Medication Overuse Headache: A Provider’s Guide to Deprescribing Acute Headache Treatment

The patient in front of you has had migraine for fifteen years, is now describing headache on most days of the month, and has failed two preventives. The medication list shows a combination analgesic filled every month without interruption. Before you reach for a third preventive, it is worth asking whether the acute treatment has become part of the problem. Medication overuse headache is one of the few causes of chronic daily headache that is fully reversible, and it is also one of the most commonly missed.

Why is this diagnosis still being missed?

Partly because the presentation is unremarkable. Patients describe their usual headache, only more often and somewhat worse, with a physical examination that is typically non-focal. There is no confirmatory test. Nothing about the character or location of the pain establishes the diagnosis, and frequency is what makes it.^[1]

Partly because of how the drug histories are taken. Around 90 percent of patients with medication overuse headache are using more than one product for relief, and over-the-counter agents rarely make it onto the reconciled medication list unless you ask directly.^[1] A patient who reports taking a prescription combination analgesic on eight days a month may also be taking an over-the-counter combination on another eight.

And partly because the diagnosis is often hiding inside a preventive failure. Responsiveness to preventive treatment is reduced while an acute medication is being overused and improves after discontinuation.^[1] The two-preventive failure in the vignette above may be a consequence of the overuse rather than evidence of refractory disease.

The scale of it argues for looking harder. More than half of people with headache on 15 or more days a month meet criteria for medication overuse headache, and the population prevalence sits around 1 to 2 percent.^[2][3] It is also worth knowing that a plain explanation delivered in primary care is often enough on its own. The classification itself notes that simple advice about the causes and consequences can be provided with success in primary care, and that an explanatory brochure is sometimes all that is needed.^[2]

What exactly are the diagnostic thresholds?

The ICHD-3 criteria for 8.2 Medication-overuse headache are three lines: headache on 15 or more days a month in a patient with a pre-existing headache disorder; regular overuse for more than three months of one or more drugs that can be taken for acute or symptomatic treatment of headache; and not better accounted for by another ICHD-3 diagnosis.^[2]

The days-per-month threshold that constitutes overuse depends on the agent:^[1][2]

• 10 or more days a month for ergotamine, triptans, opioids, and combination analgesics.
• 15 or more days a month for acetaminophen, aspirin, and other NSAIDs taken as single agents.
• 10 or more days a month in aggregate when multiple classes are used without any single one crossing its own threshold. That is coded separately as 8.2.6, and it is easy to miss.^[2]

A useful coding note: patients who overuse a combination analgesic are coded as combination-analgesic-overuse headache, not according to each constituent.^[4] So a butalbital, acetaminophen and caffeine product is judged against the 10-day threshold, not the 15-day one that acetaminophen alone would attract. That distinction changes who qualifies.

Patients meeting criteria for both chronic migraine and medication overuse headache should carry both diagnoses.^[2] The classification also acknowledges that the specific day counts rest on expert opinion rather than formal evidence, which is worth remembering when a patient sits just under a threshold and is clearly deteriorating.^[2]

Which drug classes carry the most risk?

Not all acute treatments carry equal risk, and the ordering is fairly consistent across reviews. Highest incidence sits with opioids, butalbital-containing combination analgesics, and acetaminophen-aspirin-caffeine combinations, followed by triptans, then NSAIDs.^[1] Progression to medication overuse headache is also substance-specific in speed, occurring faster with triptans, opioids and combination analgesics than with simple analgesics.^[1]

The American Academy of Neurology addressed two of these directly in its Choosing Wisely list, advising against opioid or butalbital treatment for migraine except as a last resort, on the grounds that migraine-specific alternatives exist and frequent use of these agents can worsen headache over time.^[5] That recommendation is now more than a decade old, and butalbital combinations remain in wide circulation, frequently continued by refill inertia rather than by an active decision.

Two comorbidity patterns are worth screening for at the same time. Dependence-type behaviour is common, with up to half of patients showing tolerance or an inability to control use, and it presents more often with opioid and triptan overuse than with aspirin or ibuprofen.^[1] Mood and anxiety disorders are also frequent travelling companions, with reported rates around 40 percent for depression and up to 58 percent for anxiety in some series.^[1] A family history of medication overuse headache or of substance use disorder has been associated with roughly a threefold risk.^[1]

Withdrawal, switching, or neither: what does the evidence support?

Until recently the default advice was to withdraw the overused agent and start prevention. The Medication Overuse Treatment Strategy trial complicates that in a useful way.

MOTS randomised adults with chronic migraine and medication overuse to one of two strategies: migraine preventive therapy plus switching off the overused medication to an alternative used with limited frequency, or migraine preventive therapy without switching or limiting the overused drug. Roughly 720 patients were enrolled across the two arms. Over the first 12 weeks, the preventive-only strategy was non-inferior on moderate to severe headache day frequency.^[6]

The practical reading is not that overuse stopped mattering. It is that you have a second acceptable route when a patient is unwilling or unable to face withdrawal immediately. Starting or optimising prevention first, and revisiting the acute agent once headache frequency falls, is a defensible plan rather than a compromise.

Three caveats keep that from becoming a general licence. First, MOTS enrolled patients with chronic migraine, so it does not speak to every headache phenotype. Second, the non-inferiority window was 12 weeks, which says little about a patient in year six of daily use. Third, and most importantly, the trial’s logic does not extend to the sedative-hypnotic classes. Continuing a barbiturate or an opioid indefinitely is a different risk calculation from continuing a triptan, and it is where the safety argument, not the headache-frequency argument, governs.

European guidance has generally favoured a more direct approach, recommending abrupt discontinuation of the overused medication with prophylaxis initiated just before or at the time of discontinuation, with close follow-up.^[1] A German S1 guideline on management of medication overuse and medication overuse headache similarly supports structured withdrawal with symptomatic cover.^[7] Reasonable clinicians land in different places, and the patient’s own preference carries real weight in a condition where motivation predicts outcome.^[1]

How should the taper differ by drug class?

This is the part where a single protocol will not serve. The approach splits cleanly along one line: whether the overused agent produces physical dependence with a dangerous withdrawal syndrome.

For non-sedative agents (triptans, NSAIDs, simple analgesics, ergots): start or optimise preventive therapy, then either switch to an alternative class capped at no more than two days a week, or taper the acute medication as headache frequency falls in response to prevention. Where headaches persist through the transition, bridge with a long-acting NSAID.^[1] Preventives are typically started at low dose at the point of withdrawal and titrated, with the choice guided by headache phenotype, comorbidity, prior trials and adverse-effect profile. Beta-blockers, calcium channel blockers, tricyclics and anticonvulsants have all been used; topiramate has evidence in this population and performs better in patients with chronic migraine.^[1] OnabotulinumtoxinA subgroup analyses from the PREEMPT programme showed benefit in patients with medication overuse, though those trials did not include opioid overusers.^[1]

For opioids, barbiturates and benzodiazepines: taper gradually, typically over two to four weeks, when higher doses have been used. Lower doses may be discontinued more directly, but that is a judgement made against the individual’s dose and duration.^[1] Abrupt cessation of a barbiturate after sustained use can produce anxiety, tremor, autonomic instability, delirium and seizures, and tapering under supervision is the recommended mitigation.^[8]

Butalbital specifically has a well-described alternative to a stepwise dose reduction. Substituting phenobarbital exploits its long half-life, which supplies a natural taper without repeated dose adjustments. Loder and Biondi described an oral phenobarbital loading protocol for withdrawing patients with headache from butalbital compounds and concluded that management can be simplified this way.^[9] The working equivalence clinicians use is 30 mg of phenobarbital for 100 mg of butalbital, which puts a maximum of about 90 mg of phenobarbital against 300 mg of butalbital.^[1] Note also that butalbital’s own plasma half-life is roughly 35 hours, so steady-state accumulation is meaningful in daily users.^[10]

For opioid withdrawal, a transdermal clonidine patch at 0.1 to 0.2 mg per 24 hours, applied weekly for one to two weeks, has been used to manage withdrawal symptoms.^[1]

One further consideration applies to any combination product containing acetaminophen. Withdrawal is a natural moment to audit total daily acetaminophen exposure across every source, given the threshold for severe hepatic injury and the ease with which prescription combinations and over-the-counter products stack.^[11] Butalbital combinations also carry caffeine, so an abrupt stop layers a caffeine-withdrawal headache onto the picture, typically beginning within 24 hours of interruption in regular consumers of more than 200 mg a day.^[12]

What bridge therapy is worth using?

Transitional treatment is indicated for patients with severe or frequent headache who are likely to struggle during the withdrawal window. The governing rule is that the bridge agent must come from a class other than the one being withdrawn.^[1]

• NSAIDs. Naproxen, indomethacin and ketorolac have all been used, with tizanidine described as a useful adjunct to NSAID therapy.^[1]
• Antiemetics and neuroleptics. Prochlorperazine, metoclopramide, promethazine, diphenhydramine and chlorpromazine have roles during this phase.^[1]
• Corticosteroids. The evidence here is genuinely mixed. Several randomised studies have found minimal or no effect of steroids on withdrawal headache, so the enthusiasm they sometimes attract is not well supported.^[1]
• Infusion protocols. For more complex cases, dihydroergotamine with metoclopramide has been used, with the well-known vascular-disease contraindication; intravenous lidocaine is an alternative in patients where DHE is unsuitable. Hydration, antiemetics, NSAIDs and magnesium are all used supportively.^[1]

Setting is worth a thought too. Dropout rates are higher in outpatient than inpatient withdrawal programmes, which is an argument for front-loading contact in the first fortnight rather than for admitting people.^[1] A scheduled call at day three and a visit at week two costs little and lands at the two points where patients most often abandon the plan.

What outcomes should you set expectations by?

Give patients real numbers. They tolerate a difficult fortnight better when they know what it buys.

Successful withdrawal is generally reported in around 50 to 70 percent of patients at one year, and remaining off the overused agent at one year is itself a good predictor of longer-term success.^[1] A structured multi-centre programme combining detoxification with prophylaxis reported that about two thirds of patients were no longer overusing medication at the end of the protocol and roughly 47 percent had reverted to an episodic headache pattern, with parallel reductions in disability, depression and anxiety scores.^[1]

Relapse deserves equal honesty. Rates in the range of 40 to 50 percent have been reported over six years.^[1] Recognised predictors include tension-type headache as the underlying disorder, longer duration of regular intake, a higher number of acute treatments, smoking and alcohol use, return to the previously overused drug, and no improvement by around two months.^[1] That last one is a practical trigger: if a patient is two months in with nothing to show, revisit the plan rather than wait.

The most useful intervention against relapse may be the least pharmacological. In one series, 42 percent of patients relapsed by three years, falling to 12.5 percent when behavioural therapy was added to the treatment package.^[1] Addressing mood, sleep, smoking and inactivity alongside the taper is not adjunctive nicety; it changes the outcome.

For related protocols, see our provider guides to deprescribing opioids in chronic pain, NSAIDs in older adults, and hyperbolic tapering. Patients coming off a butalbital combination may find our patient-facing guide to coming off butalbital useful alongside the consultation. Our deprescribing playbook sets out the general framework, you can see how a deprescribing consultation is structured, refer patients through find a provider, or browse everything on the MyDeprescriber blog.

Frequently asked questions