CMS source for the current year, and every page shows its last-updated date.
Where we are not yet where we want to be: we do not currently publish a named, individually credentialed reviewer against each technical billing article. That is a real gap in this policy, and naming a person we do not have would be exactly the kind of claim the rest of this page exists to prevent. It is stated here rather than omitted.
Corrections log
Material corrections are listed here, newest first, with the date and the reason. A corrections policy with an empty log is a statement of intent; this one is a record.
Removed five fabricated research citations from four service pages
A claims page cited an "AMA 2024 National Health Insurer Report Card" — the AMA discontinued that report after 2013, so no 2024 edition exists. The same page cited an "HFMA 2024 Claims Management survey" and a claims-processing page cited an "AAPC 2024 industry survey"; neither publication could be located. A company page cited Black Book Research and Medical Economics for a claim that certified staff produce 8–12% higher net collection rates. KFF's ACA-Marketplace denial rate was presented as a Medicare fee-for-service figure, which is a different population entirely. All removed and replaced with MGMA's published first-submission denial rate and AAFP's published targets.
Removed an unverifiable urgent-care throughput benchmark
The urgent care page carried a "UCAOA benchmark of 2.7 patients per provider hour" and a monthly revenue range derived from it. The figure could not be traced to any UCAOA publication. The benchmark and both derived figures were removed rather than re-attributed.
Removed five unsourced cost figures from the credentialing timeline page
Per-day revenue-loss figures ("$1,500–$3,000 per day in billable services") and the annual exposure ranges derived from them had no source. Replaced with the structurally true statement — your exposure is your own daily billable charges multiplied by the days you are uncredentialed — plus an explicit note that no credible universal per-day figure exists.
Retracted the "98% clean claim rate" and "97–99% net collection rate" benchmarks
Both figures, and an "A/R over 90 days under 10%" target, were attributed on this site to HFMA and MGMA. They trace instead to a single article on hfma.org carrying a "Sponsored by Conifer Health Solutions" label, which itself attributes its clean-claim figure to a trade publication. HFMA's MAP Keys publish metric definitions and equations and no target values at all; MGMA's benchmark values sit inside licensed products. Re-sourced to AAFP, which publishes days in A/R, adjusted collection rate and denial rate targets openly — and publishes no clean claim rate figure, so we now publish that as a blank.
Corrected a calculation error in the billing cost calculator
The model applied the BLS employee-benefits share to salary. BLS reports benefits as 30.1% of total compensation, which is a 43.06% load on wages — so the calculator understated in-house employer cost by roughly a third. Fixed, and a build-time assertion now fails if the wrong denominator is reintroduced. An unsourced "outsourced billing typically increases collections by 5–15%" claim was removed at the same time; a dedicated search found no source stating a methodology.
Removed 128 verbatim X12 code descriptions across 20 pages
X12 prohibits republication of its copyrighted code descriptions on a third-party public website, and prohibits paraphrasing separately. All 128 strings were replaced with original wording, and each code now deep-links to x12.org, which publishes the official description free. Denial code CO-15 was also corrected — it was deactivated on 2018-05-01 and was being presented as live.
Corrected two clinical billing errors found in a source review
The physical therapy page carried an outdated KX modifier threshold and described CPT 97530 as untimed; it is a timed code. The gastroenterology page stated that a screening colonoscopy converted to diagnostic carries no Medicare cost-share — the deductible is waived but 15% coinsurance applies to both the provider and the facility fee, per medicare.gov.
Removed 59 case studies and 243 statistics that could not be traced to a source
A sitewide audit found published client case studies and performance statistics with no verifiable origin. All were removed rather than re-worded. This site publishes no case studies. Where a worked example is useful, we publish a clearly labelled illustrative one built on a hypothetical practice, and every benchmark comparator inside it carries a real citation.
How to report an error
If you find a figure on this site you believe is wrong, unsourced, or attributed to a body that does not publish it, email [email protected] with the page URL and the figure. We would rather hear it from you than leave it up. Substantiated corrections are made to the page and added to the log above.
Common questions
Why does this site refuse to publish a clean claim rate benchmark?
Because no professional body publishes one. The 98% figure quoted across this industry traces to a vendor-sponsored article, which attributes it in turn to a trade publication, with no methodology stated at any link in the chain. AAFP publishes targets for days in A/R, adjusted collection rate and denial rate but publishes no clean claim rate figure. HFMA's MAP Keys define the metric without attaching a value. We report clean claim rate as a trend against your own baseline instead.
Do you use AI to write this content?
Drafting assistance is used, and it changes nothing about the standard. Every figure is checked against its primary source by a person before publication, the source URL is fetched rather than recalled, and the build blocks a benchmark attributed to a body in bare prose. The failure mode of generated text is a confident, plausible, unsourceable number — which is precisely the class this policy exists to catch, and precisely what the corrections log below records us catching.
How do you decide whether a source is authoritative?
Primary sources rank first: federal regulators and their own publications (CMS, HHS, OIG, the Federal Register, BLS), and a payer for facts about its own process. Professional societies rank next, and only for what they actually publish — a definition is not a benchmark value. Vendor research is used when the methodology and sample are disclosed, and is always labelled as vendor research. Sponsored content is not a source, and the corrections log records what happened the last time this site treated it as one.
Do you correct pages silently?
Material corrections — a wrong figure, a misattributed source, a retracted claim — are logged on this page with a date and the reason. Routine copy edits, formatting and link maintenance are not logged. If a page you cited has changed in a way that affects your use of it, contact us and we will tell you exactly what changed and when.
Can journalists and researchers cite your data?
Yes. Our benchmarks hub carries a citation line, downloadable CSVs with a source URL on every row, and charts built from the underlying data. Please cite the primary source alongside us wherever we are relaying one — the whole point of publishing the URL is that you should not have to take our word for it.
Related: our 2026 medical billing benchmarks, where every figure carries its source, population and denominator; the days in A/R formula and benchmark; our operating framework; and our HIPAA compliance program. The reporting standard described here is the one applied in our revenue cycle analytics service.
Want the same standard applied to your monthly reporting — every KPI against a source you can open? Start with a free billing audit.
Get a Free Billing Audit arrow_forwardLast reviewed: August 5, 2026
This policy governs content published on medprecisionbilling.com. It is not legal, tax or clinical advice. Figures cited from third parties are the responsibility of the publishing body; where we relay one, we link it so you can check it yourself.