Which Acute Migraine Medications Work Best? Taking a Big Data Approach

By Kayt Sukel | February 22, 2024 | Posted in

Researchers use e-diary smartphone data for head-to-head comparisons of 25 acute migraine medications. Reassuringly, the results are consistent with current clinical guidelines.

There are dozens of different medications, available both over the counter (OTC) and by prescription, to help those living with migraine. But finding the best option can be a challenge: OTC options may not be powerful enough, and physicians may not always know the best drug to prescribe to their patients, in part because of a lack of controlled studies comparing different drugs and drug classes.

Now, researchers including co-first authors Chia-Chun Chiang, a headache specialist at Mayo Clinic, Rochester, Minnesota, US, and data scientist Xuemin Fang, Kanagawa University of Human Services, Kawasaki, Japan, and IBM Client Engineering, Tokyo, Japan, have taken a big data approach to directly compare 25 acute migraine medications, using patient-generated medication-outcome data from an e-diary smartphone application called Migraine Buddy. Their analysis – based on nearly 11 million migraine attacks – found that Migraine Buddy users reported the greatest relief from triptans, followed by ergots, and then antiemetics.

The results held up when the team compared users from the US, UK, and Canada. And, notably, the findings were in line with current clinical guidelines, revealing an encouraging alignment of official recommendations with what patients say works for them.

Matthew Robbins, New York-Presbyterian and Weill Cornell Medicine, New York City, US, a neurologist who co-wrote an accompanying editorial but was not involved with the study, told Migraine Science Collaborative in an interview that the use of real-world data is a key strength of the study.

“Most studies that look at whether a drug is effective are very controlled, like the clinical trials used to determine whether different treatments will be approved by the [US] Food and Drug Administration,” Robbins said. “There is a lot we can learn from those studies, but they tend to only provide one-to-one comparisons. Sometimes it’s the real-world data that provide those little clues and nuances into how we should compare approved treatments, how we should sequence them in practice, and how we can better match up the right treatment for the right patient.”

The study and editorial appeared in the December 12, 2023, issue of Neurology.

Finding a way to compare medications
Historically, researchers have only been able to compare the efficacy of different medications through one-to-one comparisons in costly clinical trials, or via meta-analyses. A danger of the latter, which combine and analyze results from different clinical trials, is that they may be comparing apples and oranges, by bringing together studies with different individual designs.

Chiang said that a big data approach analyzing real-world, patient-generated data can provide information that is otherwise hard to come by.

“It’s challenging to compare multiple medications in a single study, and our ability to do so is quite limited by the data sources we have,” she said. “We know that real-world patient experience is important. And we can take this real-world experience with different medications so that we can compare many, many different medication options head to head.”

Chiang and colleagues conducted a retrospective analysis of 10,842,795 migraine attack records entered into Migraine Buddy. In the app, users can document the start and end time of a migraine attack, pain intensity, the treatment they used, and whether that treatment was helpful. The app also records the sex, age, and geographical location of each user.

According to Chiang, it was a challenge to clean the data for analysis, but the investigators were eventually able to classify each attack treatment entry into a medication group.

“We went into the raw data to go over each medication entry and made sure it was correct, as some people selected a drug from a drop-down menu and others entered it as free text,” she said. “We understood there would be different formulations and doses for each medication. And we spent a lot of time considering which medication types to include because of the complexity of this patient-generated dataset.”

The researchers ultimately decided to include 25 different medications, ranging from triptans like sumatriptan to opioids like oxycodone. Chiang said she and her colleagues did not include any combination treatments, with the exception of acetaminophen, aspirin and caffeine, the formulation of Excedrin Migraine and one of the most popular OTC medications on the market. The analysis also excluded calcitonin gene-related peptide (CGRP) antagonists and ditans, as the data were collected between 2014 and 2020, before such drugs became readily available.

“We reviewed the American Headache Society’s level of evidence guidelines and only included medications that had established efficacy or were probably effective per their guideline recommendations,” said Chiang. “That led to the inclusion of seven different classes of medications including triptans, ergots, non-steroid anti-inflammatory drugs [NSAIDs], acetaminophen, antiemetics, opioids, and combination analgesics.”

The primary outcome of the study would be how users rated the effectiveness of the medication they took, with the options being “Helpful,” “Somewhat helpful,” or “Unhelpful.” The researchers considered “Helpful” a positive outcome, and the other two options a negative outcome.

Triptans, ergots, and antiemetics lead the pack
The group was able to run an analysis of 4.7 million medication-outcome pairs based on more than three million attacks in 278,006 app users. Of the approximately 96,000 users who provided both sex and age information, roughly 91% of users were women, and the average age for men was 38 years, and 36 years for women.

Using ibuprofen as a reference (the most commonly used drug by app users), the researchers discovered that triptans were the most effective drug class, with a mean odds ratio (OR) of 4.8, meaning that triptans were almost five times more likely to be helpful compared to ibuprofen. Among triptans, the most effective ones were eletriptan, zolmitriptan, and sumatriptan.

Ergots (a mean OR of 3.02) and antiemetics (a mean OR of 2.67), were the second and third most helpful classes of medications. In contrast, acetaminophen, NSAIDS other than ibuprofen, combination analgesics, and opioids were only as effective, or less effective, than ibuprofen.

The team then ran subgroup analyses for Migraine Buddy users in the US, the UK, and Canada, finding similar ORs for the different medications in those regions.

“We were delighted to see that the results were consistent in the United States, United Kingdom, and Canada, and that they were also quite consistent with clinical guideline recommendations for migraine treatment,” Chiang said. “The results were very reassuring and really speak to the efficacy of migraine-specific medications.”

Many answers, more questions
Robbins said he was struck by the “massive” number of patients and migraine attack records involved in the study and appreciated the innovative methodology of the work. But he also pointed to some shortcomings in the data.

“This is the largest study of migraine attacks that’s ever been published. The use of this app really democratizes access to this kind of data and opens up possibilities for further study,” he said. “Yet, there’s a lot of information we’d like to know about patients with migraine because this kind of data capture just isn’t specific. We don’t have much in terms of race or ethnicity, gender minorities, or others who may be disproportionately affected by migraine and may be underserved. There’s also no nuance about when the medication was taken in terms of the attack progression or what dose was taken. There’s a lot more we could learn about individual treatment that simply can’t be captured here.”

Chiang said she would like to do a similar analysis that includes gepants in the future. Robbins added that he’d like to see an analysis that brings in CGRP inhibitors.

In the meantime, Chiang hopes that clinicians will feel more confident in prescribing triptans, ergots, and antiemetics to patients.

“This kind of real-world, patient-reported data can reassure clinicians to leverage and utilize migraine-specific medications to help patients,” she said. “I also hope clinicians and researchers recognize the potent insights gained from large-scale, real-world patient-experience data, which can significantly enhance our comprehension of treatment outcomes and guide future research directions.”

Kayt Sukel is a freelance writer based outside of Houston, Texas.

Simultaneous comparisons of 25 acute migraine medications based on 10 million users’ self-reported records from a smartphone application.
Chiang et al.
Neurology. 2023 Dec 12;101(24):e2560-70.

Millions of migraine attacks, many answers, more questions.
Friedman SA, Robbins MS.
Neurology. 2023 Dec 12;101(24):1089-90.

Image credit: 123RF Stock Photo.

Sign Up For An MSC Newsletter

Kayt Sukel is a passionate traveler and science writer who has no problem tackling interesting (and often taboo) subjects spanning love, sex, science, technology, travel and politics. Her work has appeared in the Atlantic Monthly, New Scientist, USA Today, Pacific Standard, the Washington Post, ISLANDS, Parenting, the Bark, American Baby, National Geographic Traveler, and the AARP Bulletin, among others. She has written stories about out-of-body experiences, artificial intelligence in medicine, new advances in pain treatments, and why one should travel to exotic lands with young children.

She is the author of two books: Dirty Minds: How Our Brains Influence Love, Sex, and Relationships (re-titled as This Is Your Brain on Sex: The Science Behind the Search for Love in paperback) and The Art of Risk: The New Science of Courage, Caution, and Chance.



  • Content Types

  • Paper Categories

  • Paper Dates

Recent News

Recent Papers