CGRP Prior Authorization: What Your Diary Has to Show

In March 2024 the American Headache Society published a position statement saying CGRP-targeting therapies are a first-line option for migraine prevention, and that starting one should not require trial and failure of older preventive drugs. Coverage rules did not all move with it.

So there is a gap. On one side, the specialty society that writes the clinical guidance. On the other, a plan document that may still say two classes, eight weeks each, documented. Your neurologist writes the prescription. What decides whether it gets filled is a form, and what fills the form is your diary.

Two limits on what follows. This is a US-specific piece of admin — readers outside the States are dealing with a different set of gatekeepers, and in the UK the equivalent hurdle is a NICE threshold rather than an insurer’s policy PDF. And coverage criteria move independently of clinical guidance, on each plan’s own schedule. Nothing here is a substitute for the document your own plan is deciding against, which is why the first section is about finding it. Treat any general statement about “plans” on this page as a prompt to go and check yours.

Read the document the decision is made against

Prior authorization is not decided by a person weighing your case in the abstract. It is decided against a written policy, and most insurers publish theirs. Search your plan’s name together with the drug name and the phrase “medical policy” or “clinical criteria”, and you will usually surface a PDF listing the exact conditions for approval.

Read it before the appointment, not after the denial. It tells you three things you otherwise have to guess: the migraine-day threshold, which prior drugs count as an adequate trial, and — the part people miss — what renewal will require six months from now.

The four things the form asks

Plans differ, and yours is the one that matters. But prior-authorization criteria for preventive migraine treatment tend to circle the same four items, and each maps onto something you either have in a diary or don’t.

A diagnosis with a code. That comes from the clinician, not from you. What helps is that their impression matches your record — the ICHD-3 criteria are what the diagnosis is written against, and a diary showing duration, one-sided throbbing pain, nausea and light sensitivity is what makes the diagnosis easy to write.

A migraine-day count over a defined baseline period. Usually a month, sometimes three. This is the number everything turns on, and it is the number most people get wrong, because a migraine day is a calendar day carrying a qualifying headache rather than an attack. One long attack across a weekend is two or three of them.

Prior preventive trials. Not “I tried topiramate once.” A plan wants the drug, the dose, the start and stop dates, and why it stopped — no benefit, or an adverse effect you couldn’t live with. Eight weeks at an adequate dose is the usual bar for calling a trial adequate, and stopping at week three because of fog does not count as a failure of efficacy; it counts as intolerance, which is a different box on the form and often an easier one.

A disability measure. MIDAS or HIT-6, both validated questionnaires, both quick. Some policies require a score, and even when they don’t it converts “bad migraines” into a number an assessor can act on. Migraine Journal includes both and keeps the dated results with your entries, which matters because a score from eighteen months ago is not evidence about now.

The renewal trap

Approval is not the end of it. Continuation typically requires showing the drug worked — commonly a meaningful reduction in monthly migraine days against your pre-treatment baseline, assessed after three to six months.

Here is how people lose that. The prescription finally gets approved after weeks of back and forth. Relief arrives. The first injection goes in. Nobody wrote down what the month before looked like, in detail, day by day.

Six months later the plan asks for evidence of reduction and there is nothing to reduce from. Not because the drug didn’t work — because the comparison doesn’t exist. “It’s a lot better” is not a percentage.

Record the baseline month before the first dose, and keep recording at the same level of detail afterwards. Same fields, same discipline, so the two periods are comparable. If your acute medication use dropped from twelve days a month to four, that belongs in the record too; it’s often the most persuasive single line in an appeal, and it’s the same number that keeps you clear of the medication-overuse threshold.

The eight-week and twelve-week checkpoints are worth understanding for their own sake as well, since judging whether a preventive is working is a different question from proving it to an insurer, and the honest answer to the first one is what you want going into the second.

If it’s denied

Denials at the first pass are common enough that clinic staff treat them as a step rather than an outcome. Two things help.

The first is a peer-to-peer review, where your prescriber talks to a clinician at the plan. Your job before that call is to make sure the office has your actual numbers — migraine days, acute medication days, prior drugs with dates and reasons for stopping — rather than a note that says “frequent headaches, tried several things.”

The second is that a written appeal takes attachments. A one-page summary showing the count per month, with the day-level detail behind it, does more work than a paragraph of description. Print it once and hand it in; don’t send forty pages of raw log entries and expect someone to total them for you. There’s more on assembling that in what to bring to a neurologist appointment, and the same packet serves both purposes.

Anything about whether a specific drug is right for you belongs with your prescriber. What’s on your side of the line is the record, and the record is the part that decides how much of the argument you have to make twice.