Letters of Medical Necessity

A letter of medical necessity is the document that converts a clinical judgment into something a payer can approve. It is read by a reviewer who may not be a geneticist, often under time pressure, against a written policy. Writing for that reader rather than for a colleague is what separates letters that succeed from letters that are clinically excellent and still denied.

• Write to the payer's own criteria. Obtain the medical policy for the test, identify the criterion the patient meets, and quote it. A reviewer whose job is to check a policy box will approve a letter that shows the box is checked, and will deny one that argues around it.
• Structure that works: patient identification and the specific test requested with its CPT code; the clinical presentation; the family history in specific detail; the guideline criteria met, cited by name and section; why alternative or stepwise testing is inadequate; and the management changes that will follow each possible result.
• The management-change paragraph is the one that decides it. Payers approve tests that alter care. State concretely what changes: surveillance beginning at a given age, a specific surgical decision, a drug started or avoided, cascade testing of named at-risk relatives, or the ending of a diagnostic odyssey with its own costs.
• Specificity beats eloquence. "Strong family history of cancer" is weak. "Mother diagnosed with breast cancer at 41, maternal aunt with ovarian cancer at 55, maternal grandmother with breast cancer at 48" is what meets criteria.
• Address the stated denial reason directly when appealing. If the denial says investigational, the letter must supply evidence of clinical validity and utility, with citations. If it says criteria not met, the letter must show which criterion is met and where it is documented.
• Cite guidelines by name, version, and section: NCCN, ACMG, ACOG, and specialty society statements. Attaching or quoting the specific line is stronger than naming the organization.
• Cost arguments are legitimate. Where a single test replaces a sequence of more expensive investigations, saying so with figures addresses the reviewer's actual mandate.
• Keep it to one or two pages. A ten-page letter is skimmed. The reviewer needs the criterion, the evidence it is met, and the consequence.
• Who signs it matters for some payers, which require a physician signature. Genetic counselors frequently draft the letter and a supervising or ordering physician signs, and in states with licensure and appropriate payer recognition counselors may sign directly.

• A child with unexplained global developmental delay is denied exome sequencing as investigational. The letter cites the ACMG statement recommending exome or genome as a first-tier test for this indication, gives the published diagnostic yield, lists the prior non-diagnostic workup with its costs, and states that a diagnosis would end further testing and direct surveillance for known complications. Approved on appeal.
• A woman with a paternal family history of breast cancer is denied on the basis that the history is on the father's side. The letter states plainly that BRCA1 and BRCA2 are autosomally inherited and transmitted equally through fathers, quotes the NCCN criterion that does not distinguish lineage, and lists the affected paternal relatives with ages at diagnosis. Approved.
• A patient's carrier testing is denied because the payer requires stepwise testing. The letter documents that the stepwise option was performed and was uninformative, attaches the report, and requests the expanded panel as the next indicated step rather than as a first-line request.
• A metabolic patient needs an enzyme assay unavailable in network. The letter requests a network adequacy exception, names the two laboratories nationally performing the assay, and documents that neither is in network, converting the request from a coverage question into an access question the payer must answer.

• Do not write a letter without reading the payer's policy. Arguing from general clinical reasoning against a specific written criterion is the most common way a strong case fails.
• Do not summarize the family history. Ages at diagnosis and specific relationships are what criteria are written in terms of.
• Do not omit the management change. A test that does not alter care is, from the payer's perspective, correctly denied, so if care will change the letter must say how.
• Do not ignore the stated denial reason. Resubmitting the original argument unchanged reliably produces the same result.
• Do not exceed two pages. Length reduces the chance the decisive paragraph is read.