Healthcare & MedTech// definition

The letter of medical necessity: the 4 things it has to establish

In short

A letter of medical necessity is an evidentiary document that has to establish 4 things: the diagnosis and how it was reached, the functional impact on this patient, what has already been tried and why it failed, and why this service is the right next step now. Anything the reviewer cannot locate is treated as absent.

Key takeaways

  • The letter is an argument with a burden of proof, not a covering note attached to a request form.
  • Four assertions have to land: diagnosis, functional impact, failed alternatives, and why this service now.
  • The failed-alternatives history is the assertion most often missing and most often denied on.
  • Every claim should trace to something already in the record, with dates, or it reads as advocacy.
  • No letter rescues a request the payer's own criteria document already rules out on other grounds.

A letter of medical necessity carries the clinical argument for a service a payer will not approve on the request form alone. It is evidentiary, not rhetorical: the reviewer checks whether 4 assertions are present and supported, and a letter making 3 of them beautifully loses to a plainer one that makes all 4.

Treat the reader as an assessor with a checklist, not a colleague to be convinced. They match what you wrote against criteria fixed before your request arrived, and their default is not approval.

The 4 assertions, and where each one's evidence lives

AssertionWhat it must stateWhere the evidence comes fromIf it is missing
Diagnosis, and how it was establishedThe condition, its severity, the findings behind itExamination findings, imaging and results, datedReturned for information
Functional impact on this patientWhat the patient cannot do, in specifics not adjectivesHistory, functional scores, work and activity limits over timeReads as preference, not need
What was tried and failedEach prior treatment, duration, dose or intensity, outcomeMedication history, therapy notes, prior procedure recordsThe most common denial: alternatives not exhausted
Why this service, and why nowThe link from these findings to this interventionClinical reasoning in the plan; criteria metApproved for something cheaper, or deferred
Each assertion needs a source in the record, not a sentence

Notice what the third column is doing. Every assertion must trace to material already in the chart, with dates, because claims that appear nowhere else in the record read as advocacy written for the occasion. Where the notes are thin the fix is upstream, in the structure that keeps findings, reasoning and plan separable — the 4 sections of a SOAP note.

The paragraph that decides most requests

One part is worth writing carefully above all others: the failed-alternatives history. Payer criteria for expensive services are stepped — the more conservative option must have been tried, for a stated period, at a stated intensity, without adequate benefit. Saying conservative management was unsuccessful states a conclusion. Naming the treatment, the dates, the dose or session count, and what did not improve supplies evidence.

The same paragraph must handle steps never tried, and this is where letters go quiet. If one was skipped as contraindicated, or because it previously caused an adverse reaction, say so and give the reason. Silence on a required step reads as an omission and is denied as one; a stated contraindication is an argument the reviewer can act on.

Conservative management failed is a conclusion. Twelve weeks of supervised therapy from March to June with no change in the recorded functional score is evidence. Reviewers approve the second.

The letter is answering a document you can read first

A letter written in a vacuum is guesswork. The criteria the reviewer applies are published for most non-trivial services, and reading them first turns the letter from an essay into a point-by-point response — which is why the payer document that decides your authorisation first comes before drafting. Where the criteria name a required duration of prior therapy, state it; where they name a test, say when it was done and what it showed.

2 different documents share this name

The phrase also names a far shorter document in a tax and benefits context: a clinician's written statement that an expense treats a specific condition, used to substantiate a claim against a health spending account. The IRS publication on medical and dental expenses, Publication 502, sets out categories where a physician's statement makes an expense eligible — treatment at a health institute counts only where a physician prescribes it and states it is necessary. That letter is a paragraph and a signature. Using it for the authorisation job gets a request denied for lack of clinical information.

Who signs it, and what a drafting tool may do

The treating clinician signs, because the letter asserts clinical judgement. Almost everything else is assembly: pulling the medication and therapy history with dates, retrieving prior results, checking the current criteria, and ordering the argument as the criteria are written. That is worth automating — the shape of work we scope as AI agents and automation, under the constraints in AI agents in production.

Two boundaries hold however the draft is produced. A tool may surface what the record contains and flag what it lacks; it may not supply an assertion the record does not support. And no letter rescues a request that failed for a non-clinical reason — a submission that never reached the payer is a different problem, and a denial pointing at eligibility or coding rather than necessity says not to rewrite the argument at all, as reason codes and remark codes explains. The workflow around it sits in revenue cycle, prior authorisation and denials, inside our healthcare operations work.

Frequently asked questions

Short answers to the follow-ups this page tends to raise.

What is a letter of medical necessity?

It is a signed clinical document establishing why a specific patient needs a specific service, used where a payer will not decide from the request form alone. It must establish 4 things: the diagnosis and how it was reached, the functional impact, what was already tried and failed, and why this service now.

Who has to write and sign it?

The treating clinician signs it, because it asserts clinical judgement about this patient. Staff or software can assemble everything else — medication and therapy history with dates, prior results, the current criteria — but the reasoning and the signature belong to the clinician responsible for the decision.

Why do letters get denied when the service is clearly needed?

Most often because the failed-alternatives history is missing or vague. Payer criteria for costly services are stepped, and the reviewer checks that a more conservative option was genuinely tried, for long enough, without adequate benefit. Stating that conservative treatment failed, without naming it, the dates and the outcome, gives them nothing to verify.

Should the letter quote the payer's policy criteria?

Answer the criteria point by point, but describe them rather than reciting a clause reference from memory. Name the document and its effective date from the copy in front of you, and address each requirement in order. An outdated reference undermines the letter more than omitting it would.

  • medical necessity
  • prior authorisation
  • revenue cycle
  • documentation
// shipped work

The work behind this page

Builds from our portfolio that this page draws on.

Read next

Working on something in this space?

Tell us where you are in a sentence or two. We'll tell you honestly whether we're the right team, and what a sensible first slice of the work looks like.

Start the conversation