4.4 Reviewing Documentation to Support the Level of Codes Billed
Key Takeaways
- Documentation must be legible, complete, dated, and authenticated by the rendering provider; an unsigned or unsigned-by-the-wrong-person note cannot support any claim.
- The billing rule is that if a service is not documented it was not performed, so the record and not the claim is the object of every payer review.
- Validating an office E/M level means testing the documented medical decision making against the two-of-three rule, or testing the documented total time on the date of the encounter.
- Copy-forward and cloned documentation is a primary audit target because identical narrative across encounters cannot establish that the work was performed at each visit.
- Auditors who find a high error rate in a sample may extrapolate the overpayment across the whole universe of claims, turning a small sample finding into a large repayment demand.
4.4 Reviewing Documentation to Support the Level of Codes Billed
The Detailed Test Plan's second claims-submission task is "Review medical record documentation supporting level of codes billed." This is the internal-audit function of the Insurance and Coding Specialist role, and it is what stands between a practice and a repayment demand.
The governing principle has one sentence: if it is not documented, it was not performed. A payer reviewing a claim does not review the claim. It requests the record and asks whether the record independently supports what the claim asserted.
1. The Four Threshold Standards
Before any level-of-service question is reached, the note must clear four gates. Failing any one of them makes the encounter non-billable regardless of how good the clinical care was.
| Standard | Requirement | Common Failure |
|---|---|---|
| Legible | A reviewer other than the author must be able to read it | Handwritten notes that the auditor cannot decipher are treated as absent |
| Complete | Reason for the encounter, findings, assessment, plan, and the services rendered | A plan with no assessment; an order with no indication |
| Dated | Each entry carries the date of service | Late entries not identified as late |
| Authenticated | Signed by the rendering provider, handwritten or by compliant electronic signature | "Signed by" a scribe or an unsigned note; a signature stamp, which Medicare does not accept |
Signature specifics that appear in audit findings. An electronic signature must be attributable and locked (for example, "Electronically signed by Jane Rivera, MD, 03/14/2026 14:22"). "Dictated but not read" does not authenticate a note. If a signature is illegible or missing, some payers will accept a signature attestation statement completed by the provider, but an attestation cannot be created for a note the provider did not author.
2. Validating an Office E/M Level
E/M upcoding is the single most audited area in physician billing and appears every year on the OIG Work Plan. Under the guidelines in force for office and outpatient visits, level selection rests on either medical decision making or total time — never on the volume of history and exam documented.
Path A: Validating by Medical Decision Making
Test the note against all three MDM elements and confirm that at least two of the three reach the level billed.
| Element | What the auditor looks for in the note |
|---|---|
| Number and complexity of problems addressed | Are the problems addressed at this visit — evaluated, treated, or managed — or merely listed in a history? A stable chronic condition mentioned but not managed does not count. |
| Amount and complexity of data reviewed and analyzed | Are the tests ordered or reviewed named? Is independent interpretation documented? Is discussion with an external provider documented with whom and about what? |
| Risk of complications, morbidity, or mortality of management | Is prescription drug management documented (drug, decision, monitoring)? Is a surgical decision and its risk factors documented? |
The recurring finding. A note lists eight chronic conditions in the history, the assessment addresses one, and the visit is billed at a moderate level on "two or more stable chronic illnesses." The auditor downcodes it, because the guideline counts problems addressed, not problems present.
Path B: Validating by Total Time
If time is the basis, the note must state the total time spent on the date of the encounter by the physician or qualified health professional. Countable activities include record review, obtaining history, the examination, counseling, ordering, care coordination, and documentation — all on that date.
| Documented statement | Supports a time-based level? |
|---|---|
| "35 minutes total time spent on the date of the encounter, including chart review, examination, counseling, and documentation." | Yes |
| "Spent 35 minutes with the patient." | Weak — face-to-face only; does not capture total time |
| "Greater than 50% of the visit spent in counseling." | No — that is the retired pre-2021 standard for office visits |
| No time statement at all | No — the level must then be validated on MDM |
3. Cloned and Copy-Forward Documentation
Electronic records make it trivial to carry a prior note forward. Auditors specifically hunt for the result.
Indicators of cloning: identical review-of-systems language across every patient in a provider's panel; a physical examination narrative unchanged across ten consecutive visits for a patient whose condition demonstrably changed; a "current medications" list that has not moved despite documented changes; assessments that repeat verbatim while the plan changes.
Why it is fatal to a claim: cloned text cannot establish that the work was performed at this encounter. When the auditor cannot distinguish today's work from last month's, the level collapses to what the unique, encounter-specific text supports — typically one or two levels lower.
The compliant use of copy-forward is carrying forward genuinely static information (surgical history, allergies, family history) while writing the history of present illness, the examination findings, and the assessment and plan fresh at each encounter.
4. The Internal Audit Workflow
| Step | Action | Practical Standard |
|---|---|---|
| 1. Scope | Select the codes and providers to review | Start with the OIG Work Plan's flagged areas and the practice's own outlier codes |
| 2. Sample | Pull a defined sample per provider | A baseline audit of 5-10 charts per provider establishes the error rate; routine audits follow quarterly or annually |
| 3. Blind review | Recode from the documentation alone | The auditor must not see the code that was billed before coding the note |
| 4. Compare | Match the audit code to the billed code | Record every variance, both over- and under-coded |
| 5. Quantify | Calculate the error rate and dollar impact | Error rate = variances ÷ charts reviewed |
| 6. Educate | Provider-specific feedback with the source guideline | Feedback without the citation does not change behavior |
| 7. Correct | Refund identified overpayments | The ACA 60-day rule starts when the overpayment is identified and quantified |
| 8. Re-audit | Verify the correction took | A focused re-audit on the same codes within one to two quarters |
Why the Sample Size Is Not the Exposure
The financial danger in a payer audit is extrapolation. When a contractor finds a high error rate in a probe sample, it may project the overpayment rate across the entire universe of that provider's claims for the review period.
Illustration. A contractor reviews 30 claims and finds $1,800 in overpayments — a 60% error rate. The provider billed 9,000 comparable claims in the period. Rather than demanding $1,800, the contractor extrapolates the per-claim overpayment across the universe: $1,800 ÷ 30 = $60 per claim × 9,000 claims = a $540,000 repayment demand from a 30-chart review.
This is the reason documentation review is a revenue function and not a clerical one. The reviewer who downcodes one visit today is preventing an extrapolated demand later, which is also why an internal audit that only ever finds under-coding is not a credible audit.
A provider's office note lists eight chronic conditions in the history section, but the assessment and plan address only the patient's hypertension. The visit was billed at a moderate-complexity level based on 'two or more stable chronic illnesses.' What should the auditor conclude?
A physician's note documents 'Greater than 50% of the visit was spent in counseling' as the basis for an office visit level. Does this support time-based level selection?
A Medicare contractor reviews 30 claims, identifies $1,800 in overpayments, and the provider billed 9,000 comparable claims during the review period. Why is the provider's exposure far greater than $1,800?