Common cluster headache treatment mistakes

Most cluster headache patients spend years on the wrong treatments before they find the right ones. And even once they're diagnosed, the right treatments are often used incorrectly. This chapter flags the most common mistakes patients and doctors make.

If you've just been diagnosed, also check out the abortive, preventive, and bridge chapters.

---

At a glance

Treatments that don't work for cluster headache

Treatment	Why it fails
Over-the-counter painkillers (ibuprofen, acetaminophen, aspirin)	Wrong mechanism. Clusters don't respond to ordinary painkillers.
Oral triptan tablets (sumatriptan, rizatriptan, eletriptan)	Too slow. The attack peaks before the drug is absorbed.
Opioids (codeine, tramadol, oxycodone, fentanyl)	Don't abort attacks; can cause rebound and dependence.
Ergotamine tablets, butalbital combinations	Too slow.
Decongestants, antibiotics for "sinus"	Based on misdiagnosis. Cluster headache is not sinusitis.
Oxygen concentrators or nasal cannulas	Wrong device. Purity and flow are too low.
Daith piercings, craniosacral therapy, acupuncture	No mechanism; no evidence.

Treatments that work but are commonly used wrong

Treatment	Common mistake	Fix
Oxygen	Wrong flow, wrong mask, wrong device	See the oxygen guide
Sumatriptan	Pills instead of injection	Use the injection or a nasal spray
Verapamil	Dose is too low	Target 360–720 mg/day with ECG monitoring
DMT	Vapor never reaches the lungs	Draw into your mouth, then breathe in fresh air to pull it deep
Steroid bridge	Used alone with no preventive	Take a preventive (usually verapamil) simultaneously
Melatonin	Dose is too low	Use 10 mg at bedtime
Lithium	No baseline labs, no level monitoring	Insist on the full lab protocol
Psychedelics (psilocybin, LSD, 5-MeO-DALT)	Interfering medications still in your system	Clear triptans, ergots, antidepressants, etc. at least 5 days before dosing

---

Treatments that don't work for cluster headache

Over-the-counter painkillers

Ibuprofen, acetaminophen (paracetamol), aspirin, and naproxen do not touch a cluster headache attack. Many patients have been told to "take some ibuprofen and lie down," but cluster headache pain does not respond to these drugs. The pathways that drive a cluster attack are different from the ones that ordinary painkillers act on. Even if you took them by injection, they wouldn't stop an attack.

Oral triptan tablets

This is a subtle and frustrating mistake, because triptans do work for cluster headache, just not in pill form. Subcutaneous sumatriptan injection and zolmitriptan or sumatriptan nasal sprays are first-line abortives in every major guideline.^[1][2] Sumatriptan tablets aren't. If a doctor offered you sumatriptan tablets and they didn't help, that doesn't mean triptans don't work for you, but that you should get injections or sprays instead.

Oxygen concentrators and nasal cannulas

Oxygen works, but only the right oxygen delivered the right way. Concentrators (the machines that pull oxygen out of room air) cap out around 90–95% purity and 5–10 L/min flow, which isn't enough. Nasal cannulas (the soft tubes that sit in the nostrils) deliver far too little oxygen for cluster headache. Patients who say "oxygen didn't work for me" usually got a concentrator and a cannula from their supplier, which physically can't deliver enough.^[5] See the oxygen guide for what does work.

Opioids

Emergency rooms commonly hand opioids (codeine, tramadol, hydrocodone, oxycodone, morphine, fentanyl) to cluster headache patients, and that's the most damaging mistake on this list. Opioids don't abort attacks, they cause rebound and dependence with repeated use, and the patient community is unanimously hostile to them. In an international survey of more than 2,000 cluster headache patients, only 6% rated opioids as "complete or very effective."^[3] Long-term opioid use makes the underlying cluster headache course worse over time, and it can trigger opioid-induced hyperalgesia, in which the drug itself starts producing more pain.

If you've been prescribed opioids for cluster headache, ask your doctor about oxygen and subcutaneous sumatriptan instead. If you're already taking opioids, do not stop them abruptly without medical supervision.

Ergotamine tablets and butalbital combinations

Older oral ergot pills (Cafergot, Ergomar) and combination painkillers containing butalbital (Fioricet, Fiorinal) used to be common prescriptions for headaches generally. They're too slow for cluster headache attacks, and triptans and DHE injections have largely replaced them.

Decongestants and antibiotics for sinus

Because cluster headache attacks come with tearing, a runny nose, and one-sided facial pain, doctors regularly misdiagnose them as sinusitis.^[4] Patients then get round after round of antibiotics, decongestants, and sinus rinses. None of these touch cluster headache pain. If you've been treated for "chronic sinus infections" for years without your scans ever changing, and the headaches still come in clusters of identical attacks at the same time of day, it's worth re-opening the diagnosis.

Daith piercings, craniosacral therapy, acupuncture, electronic pulse devices

Daith piercings, craniosacral therapy, acupuncture, and various small electronic pulse or "anti-migraine" devices marketed for pain show up in cluster headache communities periodically. None has evidence behind it. Patients specifically report that craniosacral manipulation can trigger attacks.

---

Treatments that work but are commonly used wrong

When a cluster headache treatment fails, the problem is usually in how it's being used, not in the treatment itself.

Oxygen

Oxygen is the cluster headache treatment most often used wrong. When patients say "oxygen didn't work for me," the cause is almost always one of these:

• Wrong device. Oxygen concentrators (instead of tanks) and or nasal cannulas (instead of non-rebreather masks) deliver too little oxygen, and should be avoided.
• Flow too low. You need an oxygen setup that delivers at least 15 L/min, and sometimes 25 L/min for stubborn attacks.
• Wrong mask. A standard surgical mask or a "partial-rebreather" lets room air in and dilutes the oxygen. You need a true non-rebreather mask with a reservoir bag and intact one-way side valves, or a demand valve.
• Wrong breathing technique. Lying down with shallow breaths usually fails. Sit upright. Breathe deeply and rapidly, almost like hyperventilating, as soon as the attack starts.
• Stopping too early. Stay on the oxygen for three to five minutes after the pain is gone, or the attack can come back.

You can find the full details in the oxygen guide.

Sumatriptan

Three common mistakes:

• Pills instead of injection. Sumatriptan tablets are too slow for cluster headache. Ask specifically for the 6 mg subcutaneous autoinjector (or for vials, if you plan to split the 6 mg dose into smaller doses).
• Hitting monthly limits. Insurance plans typically cover only four to eight injections a month, well below what most cluster headache patients need. A widely used community workaround is to split each 6 mg dose into two or three smaller doses (around 2–3 mg per shot) using an insulin syringe. Clinical trials support this practice. See the sumatriptan section of the abortive chapter for the technique.
• Overusing it over time. The daily ceiling is two 6 mg injections in any 24-hour period (12 mg total). Going above that risks chest pain and other cardiovascular side effects. And even within the limit, using sumatriptan daily for weeks can worsen the cycle: many patients report that attacks become more frequent and more intense with heavy use, and a small case series found the same pattern.^[6] Use oxygen when you can, and save sumatriptan for the attacks oxygen doesn't catch.

DMT

DMT works only if you get it into your lungs. The most common mistake is treating a DMT vape like a cigarette: drawing the vapor into your mouth and exhaling it straight back out without ever getting it deep into the lungs.

The right technique has four steps:

1. Draw gently for 2 to 3 seconds. Suck the vapor into your mouth first, like sipping a thick milkshake through a straw. You should feel warm vapor filling your mouth.
2. Remove the pen from your lips.
3. Breathe in deeply through your mouth. That pulls the vapor from your mouth down into your lungs, along with fresh air.
4. Hold for 10 seconds or more, then exhale slowly.

There are two other things to get right:

1. The vape's voltage: if it's too low, the DMT doesn't vaporize properly; if it's too high, the DMT might burn.
2. Drug interactions: DMT can cause seizures when combined with lithium and can trigger serotonin syndrome with MAOIs or triptans.

The DMT guide covers settings and equipment in detail, and the DMT safety chapter has the full interaction table.

Verapamil

Verapamil is the preventive most often left at too low a dose. The cluster headache dose is two to four times the dose used for blood pressure, and the most common mistake pattern goes like this:

1. A primary care doctor starts you at 80–120 mg per day (a typical blood-pressure starting dose).
2. They never raise the dose.
3. You report it "isn't working" at that dose. Which it isn't, because you're nowhere near the effective range.

The right approach is to start at 240 mg per day, increase by 80 mg every one to two weeks (with an ECG before each increase), and target 360 to 720 mg per day. Some refractory patients need up to 960 mg or even 1,200 mg per day.

ECG monitoring isn't optional. About half of patients on these doses develop changes in their heart rhythm, and a small minority develop more serious ones.^[7] The verapamil section of the preventive chapter has the full protocol.

Psychedelics as preventives (psilocybin, LSD, 5-MeO-DALT)

Patients use small doses of psilocybin (in mushrooms), LSD, or 5-MeO-DALT, taken every five days, to break or prevent a cluster cycle. The full protocol, dosing, and safety information are in the psychedelics guide. Three mistakes account for most of the cases where the protocol doesn't work.

• Not stopping interfering medications. This is the single biggest reason the protocol fails. Triptans, ergots, verapamil, SSRIs and SNRIs, tricyclic antidepressants, opioids, anti-convulsants, and corticosteroids can all blunt or block the effect of psychedelics. The conservative rule is to stop these at least five days before your first dose (SSRIs need much longer than five days; some have half-lives of weeks). Never stop a prescription medication without first talking to the doctor who prescribed it.
• Dosing too often. Tolerance after one psychedelic dose lasts about five days. Dosing sooner doesn't work and may actually slow things down. Stick to the five-day spacing, even if you're tempted to dose sooner.
• Reaching for a triptan during a "slap-back attack." In the day or two after a dose, some patients get extra attacks, often more frequent and sometimes more intense than their usual ones. Patients call these "slap-back" or "rebound" attacks. Using a triptan to deal with one resets your clock, since you'd need to wait another five days before the next preventive dose works. Use oxygen or DMT instead to abort attacks.

One additional warning that's less about effectiveness and more about safety: never combine these substances with lithium or MAOI antidepressants. Lithium can cause seizures; MAOIs can cause serotonin syndrome. The psychedelics safety chapter has the full interaction table.

Steroid bridges that don't bridge to anything

A prednisone taper or a greater occipital nerve block is called a bridge because they temporarily stop attacks while a longer-acting preventive starts to work. If you don't start a preventive together with the bridge, the attacks come straight back when the steroid wears off, and you're back where you started, only now with steroid side effects.

The fix is to start the bridge and the preventive (usually verapamil) on the same day (or, ideally, you start the preventive before your cycle starts). The bridge chapter covers the timing.

A related mistake is doing repeated steroid bridges every cycle, year after year, without ever asking why the preventive isn't working. Cumulative steroid exposure causes weakened bones, weight gain, high blood pressure, and at high cumulative doses, hip-joint damage. If you're on a third or fourth prednisone taper in a year, the question to bring to your doctor is "why isn't my preventive working?" not "can I have another taper?"

Melatonin at the wrong dose

The effective dose for cluster headache prevention is 10 mg at bedtime, sometimes 15 mg or higher. Many patients try a 1 to 3 mg drugstore sleep-aid bottle, see no effect, and conclude melatonin doesn't work for them. They've taken roughly a tenth of the cluster headache dose. See the melatonin section of the preventive chapter.

Lithium without monitoring

Lithium has a narrow therapeutic window: it stops working below a certain blood level and becomes toxic above another. You need baseline thyroid, kidney, and electrolyte tests before starting, blood levels checked one week after every dose change, and periodic monitoring afterwards. Lithium also interacts dangerously with common over-the-counter NSAIDs (ibuprofen, naproxen), with ACE inhibitors for blood pressure, and with dehydration, any of which can push your blood level into the toxic range. If a doctor prescribes lithium without these checks and without warning you about NSAID interactions, ask for a different prescriber.

---

Other common mistakes patients and doctors make

• No attack diary. Without a record of timing, side, duration, and triggers, you have no leverage for adjusting a preventive dose or noticing patterns.^[8] A paper notebook works, but a purpose-built app makes it easier. MyClusters is a free cluster-headache-specific tracker app built by a patient.
• Drinking alcohol during a cycle. Alcohol is the single most reliable cluster headache trigger during an active bout, often setting off an attack within 30 to 90 minutes. Outside of cycles, most patients can drink normally.
• Stopping verapamil the moment the cycle ends. Specialists typically continue verapamil for at least two weeks past the last attack before tapering. Stopping at the first pain-free day often restarts the cycle.
• Having an abortive but no preventive (or the other way around). Most patients need both: a fast abortive (oxygen, sumatriptan, DMT) for individual attacks, and a preventive running in the background to reduce how many attacks come in the first place.

---

References

1. ↩ May A, Evers S, Goadsby PJ, Leone M, Manzoni GC, Pascual J, et al. (2023). European Academy of Neurology guidelines on the treatment of cluster headache. European Journal of Neurology, 30(10), 2955–2979. doi:10.1111/ene.15956
2. ↩ Robbins MS, Starling AJ, Pringsheim TM, Becker WJ, Schwedt TJ (2016). Treatment of Cluster Headache: The American Headache Society Evidence-Based Guidelines. Headache, 56(7), 1093–1106. doi:10.1111/head.12866
3. ↩ Schor LI, Pearson SM, Shapiro RE, Zhang W, Miao H, Burish MJ (2021). Cluster headache epidemiology including pediatric onset, sex, and ICHD criteria: Results from the International Cluster Headache Questionnaire. Headache, 61(10), 1511–1520. Link
4. ↩ Klapper JA, Klapper A, Voss T (2000). The misdiagnosis of cluster headache: a nonclinic, population-based, internet survey. Headache, 40(9), 730–735. Link
5. ↩ Cohen AS, Burns B, Goadsby PJ (2009). High-flow oxygen for treatment of cluster headache: a randomized trial. JAMA, 302(22), 2451–2457. Link
6. ↩ Rossi P, Di Lorenzo G, Formisano R, Buzzi MG (2004). Sub-cutaneous sumatriptan induces changes in frequency pattern in cluster headache patients. Headache. Link
7. ↩ Chan CK, Lambru G, Matharu MS (2013). Fast and slow titration of verapamil in cluster headache: comparison of electrocardiographic abnormalities. Journal of Headache and Pain, 14(Suppl 1), P45. Link
8. ↩ Rozen TD, Fishman RS (2012). Cluster headache in the United States of America: demographics, clinical characteristics, triggers, suicidality, and personal burden. Headache, 52(7), 1079–1087. Link