Knowledge of environmental triggers for migraine attacks is limited and has mostly been acquired by studies using emergency room (ER) visits. However, it is unlikely that ER visits are a random sample of migraine events, even within strata of migraine severity. Additionally, time lags between attack onset and ER visits may vary across the population, posing challenges for assessing causal links of migraine with community-level or ecologic exposures.

Our objective was to assess the relationship between demographic and geographic measures and self-reported migraine-related ER visits.

We analyzed a targeted non-probability survey of ER use related to migraine in Canada and the USA. The 18-question online survey addressed ER use and behaviors related to recording attacks.

The final dataset included 389 respondents (Canada = 164 [42.2%], USA = 225 [57.8%]); 51 (13.1%) were Migraine Buddy app users who shared their diaries. In both countries, participants reported similar migraine symptoms. Barriers to attending the ER included cost and wait times. There was more variability in delays between attack onset and arrival to the ER than between onset and recording in the smartphone app. Younger participants and participants living in Canada were significantly more likely to present to the ER.

The sample of patients presenting to the ER for migraine may be biased toward younger patients and depend on the jurisdiction. Smartphone app records may have fewer barriers to creation and more consistent time lags compared to ER visit records.