To describe demographic and clinical characteristics, healthcare resource use, costs, and treatment patterns in three migraine cohorts.

This retrospective observational study using administrative data examined patients with episodic migraine (EM), chronic migraine (CM) (without medication overuse headache [MOH]), and medication overuse headache in Alberta, Canada. Migraine patients were identified between 2012 and 2018 based on ≥ 1 diagnostic codes or triptan prescription. Patients with CM were defined using parameter estimates of a logistic regression model, and MOH was defined as patients with an average of ≥ 15 supply days covered of acute medications. EM was defined as patients without CM or MOH. Study outcomes were summarized using descriptive statistics.

Patients with EM (n = 144,574), CM (n = 27,283), and MOH (n = 11,485) were included. Higher rates of healthcare use and costs were observed for CM (mean [SD] all-cause cost: ($12,693 [40,664]) and MOH ($16,611.5 [$38,748]) versus episodic migraine ($4,251 [$40,637]). Across all cohorts, opioids were the most dispensed acute medication (range across cohorts: 31.7%–89.8%), while antidepressants and anticonvulsants were the most dispensed preventive medication. Preventative medication classes were used by a minority of patients in each cohort, except anticonvulsants, where 50% of medication overuse patients had a dispensation.

Patients with CM and MOH have a greater burden of illness compared to patients with EM. The overutilization of acute medication, particularly opioids, and the underutilization of preventive medications highlight an unmet need to more effectively manage migraine.

Migraine, including episodic migraine (EM) and chronic migraine (CM), is a common neurological disorder that imparts a substantial health burden.

Understand the characteristics and treatment of EM and CM from a population-based perspective.

This retrospective population-based cross-sectional study utilized administrative data from Alberta. Among those with a migraine diagnostic code, CM and EM were identified by an algorithm and through exclusion, respectively; characteristics and migraine medication use were examined with descriptive statistics.

From 79,076 adults with a migraine diagnostic code, 12,700 met the criteria for CM and 54,686 were considered to have EM. The majority of migraineurs were female, the most common comorbidity was depression, and individuals with CM had more comorbidities than EM. A larger proportion of individuals with CM versus EM were dispensed acute (80.6%: CM; 63.4%: EM) and preventative (58.0%: CM; 28.9%: EM) migraine medications over 1 year. Among those with a dispensation, individuals with CM had more acute (13.6 ± 32.2 vs. 4.6 ± 10.9 [mean ± standard deviation], 95% confidence interval [CI] 7.7-8.3), and preventative (12.6 ± 43.5 vs. 5.0 ± 12.6, 95% CI 6.9-8.4) migraine medication dispensations than EM, over 1-year. Opioids were commonly used in both groups (proportion of individuals dispensed an opioid over 1-year: 53.1%: CM; 25.7%: EM).

Individuals with EM and CM displayed characteristics and medication use patterns consistent with other reports. Application of this algorithm for CM may be a useful and efficient means of identifying subgroups of migraine using routinely collected health data in Canada.