This was a retrospective, observational, claims-based study, which used data from the Merative™ MarketScan® Commercial and Medicare Databases. Access to the data was granted through an internal agreement with Eli Lilly and Company and Merative. These databases contain data for several million individuals (employees) and their spouses and dependents who are covered by employer-sponsored private health insurance in the US. All database records are de-identified and fully compliant with US patient confidentiality requirements, including the Health Insurance Portability and Accountability Act of 1996.

The index date was defined by the first date of galcanezumab or rimegepant claim between June 1, 2021, and June 30, 2022. The 12 months prior to the index date comprised the baseline period and the 12 months after the index date comprised the follow-up period (Fig. 1). Patients were assigned to one of two cohorts based on the migraine treatment initiated on the index date: galcanezumab or rimegepant cohort.

Institutional review board approval to conduct this study was not required, as the study used only de-identified patient records and did not involve the collection, use, or transmittal of individually identifiable data.

Patients were included if they were ≥ 18 years old on the index date and had ≥ 1 claim for galcanezumab 120 mg or rimegepant 75 mg between June 1, 2021, and June 30, 2022 (for rimegepant, only claims with a quantity of ≥ 15 were counted, as rimegepant is recommended for administration every other day for preventive use). Patients were continuously enrolled in medical and pharmacy benefits in the 12 months before and after the index date (baseline and follow-up periods) and had ≥ 1 medical claim with an International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) diagnosis code for episodic migraine during the baseline period or on the index date.

Patients were excluded from the study if there was evidence of cluster headache diagnosis, pregnancy, cancer, and epilepsy diagnosis during the study period. For the galcanezumab cohort, patients with any claims for other CGRP mAbs (erenumab, fremanezumab, eptinezumab) and gepants indicated for prevention (rimegepant quantity of ≥ 15 and atogepant) during the baseline period or on the index date were excluded. For the rimegepant cohort, patients with any claims for CGRP mAbs (erenumab, fremanezumab, galcanezumab, eptinezumab) and atogepant during the baseline period or on the index date were excluded.

Demographics measured on the index date included age, sex, geographical region of residence, insurance type, and type of provider. Provider type was determined based on the closest medical claim in the 45 days prior to and including the index claim. Charlson Comorbidity Index, a weighted composite score to measure mortality risk based on 19 comorbid conditions, was obtained during the 12-month baseline period. Comorbid medical conditions were identified by the presence of ≥ 1 inpatient or any position on non-diagnostic outpatient medical claim with an ICD-10-CM diagnosis code during the 12-month baseline period. Acute and preventive medications, identified by the presence of ≥ 1 outpatient prescription claim or medical claim with a procedure code for the administration of the medications, were reported during the 12-month baseline period. Acute medication included: antiemetics, acetaminophen, nonsteroidal anti-inflammatory drugs (NSAIDs), barbiturate (butalbital), ergotamine, isometheptene, opioids, triptans, lasmiditan, ubrogepant, and rimegepant (quantity of ≤ 8 or “1 pack”). Preventive medications included anticonvulsants, tricyclic antidepressants, beta-blockers, angiotensin receptor blockers, calcium channel blockers, botulinum toxin A, CGRP mAbs, and gepants.

Measures of all-cause and migraine-related HCRU during the 12-month baseline and follow-up periods included: number of inpatient admissions, emergency room visits, outpatient office visits, other outpatient services, and outpatient pharmacy claims. Migraine-related HCRU and costs were calculated based on the presence of a migraine diagnosis on the claim line (in the primary position on inpatient claims and any position on outpatient claims) and migraine treatments, while all-cause HCRU and direct costs were calculated based on all claims in all sites of care.

Discontinuation of index medication was defined as failure to refill the index medication within 60 days after the depletion of the days’ supply from previous fills. The date of discontinuation was defined as the date of run out of days’ supply of the last prescription filled prior to the first observed gap in therapy. Sensitivity analysis using a 30-day gap was also reported.

Categorical variables were presented as the count and percentage of patients in each category. Continuous variables were summarized by the mean, standard deviations (SDs), and medians. Paired t-test and McNemar test were used to compare the HCRU and costs between the baseline and follow-up periods. As cost data were skewed, both mean and median were reported. Cox proportional hazards regression analyses were conducted to estimate the hazard ratio of discontinuation for the galcanezumab and rimegepant cohorts. Kaplan–Meier survival curves were used to show the probability of discontinuation of each treatment.

Greedy propensity score matching (1:1, with a caliper of 0.02, standardized distance on the propensity score scale between 0 and 1) was used to minimize baseline differences between patients initiating galcanezumab and rimegepant cohorts. Cohorts were matched on the following 34 covariates: age, sex, region, provider type, and the baseline characteristics including back pain, coronavirus disease 2019, chronic pain, gastroesophageal reflux disease [GERD], headache, hyperlipidemia, nausea, obesity, stroke, anxiety, asthma, depression, diabetes, fibromyalgia, osteoarthritis, sleep disorder, any preventive migraine medication use, beta blocker use, anti-epileptic use, tricyclic antidepressant use, any acute migraine medication use, CGRP/gepant use, triptan use, anti-emetic use, NSAID use, opioid use, ubrogepant use, rimegepant use (≤ quantity of 8), all-cause total medical + pharmacy costs, and migraine-related total medical + pharmacy costs. Patient factors are shown in Table 1 as covariates in the model to generate the propensity score based on previous studies and clinical relevance.

We assessed the balance of baseline characteristics between the cohorts using standardized differences (std diff), quantifying the difference in means between the two cohorts relative to their pooled SD. A std diff greater than 10% is considered indicative of an imbalance between cohorts in observational studies [22]. This threshold was applied to identify covariates that needed adjustment reducing potential bias. A significance level of 0.05 was specified a priori as the threshold for statistical significance for any statistical comparisons across groups.